TWI663575B - Method and electrical device for image motion detection - Google Patents

Abstract

The invention proposes an image motion detection method, which includes: obtaining a first image and a second image, dividing the first image into a plurality of first blocks, and dividing the second image into a plurality of second blocks; calculating Block difference value between each first block and corresponding second block; calculate the number of difference blocks, and the number of difference blocks indicates how many blocks have a difference value greater than the first threshold value; and if the difference If the number of blocks is greater than the second threshold, it is determined that the action mode is a dynamic mode, otherwise the action mode is a static mode, where the first threshold value in the dynamic mode is smaller than the first threshold value in the static mode or the first threshold value in the dynamic mode. The second critical value is smaller than the second critical value in the static mode.

Description

Image motion detection method and electronic device

The invention relates to an image motion detection method, and in particular to an image motion detection method using different threshold values in different modes.

Video motion detection has a wide range of applications. For example, a monitor can start recording when motion is detected, or trigger an alarm or send some specific messages when motion is detected. In some applications, an always-on sensor and an arithmetic circuit are used to execute a motion detection algorithm. In such applications, very low power consumption is required, so how to propose a low power consumption and effective Motion detection algorithms are a topic of concern to those skilled in the art.

An embodiment of the present invention provides an image motion detection method, which is suitable for an electronic device. The motion detection method includes: obtaining a first image and a second image, dividing the first image into a plurality of first blocks, and dividing the second image into a plurality of second blocks, where the first block is respectively Corresponds to the second block; Calculate a block difference value between each first block and the corresponding second block, and determine whether the block difference value is greater than the first critical value; calculate the number of difference blocks, and the number of difference blocks represents the first There are several blocks in the image or the second image whose difference value is greater than the first threshold value; and if the number of difference blocks is greater than the second threshold value, it is judged that the motion mode is a dynamic mode, otherwise the motion mode is a static mode, in which the The first critical value in the lower mode is smaller than the first critical value in the static mode, or the second critical value in the dynamic mode is smaller than the second critical value in the static mode.

In some embodiments, the step of calculating a block difference value between the first block and the corresponding second block includes: calculating a first histogram for pixels in the first block; and for each second region The pixels in the block calculate a second histogram; and calculate the difference between the first histogram and the corresponding second histogram as a block difference value.

In some embodiments, the above block difference value is an absolute difference sum.

In some embodiments, the motion detection method further includes: setting a first threshold value or a second threshold value according to a time distance between the first image and the second image, wherein the time distance is positively correlated with the first threshold value, or The time distance is positively correlated with the second critical value.

In some embodiments, the motion detection method further includes: determining a motion mode between the second image and the third image, wherein a time distance between the first image and the second image is smaller than a distance between the second image and the third image Time distance.

From another perspective, the embodiment of the present invention proposes a An electronic device includes an always-on sensor and a computing circuit. The sensor is always activated to obtain the first image and the second image. The arithmetic circuit is used to divide the first image into a plurality of first blocks and the second image into a plurality of second blocks, wherein the first block is respectively corresponding to the second block. The arithmetic circuit is configured to calculate a block difference value between each first block and a corresponding second block, and determine whether the block difference value is greater than a first critical value. The arithmetic circuit is used to calculate a number of difference blocks. The number of difference blocks indicates that the difference value of several blocks in the first image or the second image is greater than the first critical value. If it is determined that the number of difference blocks is greater than the second critical value, the arithmetic circuit determines that the operation mode is a dynamic mode, otherwise the operation mode is a static mode. The first critical value in the dynamic mode is smaller than the first critical value in the static mode, or the second critical value in the dynamic mode is smaller than the second critical value in the static mode.

In some embodiments, the arithmetic circuit is further configured to calculate a first histogram for the pixels in the first block, calculate a second histogram for the pixels in the second block, and calculate the first histogram and the corresponding The difference between the second histograms is used as the block difference value.

In some embodiments, the arithmetic circuit is further configured to set the first threshold value or the second threshold value according to a time distance between the first image and the second image, where the time distance is positively correlated with the first threshold value, or time The distance is positively related to the second critical value.

In some embodiments, the arithmetic circuit is further configured to determine an action mode between the second image and the third image. The time distance between the first image and the second image is smaller than the time distance between the second image and the third image.

In order to make the above features and advantages of the present invention more comprehensible, The embodiments are exemplified below and described in detail with the accompanying drawings.

100‧‧‧ electronic device

110‧‧‧Always activate the sensor

120‧‧‧ Operation Circuit

130‧‧‧Logic Circuit

210‧‧‧ first image

220‧‧‧Second image

230‧‧‧ third image

Blocks 211 ~ 213, 221 ~ 223‧‧‧

d ₁ , d ₂ ‧‧‧ time distance

T _{1, s} , T _{1, d} ‧‧‧ first critical value

T _{2, s} , T _{2, d} ‧‧‧ second critical value

401 ~ 406‧‧‧step

1 is a schematic diagram illustrating an electronic device according to an embodiment.

FIG. 2 is a schematic diagram illustrating multiple images on a time axis according to an embodiment.

