JP5094752B2 - Image processing apparatus and image processing method - Google Patents

Description

  The present invention relates to an image encoding technique.

  There is lossless encoding as a method for suppressing the amount of data without losing information contained in the original image data. For example, DPCM conversion (Difference Pulse Code Modulation) is known as a lossless encoding method. In DPCM conversion, the difference between adjacent pixels in the original image data is taken, entropy is reduced, and then Huffman coding is performed to encode the original image data.

  However, when the correlation between adjacent pixels is low, the amount of code data obtained by Huffman coding may be larger than the amount of original image data. The following methods are known as methods for guaranteeing the maximum data amount when using lossless encoding.

The amount of lossless encoded data is compared with a predetermined threshold. If the amount of data is equal to or greater than the threshold, the original image data is inserted into the encoded stream. If the amount is less than the threshold, the lossless encoded data is encoded. Insert into. Thereby, even when lossless encoding is performed, the upper limit of the memory bandwidth and the data transmission amount can be guaranteed.
JP 2008-092126 A JP 2008-028534 A

  In order for the decoding side to determine whether the data format of the original image data or encoded data is inserted into the encoded stream, an identifier for determining the data format must be inserted into the encoded stream. should not. Further, in order to determine whether the data is the original image data or the encoded data, 1 bit is usually sufficient. In this specification, a method of sequentially adding a fixed-length identifier in order to determine the data format will be referred to as “method A” in this specification.

  When the method A is used, the data length of the identifier is a value obtained by multiplying the number of image divisions by the identifier length. Therefore, the maximum value of the encoded stream length is a value obtained by adding the data length of the previous identifier to the data length of the original image data. That is, since the data length of the identifier is short, the method A has an advantage that the maximum data length of the encoded stream of the entire image can be shortened. However, in general, in an image transmitted from an image sensor, the compression rate hardly exceeds 100% even when lossless encoding is performed, and the frequency of selecting original image data is low. Nevertheless, inserting an identifier for every divided image will unnecessarily increase the data length of the encoded stream.

  As a method of solving this problem, there is a method of inserting an identifier having a different bit length for each data format. This method will be referred to as “method B” in this specification. For example, an identifier is not inserted into encoded data, but an identifier is inserted only into original image data. The identifier of the original image data must be a bit pattern that can never occur as a code code of entropy encoding of lossless encoding. For example, when Huffman coding is used for entropy coding, the maximum length of the Huffman code is 16 bits, so the identifier length needs to be 16 bits or more. As a specific example, when the identifier length is 16 bits, the bit pattern is 0xFFNN (N: arbitrary value) (example: 0xFFB0).

  In addition, identifiers must be inserted at the byte boundary. For this reason, bits (padding bits) must be added so that the encoded stream before the identifier insertion ends at the byte boundary.

  However, in general, in an image transmitted from an image sensor, even when lossless encoding is performed, the compression rate rarely exceeds 100%, so the encoded stream length is guaranteed while ensuring the maximum compression rate. Can be suppressed. However, when the compression rate of all the divided images exceeds 100% and the original image data is selected, the encoded data requires identifier lengths and padding bits for the number of image divisions. Therefore, when original image data is selected for all divided images, it is obvious that the encoding efficiency of the method B is lower than that of the method A.

  The present invention has been made in view of the above problems, and has as its object to provide a technique for solving the problems related to the above two systems and generating an encoded stream that guarantees a compression rate of 100%. To do.

  In order to achieve the object of the present invention, for example, an image processing apparatus of the present invention comprises the following arrangement.

That is, an input image is divided into a plurality of divided images, by performing lossless encoding on the divided image, an image processing apparatus for generating an encoded divided picture of the divided images,
The data amount of the encoded divided image and the data amount of the divided image are compared, and based on the result of the size comparison, either the divided image or the encoded divided image is the divided image. Selecting means for selecting as an encoding result of
First storage control means for storing an encoding result selected by the selection means in an encoded stream;
An identifier for specifying whether the encoding result selected by the selection unit for each divided image constituting the input image is the divided image or the encoded divided image. Second storage control means for storing, in the encoded stream, an identifier having a data structure determined in advance among a plurality of types of data structures required for the identifier ;
Based on the data amount of the encoded stream completed by the first storage control means and the second storage control means, the second storage control means uses the image input next to the input image. Determining means for determining the data structure of the identifier.

  According to the configuration of the present invention, it is possible to solve the problems related to the above two methods and generate an encoded stream that guarantees a compression rate of 100%.

  Preferred embodiments of the present invention will be described below with reference to the accompanying drawings. The embodiment described below shows an example when the present invention is specifically implemented, and is one of the specific examples of the configurations described in the claims.

[First Embodiment]
In the present embodiment, an image processing apparatus having a function for encoding an image and a function for decoding an encoded image will be described. The function of encoding an image is to encode an image read from an image sensor included in a digital camera or digital camcorder or an image received via a network. Then, the data amount of the image is compressed (encoded) to a data amount equal to or less than the data amount at the time of acquisition without losing information included in the image.

  FIG. 16 is a diagram illustrating a hardware configuration example of a digital camera as an image processing apparatus according to the present embodiment. An image to be imaged is formed on an image sensor 1602 such as a CCD or CMOS sensor via a lens 1601. The image sensor 1602 converts the image thus formed into an analog signal, and sends this analog signal to the A / D converter 1603 at the subsequent stage.

  The A / D converter 1603 converts the analog signal received from the image sensor 1602 into a digital signal. When the memory control circuit 1607 acquires the converted digital signal as data (image data) from the A / D converter 1603, the memory control circuit 1607 sends the image data to the image processing unit 1604. The image processing unit 1604 performs known image interpolation processing, color conversion processing, and the like on this image data. For example, when the image data is a Bayer array, the image data is converted into RGB image data.

  When the memory control circuit 1607 acquires the image data processed by the image processing unit 1604 from the image processing unit 1604, the memory control circuit 1607 stores the acquired image data in the memory 1605. The memory 1605 is a memory for temporarily storing captured still image and moving image data, and has an area for storing a predetermined number of still images (a predetermined number of still images). Since the memory 1605 is a readable / writable memory, the memory control circuit 1607 includes a plurality of memory control units dedicated to writing data to the memory 1605 and a plurality of memory control units dedicated to reading data from the memory 1605.

  The image data stored in the memory 1605 is read again by the memory control circuit 1607 and sent to the D / A converter 1608 and the image dividing unit 1610. The D / A converter 1608 converts the image data into an analog signal, and sends the converted analog signal to the image display unit 1609. As a result, the image (captured image) indicated by the analog signal is displayed (reproduced) on the display screen of the image display unit 1609.

  On the other hand, the image dividing unit 1610 divides the image data (input image) received from the memory control circuit 1607 for each encoding processing unit (divided image). Then, the image dividing unit 1610 sends each divided image to the encoding unit 1611. The encoding unit 1611 generates an encoded stream including an encoding result for each divided image and an identifier of the encoding result for each divided image. The memory control circuit 1607 records the encoded stream generated by the encoding unit 1611 in the storage medium 1606. As the storage medium 1606, a removable medium such as an SD card that can be attached to and detached from the image processing apparatus is used.

  The recording mode switch 1622 is used to instruct the start of imaging. When the user presses the recording mode switch 1622, an instruction to start imaging is input to the system control unit 1620.

  Here, this digital camera is provided with a mode dial 1621 for user operation. A mode dial 1621 is used to select either the imaging mode or the playback mode. When the user operates the mode dial 1621 to select the imaging mode, the system control unit 1620 causes the image display unit 1609 to display an image based on the image data of the image obtained through the lens 1601 and also sets the recording mode switch 1622. Enable it. Only when the recording mode switch 1622 is in an effective state can an image be captured. When the user presses the recording mode switch 1622 in the imaging mode, an imaging start instruction is input to the system control unit 1620 as described above. Therefore, the system control unit 1620 performs operation control of each unit constituting the image processing apparatus. The above-described processing is realized. That is, the image data of the image obtained through the lens 1601 is encoded, and a process of recording the encoded data as a coded stream on the storage medium 1606 is realized.

  On the other hand, when the user operates the mode dial 1621 to select a playback mode, the system control unit 1620 controls the operation of each unit constituting the image processing apparatus, and implements each process described below.

  The memory control circuit 1607 sequentially reads the encoded stream recorded in the storage medium 1606 and sends the read encoded stream to the decoding unit 1613. The decoding unit 1613 decodes the encoded stream received from the memory control circuit 1607. The image synthesizing unit 1612 restores one image by collecting the results of decoding of the respective divided images that form one image by the decoding unit 1613. The memory control circuit 1607 acquires the data of each image restored by the image composition unit 1612 from the image composition unit 1612, and sends the obtained image data to the D / A converter 1608. The D / A converter 1608 converts the image data into an analog signal, and sends the converted analog signal to the image display unit 1609. As a result, the image (captured image) indicated by the analog signal is displayed (reproduced) on the display screen of the image display unit 1609.

  The ROM 1623 stores setting data of the image processing apparatus and a computer program to be executed by the system control unit 1620. Further, the ROM 1623 also stores data described as known data in the processing described below. In other words, the system control unit 1620 executes processing using the computer program and data stored in the ROM 1623, thereby controlling the operation of each unit constituting the image processing apparatus. As a result, the image processing apparatus according to the present embodiment realizes each process described below.

<Encoding unit 1611>
FIG. 1 is a block diagram illustrating a detailed functional configuration example of the encoding unit 1611. Although each part shown in FIG. 1 shall be comprised with hardware, you may implement | achieve part or all with a computer program.

  As described above, since a plurality of divided images constituting one image are input to the encoding unit 1611 in units of divided images, the encoding unit 1611 performs encoding processing in units of divided images. . In the following, the operation of the encoding unit 1611 for encoding one divided image will be described, but the same operation is performed for other divided images.

  The lossless encoding unit 107 performs lossless encoding on the divided image (RAW data) sent from the image dividing unit 1610 according to the lossless encoding method. Hereinafter, data of a lossless encoding result (encoded divided image) for RAW data is referred to as “CODE data”. As the lossless encoding method used by the lossless encoding unit 107, any method may be applied as long as the image can be compressed without impairing information (image quality) of the image. For example, as a lossless encoding method, a still image lossless encoding method such as JPEG DPCM conversion, JPEG-LS, or JPEG2000 may be used. A lossless encoding method for I slices in a moving image encoding method such as H.264 may be used.

  The comparison unit 101 compares the data amount of the RAW data input from the image division unit 1610 and the data amount of the CODE data generated by the lossless encoding unit 107 performing lossless encoding on the RAW data. Make a comparison. In other words, this processing is processing for determining whether or not the compression rate of the RAW data by the lossless encoding unit 107 exceeds 100%.

  The selection unit 102 compares the RAW data input from the image division unit 1610 and the CODE data generated by the lossless encoding unit 107 performing lossless encoding on the RAW data. One is selected based on the processing result. In the present embodiment, as a result of the comparison by the comparison unit 101, this RAW data is selected when the compression rate of the RAW data by the lossless encoding unit 107 exceeds 100%, and this RAW data is selected when it does not exceed. CODE data which is a result of lossless encoding of data is selected.

  The identifier insertion unit 105 obtains a magnitude comparison result from the comparison unit 101 to identify whether the selection unit 102 selects RAW data or CODE data. When the selection unit 102 selects RAW data, an identifier indicating the RAW data is stored in the encoded stream together with the RAW data (first storage control and second storage control). On the other hand, when the selection unit 102 selects CODE data, an identifier indicating the CODE data is stored in the encoded stream together with the CODE data. Note that one encoded stream is provided for one image. Then, when one encoded stream is completed (when RAW data or CODE data and its identifier are stored in the encoded stream for all the divided images), the identifier inserting unit 105 outputs the completed encoded stream.

  On the other hand, the counting unit 103 counts the data amount of the encoded stream output from the identifier inserting unit 105. Then, the remaining amount of the data transmission amount that can be output by the image processing apparatus is obtained from the counted data amount.

  The determination unit 104 determines the data structure of the identifier used by the identifier insertion unit 105 for the next image to be encoded based on the counting result obtained by the counting unit 103. That is, the identifier used by the identifier insertion unit 105 for the image of interest is an identifier having a data structure that is determined based on the determination result of the determination unit 104 for the image encoded immediately before the image of interest. In the present embodiment, it is assumed that there are two types of identifier data structures. One method is called method A (first method), and the other method is called method B (second method).

  FIG. 3 is a diagram showing each of the method A and the method B in a table format.

  As shown in FIG. 3, the method A stipulates that a 1-bit identifier is used as an identifier indicating RAW data and a 1-bit identifier is used as an identifier indicating CODE data. That is, when the method A is used, when the selection unit 102 selects RAW data, the identifier insertion unit 105 adds a 1-bit identifier (for example, “0”) to the RAW data. On the other hand, when the selection unit 102 selects CODE data, the identifier insertion unit 105 adds a 1-bit identifier (for example, “1”) to the CODE data.