[FIG. 3A] and [FIG. 3B] are schematic diagrams illustrating setting different thresholds in different modes according to an embodiment.

4 is a flowchart illustrating a motion detection method according to an embodiment.

Regarding the "first", "second", ... and the like used herein, they do not specifically mean the order or the order, but merely the difference between the elements or operations described in the same technical terms.

FIG. 1 is a schematic diagram illustrating an electronic device according to an embodiment. Referring to FIG. 1, the electronic device 100 is, for example, a monitor, a camera, a computer system, or any other suitable electronic device. The electronic device 100 includes an always-on sensor 110, an arithmetic circuit 120 and a logic circuit 130. The sensor 110 is always activated to continuously capture images, and the arithmetic circuit 120 performs motion detection based on these images, and the detection results are transmitted to the logic circuit 130 for subsequent processing. In some embodiments, the arithmetic circuit 120 may be a central processing unit, a microprocessor, a microcontroller, a digital signal processor, an image processing chip, a special application integrated circuit, and the like.

FIG. 2 is a schematic diagram illustrating multiple images on a time axis according to an embodiment. Here, the first image 210, the second image 220, and the third image 230 are used to describe the motion detection method. The following steps are performed by the arithmetic circuit 120 and will not be described again. First, the first image 210 is divided into a plurality of first blocks, and the second image 220 is divided into a plurality of second blocks. These first blocks are respectively corresponding to the second blocks, such as the first block. 211 corresponds to the second block 221, the first block 212 corresponds to the second block 222, and so on. In the embodiment of FIG. 2, the first image 210 and the second image 220 each have 16 blocks, but the number of this block is only for illustration. The present invention does not limit how many blocks the image is divided into. There are no restrictions on the size and shape of these blocks.

Next, calculate a block difference value between each first block and the corresponding second block. In some embodiments, a histogram is calculated first, and then a block difference value is calculated based on the histogram. Specifically, the gray level of the pixel may be first divided into a plurality of bins, for example, a range from 0 to 255 is divided into 8 bins. For each pixel in a certain first block, you can determine which slot the pixel's gray level falls in, so as to count how many pixels each slot has. Those skilled in the art can understand the histogram. The calculation is not detailed here. Therefore, a first histogram can be calculated for the pixels in the first block 211, and a second histogram can be calculated for the pixels in the second block 221. Next, calculate the difference between the first histogram and the second histogram, which can be used as the block difference value. In some embodiments, the sum of absolute difference between two histograms is calculated as the block difference value, which can be expressed as the following equation (1).

SAD _i = Σ _j | bin 1 _{i, j} - bin 2 _{i, j} | ... (1)

SAD _i represents the block difference value corresponding to the i-th block. bin 1 _{i, j} represents the value of the j th slot in the histogram of the i th block in the first image. bin 2 _{i, j} represents the value of the j th slot in the histogram of the i th block in the second image. However, in other embodiments, the sum of square differences between the two histograms or other suitable differences may be calculated as the above-mentioned block difference value. Alternatively, in some embodiments, the absolute difference sum or the square difference sum of the grayscale values between the first block 211 and the second block 221 may be calculated as the block difference value (the histogram is not calculated). In some embodiments, in order to save memory usage or power consumption, pixels in the first block 211 and the second block 221 may also be subjected to down sampling to calculate the block difference value. For example, if 2 times downsampling is performed, only 1/4 of the pixels in the block will be used to calculate the above-mentioned block difference value, but the present invention does not limit the downsampling magnification.

After the block difference value is calculated, it is further judged whether the block difference value is greater than a first threshold value. If it is greater than the first threshold value, it indicates that an action has occurred in this block. After calculating the block difference value for each block in the image and determining whether it is greater than the first critical value, a number of difference blocks can be calculated. The number of difference blocks indicates that the number of block differences is greater than First critical value. For example, in FIG. 2, the block difference values corresponding to blocks 211 and 221 are greater than the first critical value; the block difference values corresponding to blocks 212 and 222 are greater than the first critical value; blocks 213, The block difference value corresponding to 223 is greater than the first critical value. Therefore, the block difference value of a total of 3 blocks is greater than the first critical value, that is, the number of the above-mentioned difference blocks is three.

In this embodiment, it is to determine whether an action occurs, so an action mode can be set, and the action mode can be a dynamic mode or a static mode. This action mode can be represented by a numerical value, such as 0 for static mode and 1 for dynamic mode. However, this action mode can also be represented by characters, strings, or one or more bits, and the present invention is not limited to what data type is used to represent the action mode. The above number of difference blocks can be used to determine the operation mode. Specifically, it can be determined whether the number of difference blocks is greater than a second critical value, and the second critical value can be 4, 8, 12, or any other value. If the number of difference blocks is greater than the second critical value, it means that there are actions in many blocks in the image, so it can be judged that the action mode is a dynamic mode, otherwise it is judged that the action mode is a static mode. After determining the dynamic mode or the static mode, the electronic device 100 can execute any subsequent programs, for example, start recording in the dynamic mode, etc. The present invention does not limit what programs are to be executed in the dynamic mode and the static mode.