  Here, when only one method A is used for encoding one image, the data length (bit length) of all identifiers finally included in the encoded stream of this image is (number of divided images × 1 bit). ) Therefore, the data length of the encoded stream for one image is obtained by adding the data length of the identifier (number of divided images × 1 bit) to the total length of the encoded data of all the divided images constituting this image. It becomes. That is, since the data length of the identifier is short, the method A has an advantage that the maximum data length of the encoded stream of the entire image can be shortened. However, since the compression rate of an image transmitted from an image sensor generally does not exceed 100% even when lossless encoding is performed, the selection unit 102 does not frequently select RAW data. Therefore, always providing a 1-bit identifier for each divided image increases the data length of the encoded stream.

  As a method of solving this problem, there is a method of using identifiers having different bit lengths for RAW data and CODE data. The method B pays attention to such a point, and defines identifiers having different bit lengths for RAW data and CODE data. For example, no identifier is provided for CODE data, and an identifier is provided only for RAW data. The identifier of RAW data must be a bit pattern that can never occur as a code code of entropy encoding of lossless encoding. For example, when Huffman coding is used for entropy coding, the maximum length of the Huffman code is 16 bits, so the bit length of the identifier needs to be 16 bits or more. As a specific example, when the data length of the identifier is 16 bits, the bit pattern is 0xFFNN (N: arbitrary value). The upper 8 bits of the above bit pattern are ITU-T recommendation T.264. The same 0xFF as the marker defined in 81. However, the lower 8 bits of the bit pattern are arbitrary values. The marker value reserved in 81 cannot be used as the identifier value. Therefore, an example of a 16-bit identifier bit pattern is 0xFFB0. Furthermore, the identifier must be stored in the encoded stream on a byte boundary. Therefore, it is necessary to add bits (padding bits) to the encoded stream so that the encoded stream before storing the identifier ends with a byte boundary.

  However, in general, an image transmitted from an image sensor hardly has a compression rate exceeding 100% even when lossless encoding is performed. Therefore, while ensuring the maximum compression rate, the data length of the encoded stream is reduced. Can be suppressed. However, when the compression rate of all the divided images exceeds 100% and RAW data is selected by the selection unit 102, the encoded stream includes {(number of divided images × 16 bits) + bit length of padding bits. } Is stored. It is obvious that the encoding efficiency is lower than that of the scheme A.

  In the present embodiment, in view of the above-described properties of the method A and the method B, the compression efficiency of the image is calculated before the image is encoded, and either one of the methods is adopted according to the calculated result. And the identifier prescribed | regulated by the employ | adopted system is stored in an encoding stream with an encoding result.

  As described above, in the present embodiment, two methods of method A and method B are used as methods for defining the data structure of the identifier. However, only one method may be used, or 3 Of course, more than one method may be used.

  Next, the operation of the encoding unit 1611 when the imaging mode is set will be described using the flowchart shown in FIG. FIG. 10 is a flowchart of processing performed by the encoding unit 1611 in the imaging mode. Here, as an example, a case where an image of a frame of interest is encoded when an image of each frame is encoded will be described. As described above, when the image of the frame of interest is encoded, the image is divided into a plurality of divided images by the image dividing unit 1610, and each divided image is sequentially input to the encoding unit 1611.

  Further, as described above, any lossless encoding may be used in the present embodiment, but here, for the sake of explanation, the DPCM method is used as an example.

  In step S1001, when the method A is used, the determination unit 104 determines the maximum capacity TA that can be allocated to one image in a memory (not shown) included in the encoding unit 1611 (hereinafter referred to as an encoding memory). Ask. Further, when the method B is used, the determination unit 104 obtains the maximum capacity TB that can be assigned to one image in the encoding memory.

  There are various methods for obtaining the maximum capacity TA. As an example, there is a method in which a quotient obtained by dividing the total capacity of the encoding memory by a preset number of images to be stored in the encoding memory is the maximum capacity TA. In this embodiment, to simplify the description, the data amount of the encoded stream generated according to the scheme A is assumed to be the maximum capacity TA. That is, the maximum capacity TA is the sum of the total value of the data amounts of all the divided images constituting one image and the total bit length (1 bit) of the identifier for each divided image.

  On the other hand, for the maximum capacity TB, the data volume of the encoded stream generated according to the scheme B is set as the maximum capacity TB. That is, the sum of the total data amount of all the divided images constituting one image and the total bit length (16 bits) of the identifier for each divided image is defined as the maximum capacity TB.

  As a supplement, as long as the maximum capacity TA is larger than the maximum capacity TB, either system A or system B may be used. On the other hand, if the maximum capacity TA is smaller than the maximum capacity TB, only the method A is used. However, when the capacity of the encoding memory is smaller than the maximum capacity TA, it is not considered in normal implementation. This is because the number of images that are continuously captured is large, and it is not guaranteed that all the captured images are stored in the memory. Therefore, in such a case, it is necessary to change the number of images that are continuously shot to an appropriate value.

  Next, in step S1002, the determination unit 104 obtains a data amount TP (N) that is actually allocated to one image in the encoding memory. Here, since N indicates the frame number of the image, TP (N) indicates the data amount (capacity) that is actually allocated to the Nth image. That is, an image (image 1 which is the first image) first input to the encoding unit 1611 as an encoding target is N = 1, and then an image (second image) input to the encoding unit 1611 as an encoding target. In the case of image 2), N = 2. In this embodiment, TP (1) = TA. TP (m) (m> 1) will be described later. Hereinafter, a case where the image x (x = 1, 2, 3,...) Is encoded will be described.

  In step S1003, the determination unit 104 determines whether the identifier used by the identifier insertion unit 105 is defined according to the method A or the method A or the method B (method selection). In this determination process, first, the size of the capacity TP (x) obtained in step S1002 for the image x is compared with the maximum capacity TB.

  In the present embodiment, as an example, it is assumed that DPCM conversion is used as a lossless encoding method and Huffman encoding is applied to entropy encoding. Therefore, as shown in FIG. 3, when the method B is used, a 16-bit identifier is added to each divided image. Accordingly, the data length of the identifier is 16 bits multiplied by the number of divided images. Note that the identifier length in the scheme B depends on the lossless encoding scheme, but this embodiment does not limit the identifier length of the scheme B.

  Then, the determination unit 104 determines that the method A is used when TP (x) <TB as a result of the size comparison between the capacity TP (x) and the maximum capacity TB. On the other hand, when TP (x) ≧ TB, the method A may be used or the method B may be used.

  Therefore, if it is determined that the method A is to be used, the process proceeds to step S1008 via step S1005. On the other hand, if it is determined that method A or method B may be used, the process proceeds to step S1007 via step S1005.

  In step S <b> 1007, the determination unit 104 selects one of method A and method B. This selection may be made by selecting a predetermined one or by the user. When the user makes a selection, a screen for the selection may be presented and a selection instruction from the user may be accepted.

  Next, in step S <b> 1008, the lossless encoding unit 107 acquires from the image dividing unit 1610 one divided image that has not yet been encoded for the image x.

  In step S1009, the lossless encoding unit 107 performs lossless encoding on the divided image (RAW data) acquired in step S1008.

  Next, in step S1010, the comparison unit 101 compares the data amount of the CODE data obtained by the lossless encoding by the lossless encoding unit 107 in step S1009 and the data amount of the RAW data acquired in step S1008. Process. As a result of the size comparison processing, if the data amount of the CODE data is smaller than the data amount of the RAW data (compression rate is less than 100%), the process proceeds to step S1011. On the other hand, if the amount of CODE data is greater than or equal to the amount of RAW data (compression rate of 100% or more), the process proceeds to step S1012.

  In step S1011, the selection unit 102 selects the CODE data generated in step S1009 as the encoded result of the divided image acquired in step S1008, and sends the selected CODE data to the identifier insertion unit 105 in the subsequent stage. The identifier insertion unit 105 attaches an identifier having a data structure defined by the determined method to the CODE data. For example, when the determined method is method A, a 1-bit identifier (for example, the value is 1) is attached to the CODE data from FIG. The identifier insertion unit 105 stores the CODE data with the identifier added to the encoded stream for the image x.

  On the other hand, in step S1012, the selection unit 102 selects the RAW data of the divided image as the encoding result of the divided image acquired in step S1008, and sends the selected RAW data to the identifier insertion unit 105 in the subsequent stage. The identifier insertion unit 105 attaches an identifier having a data structure defined by the determined method to the RAW data. For example, when the determined method is method A, a 1-bit identifier (for example, the value is 0) is attached to the RAW data from FIG. The identifier insertion unit 105 stores the RAW data with the identifier added to the encoded stream for the image x.

  When it is determined that the method B is used, the 16-bit identifier is stored on the byte boundary. Therefore, padding bits are inserted so that the length of the encoded stream already generated is a byte boundary. After inserting the padding bits, this 16-bit identifier is stored. Naturally, when the method B is selected and the data format is CODE data, no padding bit insertion is required because there is no identifier. When method A is selected, padding bits are not inserted regardless of the data format, and only an identifier is added.

In step S1013, the counting unit 103 counts the data amount L _ACT of the encoded stream when the CODE data and the identifier are stored in step S1011. Note that the initial value of _LACT is zero.

In step S1014, the counting unit 103 counts the data amount L _ACT of the encoded stream at the time when the RAW data and the identifier are stored in step S1012. Note that the initial value of _LACT is zero.

  Next, if there is a divided image that has not been encoded in the image x, the process returns to step S1008 via step S1015, and the subsequent processes are repeated. On the other hand, if all the divided images have been encoded for the image x, the process proceeds to step S1016 via step S1015.

  Next, when the encoding process has been completed for all the images, the process ends via step S1016. On the other hand, if there is an image that has not been subjected to the encoding process, the process returns to step S1002 via step S1016. Here, the description is continued on the assumption that there is an image (x + 1) that has not been encoded yet after the image x.

  A case will be described in which the processing returns from step S1016 to step S1002 to perform processing on the image (x + 1).

  In step S1002, the determination unit 104 obtains the capacity TP (x + 1) actually allocated to the image (x + 1) in the encoding memory according to the following equation.

TP (x + 1) = (TP (1) + TP (2) +... + TP (x)) − L _ACT + TA
Here, L _ACT is the data amount of the encoded stream completed for the image x. Figure 4 is a diagram for explaining TA in the memory for coding, _TP, the _{L ACT.}

  Next, in step S1003, the determination unit 104 determines whether the identifier used by the identifier insertion unit 105 is defined according to the method A or according to either the method A or the method B. In the determination process, first, the size of the capacity TP (x + 1) obtained in step S1002 for the image (x + 1) is compared with the maximum capacity TB.

  The subsequent processing is as described above.

  In order to simplify the above description, the description has been made assuming that continuously shot images in a digital camera are recorded in a memory. However, the above processing is not applied only to such a case. For example, the number of encoding target images may be one. The image may be received via a communication path such as a network. Furthermore, the encoded stream may be transmitted to the network.

  Further, the above processing does not limit the order of the processing procedures described above, and the order may be appropriately changed as long as similar results can be obtained.

  Further, in the present embodiment, for the sake of simplicity of explanation, the method of selecting either method A or method B according to the number of image frames during continuous shooting has been described. However, it is obvious that the method may be divided into image slice units and the method may be selected in slice units. Furthermore, the measurement target is not limited to the remaining capacity of the memory, and may be anything that measures the total output remaining amount of data that can be output by the apparatus, such as a network bandwidth.

<Decoding unit 1613>
Next, the decoding unit 1613 will be described. FIG. 2 is a block diagram illustrating a detailed functional configuration example of the decoding unit 1613. Although each part shown in FIG. 2 shall be comprised with hardware, you may implement | achieve part or all by a computer program.

  An encoded stream for each image is input to the decoding unit 1613. The identifier decoding unit 205 acquires the encoded stream read from the memory 1605 by the memory control circuit 1607 from the memory control circuit 1607. Then, all identifiers (identifier groups) are extracted from the acquired encoded stream.

  The data reading unit 201 acquires the encoded stream read from the memory 1605 by the memory control circuit 1607 from the memory control circuit 1607. Then, the remaining (encoded image data) obtained by removing the identifier data and the header portion is extracted from the acquired encoded stream. This encoded image data is an encoding result (RAW data or CODE data) for each divided image. The data reading unit 201 outputs the extracted encoded image data in units of divided images, and refers to the identifier group extracted by the identifier decoding unit 205 at the time of output.

  That is, the encoded image data and the identifier group are referred to in order from the beginning, and if the referenced identifier indicates RAW data, the data corresponding to the size of the divided image is read from the encoded image data, and this is used as the RAW data. The data is sent to the selection unit 202. On the other hand, if the referenced identifier indicates CODE data, the data up to the end of the CODE data is read from the encoded image data and sent to the lossless decoding unit 206. In addition, since the process for identifying each “encoding result for each divided image” constituting the encoded image data using a corresponding identifier is well known, further description is omitted.

  The lossless decoding unit 206 decodes the CODE data sent from the data reading unit 201 using a lossless decoding method corresponding to the lossless encoding method used by the lossless encoding unit 107 included in the encoding unit 1611. Then, the RAW data as the decoding result is sent to the selection unit 202.

  Since either the RAW data from the data reading unit 201 or the RAW data from the lossless decoding unit 206 is input to the selection unit 202, it is transmitted to the outside.

  The counting unit 203 counts the data amount of data transmitted from the data reading unit 201 and the data length of the identifier extracted by the identifier decoding unit 205.

  The determination unit 204 determines the identifier method based on the counting result of the counting unit 203, that is, the accumulated data amount counted by the counting unit 203 for one image. The operations of the counting unit 203 and the determining unit 204 are the same as those of the counting unit 103 and the determining unit 104 shown in FIG.

  Next, the operation of the decoding unit 1613 will be described with reference to the flowchart of FIG. FIG. 13 is a flowchart of processing performed by the decoding unit 1613.

  Here, in order to simplify the description, a method for restoring data in order to browse data continuously taken by a digital camera and stored in a storage medium such as a memory or a buffer will be described.

  Steps S1001 to S1003, S1005, and S1007 are the same as described in the first embodiment, and only the main body that performs the process is different. Therefore, the description thereof is omitted. That is, what the counting unit 103 performs in the first embodiment is performed by the counting unit 203 in the present embodiment, and what the determination unit 104 performs in the first embodiment is performed by the determination unit in the present embodiment. 204 performs.

  If it is determined that the method A is used, the process proceeds to step S1309 via step S1308. If it is determined that the method B is used, the process proceeds to step S1310 via step S1308.

  In step S1309, the identifier decoding unit 205 pre-reads the input encoded stream and interprets each identifier stored in the encoded stream according to the scheme A. Then, it is determined whether the read identifier indicates RAW data or CODE data by interpretation according to method A. For example, one bit of the encoded stream is read, and whether the data to which the identifier is attached is RAW data or CODE data is specified according to the read bit value.

  On the other hand, in step S1310, the identifier decoding unit 205 pre-reads the input encoded stream and interprets each identifier stored in the encoded stream according to the scheme B. For example, padding bits inserted up to the byte boundary in the encoded stream are discarded. Thereafter, a 16-bit bit string is read again from the encoded stream as an identifier, and the bit pattern matches the identifier indicating the RAW data. If the bit pattern of the identifier matches, the data format in the encoded stream is RAW data. If there is no match, the data format in the encoded stream is CODE data.

  If the determination result in step S1309 / step S1310 is RAW data, the process proceeds to step S1312 via step S1311. If the determination result is CODE data, the process proceeds to step S1313 via step S1311.

  In step S1312, the data reading unit 201 reads data for the size of the divided image from the encoded image data, and sends this to the selection unit 202 as RAW data. On the other hand, in step S <b> 1313, the data reading unit 201 reads up to the end of the CODE data from the encoded image data and sends it to the lossless decoding unit 206. In the case of method B, the padding bits inserted before the identifier are also discarded.

In step S1314, the counting unit 203 accumulates the total value of the data amount of the RAW data read by the data reading unit 201 and the data length of the identifier. Let this accumulated value be _LACT . In the case of method B, the data length of the padding bits added to the _{L ACT.}

On the other hand, in step S1315, the counting unit 203 accumulates the total value of the data amount of the CODE data read by the data reading unit 201 and the data length of the identifier. Let this accumulated value be _LACT . In the case of method B, the data length of the padding bits added to the _{L ACT.} Note that the initial value of _LACT is zero.

  Next, in step S1316, the lossless decoding unit 206 restores RAW data by performing lossless decoding on the CODE data read by the data reading unit 201 in step S1313. Then, the restored RAW data is sent to the selection unit 102.

  Next, if there is a divided image that has not been decoded yet, the process returns to step S1308 via step S1317, and the subsequent processes are repeated. On the other hand, if all the divided images have been decoded, the process proceeds to step S1318 via step S1317.

  Next, when the decoding process has been completed for all the images, the process ends via step S1318. On the other hand, if there is an encoded stream that has not been subjected to decoding processing, the processing returns to step S1002 via step S1318.

  In the above description of the decoding process, in order to simplify the description, it has been described assuming that a continuously shot image recorded on a storage medium is restored. It is not limited. For example, the number of decoding target images may be one. The encoded stream may be received via a communication path such as a network. Further, the restored image data may be transmitted to the network.

  Further, the above processing does not limit the order of the processing procedures described above, and the order may be appropriately changed as long as similar results can be obtained.

  As described above, according to the present embodiment, the load of the encoding process is small, but the information stored in the storage medium is guaranteed without being lost in the information of continuously shot images, and is used in the storage medium. Data capacity to be reduced.

[Second Embodiment]
Below, only the part from which this embodiment differs from 1st Embodiment is demonstrated.

<Encoding unit 1611>
FIG. 6 is a block diagram illustrating a detailed functional configuration example of the encoding unit 1611 according to the present embodiment. The configuration shown in FIG. 6 is obtained by adding a prediction unit 606 to the configuration shown in FIG. Accordingly, since the configuration other than the prediction unit 606 is the same as that of the first embodiment, only the operation of the prediction unit 606 and the operation of each unit related thereto will be described below.

  In the first embodiment, as a result of comparing the size of the capacity TP (x) and the maximum capacity TB, when TP (x) ≧ TB, the method A or the method B may be used. Either one of them was manually selected or a predetermined one was selected. In this embodiment, the prediction unit 606 determines which method to select.

  FIG. 11 is a flowchart of processing performed by the encoding unit 1611 according to this embodiment. In the flowchart shown in FIG. 11, only step S1106 is added to the flowchart shown in FIG. 10, and therefore only step S1106 and steps related thereto will be described below.

  FIG. 5 shows an example of the data format of the encoded divided image adjacent to the encoding target divided image X. In FIG. 5, a divided image whose data format is RAW data is described as “RAW”, and a divided image whose data format is CODE data is described as “CODE”.

  For example, in the case of the data format of the adjacent encoded divided image and there is even one RAW data, the compression rate may exceed 100% again. In this case, the prediction unit 606 uses the data format. As RAW data is selected.

  In addition, when the data formats of the adjacent encoded divided images are all CODE data, it is highly possible that the compression rate is continuously cut below 100%. In this case, the prediction unit 606 uses the CODE data as the data format. Is predicted to be selected.

  FIG. 5 is an example of a method for predicting a divided image, and the method for predicting the data format of the divided image is not limited.

  Accordingly, in step S1106, the prediction unit 606 predicts that RAW data is selected as the data format when at least one RAW data is included in the encoding result of the divided image adjacent to the encoding target divided image. To do. Also, the prediction unit 606 predicts that CODE data is selected as the data format when the encoding results of all the divided images adjacent to the encoding target divided image are CODE data.

  In step S1007, the determination unit 604 selects either method A or method B based on the prediction in step S1106.

  For example, if it is predicted in step S1106 that RAW data is selected as the data format, method A is selected. On the other hand, when it is predicted that CODE data is selected as the data format, the method B is selected.

  Note that other prediction methods by the prediction unit 606 are conceivable and are not limited to the processing described.

<Decoding unit 1613>
FIG. 7 is a block diagram illustrating a functional configuration example of the decoding unit 1613 according to the present embodiment. The configuration shown in FIG. 7 is obtained by adding a prediction unit 706 to the configuration shown in FIG. Therefore, since the configuration other than the prediction unit 706 is the same as that of the first embodiment, only the operation of the prediction unit 706 and the operation of each unit related thereto will be described below.

  The operation of the prediction unit 706 is basically the same as that of the prediction unit 606. A more detailed operation will be described with reference to the flowchart of FIG.

  FIG. 14 is a flowchart of processing performed by the decoding unit 1613. In the flowchart shown in FIG. 14, only step S1406 is added to the flowchart shown in FIG. 13, so only step S1406 will be described below.

  In step S1406, the prediction unit 706 performs the same process as in step S1106 using the encoded divided image identifier.

  As described above, according to the present embodiment, it is possible to select a method for adaptively specifying an identifier by predicting a data format for each divided image. Therefore, the amount of data required for the identifier can be suppressed.

[Third Embodiment]
Below, only the part from which this embodiment differs from 1st Embodiment is demonstrated.

<Encoding unit 1611>
FIG. 8 is a block diagram illustrating a detailed functional configuration example of the encoding unit 1611 according to the present embodiment. The configuration shown in FIG. 8 is obtained by adding an identifier insertion unit 808 to the configuration shown in FIG. Accordingly, since the configuration other than the identifier insertion unit 808 is the same as that of the second embodiment, only the operation of the identifier insertion unit 808 and the operation of each unit related thereto will be described below.

  The identifier inserting unit 808 uses an identifier (first number) indicating whether the identifier inserting unit 105 uses the identifier based on the method A for all the divided images constituting one image or both the method A and the method B. 2 identifier) is stored in the encoded stream.

  FIG. 12 is a flowchart of processing performed by the encoding unit 1611 according to this embodiment. In the flowchart shown in FIG. 12, step S1204 is added to the flowchart shown in FIG. Therefore, only these steps will be described below.

  In step S1204, the identifier insertion unit 808 uses an identifier (first identifier) indicating whether an identifier based on the scheme A or an identifier based on the scheme B is used for all the divided images constituting one image. And stored in the encoded stream for this image.

  The second identifier used in steps S1211 and S1212 is the same as the identifier used in the second embodiment, and the process itself is the same as steps S1011 and S1012.

<Decoding unit 1613>
FIG. 9 is a block diagram illustrating a detailed functional configuration example of the decoding unit 1613 according to the present embodiment. The configuration shown in FIG. 9 is obtained by adding an identifier decoding unit 907 to the configuration shown in FIG. Accordingly, since the configuration other than the identifier decoding unit 907 is the same as that of the second embodiment, the operation of the identifier decoding unit 907 and the operation of each unit related thereto will be described below.

  FIG. 15 is a flowchart of processing performed by the decoding unit 1613 according to this embodiment.

  In step S1501, the identifier decoding unit 907 extracts the first identifier stored in the encoded stream. Then, whether the extracted first identifier indicates that only the identifier according to the scheme A is stored in the encoded stream, or the identifier according to the scheme A and the identifier according to the scheme B are mixed. It is determined whether it is stored. As a result of such determination, when the first identifier indicates that only an identifier according to the scheme A is stored in the encoded stream, the process proceeds to step S1308 via step S1005. On the other hand, if the first identifier indicates that an identifier according to method A and an identifier according to method B are stored together, the process proceeds to step S1106 via step S1005. .

  The second identifier used in steps S1509 and S1510 is the same as the identifier used in the second embodiment, and the processing itself is the same as steps S1309 and S1310, respectively.

  As described above, according to the present embodiment, the counting process of the encoded stream on the decoding side can be omitted, so that the decoding process can be simplified. Furthermore, the mounting cost can be suppressed.

[Fourth Embodiment]
In the first to third embodiments, the configuration for encoding an image and the configuration for decoding the encoded image are provided in one image processing device, but each configuration is included in a separate device. You may make it provide. That is, the encoded stream generated by one device may be decoded by the other device. In each case, for example, a general PC (personal computer) can be applied. Moreover, various things can be considered about the structure of the system comprised by each apparatus.

  The effects of the above embodiments can be summarized as follows. That is, it is possible to realize an encoding function that guarantees the data amount of the encoded stream even in lossless encoding, and further can suppress the data amount required for the identifier indicating the data format in the encoded stream.

  Furthermore, since this encoding function can sequentially estimate a plurality of identifier schemes when decoding an encoded stream, additional information for determining the plurality of identifier schemes is not necessary. This encoding function can also suppress the amount of data required for the identifier indicating the data format in the encoded stream.

  Thereby, it is possible to realize an encoding function that guarantees the maximum amount of data without damaging the information of the original image data. In addition, this encoding function or a decoding function for decoding an encoded stream generated by a method similar to this encoding function can be realized.

[Other Embodiments]
Needless to say, the object of the present invention can be achieved as follows. That is, a recording medium (or storage medium) that records a program code (computer program) of software that implements the functions of the above-described embodiments is supplied to the system or apparatus. Needless to say, such a storage medium is a computer-readable storage medium. Then, the computer (or CPU or MPU) of the system or apparatus reads and executes the program code stored in the recording medium. In this case, the program code itself read from the recording medium realizes the functions of the above-described embodiment, and the recording medium on which the program code is recorded constitutes the present invention.

  Further, by executing the program code read by the computer, an operating system (OS) or the like running on the computer performs part or all of the actual processing based on the instruction of the program code. Needless to say, the process includes the case where the functions of the above-described embodiments are realized.

  Furthermore, it is assumed that the program code read from the recording medium is written in a memory provided in a function expansion card inserted into the computer or a function expansion unit connected to the computer. After that, based on the instruction of the program code, the CPU included in the function expansion card or function expansion unit performs part or all of the actual processing, and the function of the above-described embodiment is realized by the processing. Needless to say.

  When the present invention is applied to the recording medium, program code corresponding to the flowchart described above is stored in the recording medium.

12 is a block diagram illustrating a detailed functional configuration example of an encoding unit 1611. FIG. 24 is a block diagram illustrating a detailed functional configuration example of a decoding unit 1613. FIG. It is a figure which shows each of the system A and the system B in a table format. TA in the memory for encoding is a diagram for explaining _TP, the _{L ACT.} The example of the data format of the encoded divided image adjacent to the encoding object divided image X is shown. It is a block diagram which shows the detailed functional structural example of the encoding part 1611 which concerns on the 2nd Embodiment of this invention. It is a block diagram which shows the function structural example of the decoding part 1613 which concerns on the 2nd Embodiment of this invention. It is a block diagram which shows the detailed function structural example of the encoding part 1611 which concerns on the 3rd Embodiment of this invention. It is a block diagram which shows the detailed function structural example of the decoding part 1613 which concerns on the 3rd Embodiment of this invention. 25 is a flowchart of processing performed by an encoding unit 1611 in the imaging mode. It is a flowchart of the process which the encoding part 1611 which concerns on the 2nd Embodiment of this invention performs. It is a flowchart of the process which the encoding part 1611 which concerns on the 3rd Embodiment of this invention performs. 21 is a flowchart of processing performed by a decoding unit 1613. 21 is a flowchart of processing performed by a decoding unit 1613. It is a flowchart of the process which the decoding part 1613 which concerns on the 3rd Embodiment of this invention performs. It is a figure which shows the hardware structural example of the digital camera as an image processing apparatus which concerns on the 1st Embodiment of this invention.

Claims (18)

Dividing the input image into a plurality of divided images, by performing lossless encoding on the divided image, an image processing apparatus for generating an encoded divided picture of the divided images,
The data amount of the encoded divided image and the data amount of the divided image are compared, and based on the result of the size comparison, either the divided image or the encoded divided image is the divided image. Selecting means for selecting as an encoding result of
First storage control means for storing an encoding result selected by the selection means in an encoded stream;
An identifier for specifying whether the encoding result selected by the selection unit for each divided image constituting the input image is the divided image or the encoded divided image. Second storage control means for storing, in the encoded stream, an identifier having a data structure determined in advance among a plurality of types of data structures required for the identifier ;
Based on the data amount of the encoded stream completed by the first storage control means and the second storage control means, the second storage control means uses the image input next to the input image. An image processing apparatus comprising: a determination unit that determines a data structure of the identifier.   The selection means determines whether or not the compression rate of the encoded divided image is less than 100%. If the compression ratio is less than 100%, the encoded divided image, and if it is 100% or more, the divided image, The image processing apparatus according to claim 1, wherein the image processing apparatus is selected as an encoding result of the divided image.   The determination means is a first method for defining a data structure of an identifier indicating whether the result selected by the selection means is the divided image or the encoded divided image, and the result selected by the selection means is the The second storage control is performed on an image input next to the input image in accordance with the selected method by selecting any one of the second methods for defining the data structure of the identifier indicating whether the image is a divided image. The image processing apparatus according to claim 1, wherein the data structure of an identifier used in the means is determined. The determining means includes
The first storage control means, said second storage control unit, L the data amount of the coded stream for the finished N-th input image by _ACT, the N-th calculated data as data amount to be allocated to the input image Code when the amount is TP (N), the selection means selects a divided image as an encoding result of all the divided images constituting the Nth input image, and an identifier defined by the first method is used. the data amount of stream TA, when the N-th of the selection means as the encoding result of all the divided images to select the divided image constituting the input image and the second method is used an identifier defining When the data amount of the coded stream TB, and the (N + 1) -th assigned to the input image data amount TP (N + 1), TP (N + 1) = (TP (1) + TP (2) + + TP _{(N)) -} means for determining by calculating the _L ACT + TA,
When TP (N + 1) <TB, the first method is selected, and when TP (N + 1) ≧ TB, method selecting means for selecting the first method or the second method is provided. Prepared,
The second storage control means stores, for the (N + 1) th input image, an identifier having a data structure defined by the method selected by the method selection means in the encoded stream. Item 4. The image processing apparatus according to Item 3.   The second storage control means, when the first method is selected, is an identifier of a data structure defined by the first method for the divided image, and the first method is the encoding. The image processing apparatus according to claim 3, wherein an identifier having a data structure defining a divided image is used.   The second storage control unit uses an identifier of a data structure defined by the second method for the divided image when the second method is selected. The image processing apparatus according to 3. The TA is a total value of data amounts of all divided images constituting the Nth input image, and a total value of bit lengths of identifiers defined by the first method for the respective divided images. Is sum,
The TB is a total value of the data amounts of all the divided images constituting the Nth input image, and a total value of the bit lengths of identifiers defined by the second method for the respective divided images. the image processing apparatus according to any one of claims 4 to 6, characterized in that the sum. Dividing the input image into a plurality of divided images, by performing lossless encoding on the divided image, an image processing apparatus for generating an encoded divided picture of the divided images,
The data amount of the encoded divided image and the data amount of the divided image are compared, and based on the result of the size comparison, either the divided image or the encoded divided image is the divided image. Selecting means for selecting as an encoding result of
First storage control means for storing an encoding result selected by the selection means in an encoded stream;
An identifier for specifying whether the encoding result selected by the selection unit for each divided image constituting the input image is the divided image or the encoded divided image. Second storage control means for storing an identifier having a predetermined data structure among a plurality of types of data structures required for the identifier in the encoded stream;
The second storage control means includes a determination means for determining a data structure of the identifier for each divided image,
The determination means determines one data structure based on a coding result of a divided image selected by the selection means adjacent to the divided image when a plurality of data structures can be used. An image processing apparatus. The determining means includes
When even one divided image is included in the encoding result of the divided image selected by the selection unit adjacent to the divided image, the result selected by the selection unit is the divided image, the encoded divided image Determining a data structure based on a first method that defines a data structure of an identifier indicating which one of
An identifier indicating whether or not the result selected by the selection unit is the divided image when all of the encoding results of the divided images selected by the selection unit adjacent to the divided image are the encoded divided images. The image processing apparatus according to claim 8 , wherein a data structure based on a second method that defines the data structure of the image data is determined. Furthermore,
It did using an identifier having a data structure based on the same method for all the divided images, information indicating whether the not used according to claim 8 or 9, characterized in that it comprises means for storing the coded stream Image processing device. Dividing the input image into a plurality of divided images, generated by performing lossless encoding on the divided image, an image processing apparatus for decoding an encoded divided picture of the divided images,
For specifying whether a divided image or an encoded divided image as an encoding result of each divided image constituting the image and whether the encoding result for each divided image is a divided image or an encoded divided image Means for obtaining an encoded stream including an identifier;
Determining means for determining a method for defining a data structure of each identifier in the encoded stream;
Each identifier is interpreted based on the method determined by the determination means, and based on the interpreted identifier, it is determined whether the corresponding encoding result is an encoded divided image or a divided image, and the encoded divided image The encoded divided image is decoded and output, and when the specified format is a divided image, the output unit includes an output unit that outputs the divided image.
The image processing apparatus according to claim 1, wherein the determination unit determines a method of each identifier in an encoded stream input next to the encoded stream, based on a data amount of the encoded stream. Dividing the input image into a plurality of divided images, generated by performing lossless encoding on the divided image, an image processing apparatus for decoding an encoded divided picture of the divided images,
For specifying whether a divided image or an encoded divided image as an encoding result of each divided image constituting the image and whether the encoding result for each divided image is a divided image or an encoded divided image Means for obtaining an encoded stream including an identifier;
Determining means for determining a method for defining a data structure of each identifier in the encoded stream;
Each identifier is interpreted based on the method determined by the determination means, and based on the interpreted identifier, it is determined whether the corresponding encoding result is an encoded divided image or a divided image, and the encoded divided image The encoded divided image is decoded and output, and when the specified format is a divided image, the output unit includes an output unit that outputs the divided image.
The determining means determines the method for each divided image, and when a plurality of methods can be used, determines one method based on a coding result of a divided image adjacent to the divided image. A featured image processing apparatus. Dividing the input image into a plurality of divided images, by performing lossless encoding on the divided image, an image processing method for an image processing apparatus for generating an encoded divided picture of the divided picture is performed,
The selection unit of the image processing device performs a size comparison between the data amount of the encoded divided image and the data amount of the divided image, and based on the result of the size comparison, the divided image and the encoded division A selection step of selecting any one of the images as an encoding result of the divided image;
A first storage control step in which the first storage control means of the image processing apparatus stores the encoding result selected in the selection step in an encoded stream;
The second storage control means of the image processing device specifies whether the encoding result selected in the selection step for each divided image constituting the input image is the divided image or the encoded divided image A second storage control for storing an identifier having a predetermined data structure among a plurality of types of data structures required for the identifier for identification in the encoded stream Process,
An image to be input next to the input image based on a data amount of the encoded stream completed by the determination unit of the image processing device in the first storage control step and the second storage control step. A determination step of determining a data structure of an identifier used in the second storage control step. Dividing the input image into a plurality of divided images, by performing lossless encoding on the divided image, an image processing method for an image processing apparatus for generating an encoded divided picture of the divided picture is performed,
The selection unit of the image processing device performs a size comparison between the data amount of the encoded divided image and the data amount of the divided image, and based on the result of the size comparison, the divided image and the encoded division A selection step of selecting any one of the images as an encoding result of the divided image;
A first storage control step in which the first storage control means of the image processing apparatus stores the encoding result selected in the selection step in an encoded stream;
The second storage control means of the image processing device specifies whether the encoding result selected in the selection step for each divided image constituting the input image is the divided image or the encoded divided image A second storage control for storing an identifier having a predetermined data structure among a plurality of types of data structures required for the identifier for identification in the encoded stream A process,
The second storage control step includes a determination step of determining a data structure of the identifier for each divided image,
In the determining step, when a plurality of data structures can be used, one data structure is determined based on the encoding result of the divided image selected in the selection step adjacent to the divided image. Image processing method. Dividing the input image into a plurality of divided images, generated by performing lossless encoding on the divided image, an image processing method to be performed by the image processing apparatus for decoding an encoded divided picture of the divided images,
The acquisition unit of the image processing device is a divided image or an encoded divided image as an encoding result of each divided image constituting the image, and an encoding result for each divided image is a divided image or an encoded divided image. Obtaining an encoded stream including an identifier for specifying which is, and
A determining step in which the determining means of the image processing apparatus determines a method for defining a data structure of each identifier in the encoded stream;
The output means of the image processing apparatus interprets each identifier based on the method determined in the determination step, and based on the interpreted identifier, whether the corresponding encoding result is an encoded divided image or a divided image When the encoded divided image is a decoded image, the encoded divided image is output after being decoded. When the specified format is a divided image, an output step of outputting the divided image is performed. Prepared,
The image processing method according to claim 1, wherein, in the determining step, a method of each identifier in an encoded stream input next to the encoded stream is determined based on a data amount of the encoded stream. Dividing the input image into a plurality of divided images, generated by performing lossless encoding on the divided image, an image processing method to be performed by the image processing apparatus for decoding an encoded divided picture of the divided images,
The acquisition unit of the image processing device is a divided image or an encoded divided image as an encoding result of each divided image constituting the image, and an encoding result for each divided image is a divided image or an encoded divided image. Obtaining an encoded stream including an identifier for specifying which is, and
A determining step in which the determining means of the image processing apparatus determines a method for defining a data structure of each identifier in the encoded stream;
The output means of the image processing apparatus interprets each identifier based on the method determined in the determination step, and based on the interpreted identifier, whether the corresponding encoding result is an encoded divided image or a divided image When the encoded divided image is a decoded image, the encoded divided image is output after being decoded. When the specified format is a divided image, an output step of outputting the divided image is performed. Prepared,
In the determining step, the method is determined for each divided image, and when a plurality of methods can be used, one method is determined based on the encoding result of the divided image adjacent to the divided image. A featured image processing method. A computer program for a computer to function as each means of the image processing apparatus according to any one of claims 1 to 12. A computer-readable storage medium storing the computer program according to claim 17 .

Images