In particular, different first and second critical values are set in the dynamic mode and the static mode, respectively, in order to prevent the operation mode from being frequently switched. FIG. 3A is a schematic diagram illustrating setting different first thresholds in different modes according to an embodiment. Referring to FIG. 3A, the first critical value used in the static mode is T _{1, s} , and the first critical value used in the dynamic mode is T _{1, d} . The first critical value T _{1, d} is Less than the first critical value T _{1, s} . When the first critical value is larger (in the static mode), it means that the larger the block difference value is, it will be regarded as an action. Therefore, the number of the difference blocks calculated above will be smaller and it will not be easily judged as dynamic In other words, it is easier to maintain the static mode. Conversely, when the first threshold value is smaller (in the dynamic mode), the block difference value is likely to be greater than the first threshold value and it is considered to be an action. Therefore, the number of the difference blocks calculated above will be larger and less likely. It is easy to be judged as static mode, and it is easy to maintain in dynamic mode. If both the dynamic mode and the static mode use the same first critical value (for example, T _{1, s} ), and the block difference value changes back and forth around the first critical value T _{1, s} , then it is easy to continuously change between the dynamic mode and the static mode. Switch between. The embodiment of FIG. 3A is designed to use different first thresholds in different modes, so that the curve of mode switching in FIG. 3A is similar to the hysteresis curve.

For the same reason, different second thresholds can also be used in different modes, so a hysteresis curve can also be plotted for the second threshold. Please refer to FIG. 3B. FIG. 3B is a schematic diagram illustrating setting different second thresholds in different modes according to an embodiment. When the second threshold value used in static mode T _{2, s,} and when the second threshold value used in the dynamic mode T _{2, d,} wherein the second threshold value in the dynamic mode T _{2, d} is less than The second critical value T _{2, s in the} static mode. When the second critical value is larger (in the static mode), the number of difference blocks must be larger to be judged as the dynamic mode, so it is easy to maintain the static mode. Conversely, when the second critical value is smaller (in the dynamic mode), the number of difference blocks must be smaller to be judged as the static mode, so it is easy to maintain the dynamic mode.

In some embodiments, the above-mentioned first threshold value and the second threshold value have different values in different modes, so there are a total of 4 threshold values T _{1, s} , T _{1, d} , T _{2, s} , T _{2, d} . However, in some embodiments, it is also possible to set only the first threshold (or the second threshold) to have different values in different modes, and the second threshold (or the first threshold) to have different values in different modes. The same value is used, that is, T _{2, s} = T _{2, d} or T _{1, s} = T _{1, d} . This is to facilitate the switching of the control operation mode. In this embodiment, there are 3 critical values.

Please refer to FIG. 2. The above motion detection method between the first image 210 and the second image 220 can also be applied between the second image 220 and the third image 230, but different threshold values can be adopted when the time distance is different. . Specifically, the time distance between the first image 210 and the third image 230 is 1 second, but the time distance d ₁ between the first image 210 and the second image 220 is smaller than that between the second image 220 and the third image 230. The time distance d _{2 is} , for example, the time distance d ₁ is 1/30 seconds, and the time distance d ₂ is 29/30 seconds. Compared to the design in which the time distance d ₁ is equal to the time distance d ₂ , the design in FIG. 2 is to allow the arithmetic circuit 120 to sleep for a longer time after calculating the action mode between the first image 210 and the second image 220. To achieve the purpose of power saving. In addition, when the time distance is long, the change between the two images is also large, so a larger first threshold value and a second threshold value can be set to accurately determine the action mode. For example, the time distance can be set to have a positive correlation with the first threshold, and the time distance has a positive correlation to the second threshold. For example, since the time distance d _{1 is} less than the time distance d ₂ , when determining the action mode between the first image 210 and the second image 220, the first threshold value (or the second threshold value) used is less than The first critical value (or the second critical value) used when determining an operation mode between the second image 220 and the third image 230. In other words, the number of critical values will be multiplied by two, referring to the description in the previous paragraph, a total of 6 or 8 critical values can be set. In the embodiment, different thresholds are set according to the time interval and the operation mode. Compared with the prior art which only uses fixed thresholds (only two are required), this embodiment has at least the power consumption saving and accurate judgment of the action. Advantages of the model.

The embodiment in FIG. 2 is merely an example. In some embodiments, the time distance d ₁ may be set equal to the time distance d ₂ , or the time distance d _{1 is} greater than the time distance d ₂ , and the present invention is not limited thereto.

Referring to FIG. 1, after the above-mentioned motion detection method is executed, the detection result is transmitted to the logic circuit 130, and the logic circuit 130 can execute an arbitrary program according to the judgment result, and the present invention is not limited thereto. For example, after determining the dynamic mode, the logic circuit 130 may store the captured image or determine whether there is a specific object in the image, and so on.

FIG. 4 is a flowchart illustrating a motion detection method according to an embodiment. Referring to FIG. 4, in step 401, a first image and a second image are obtained, the first image is divided into a plurality of first blocks, and the second image is divided into a plurality of second blocks. The blocks are respectively corresponding to the second blocks. In step 402, a block difference value between each first block and a corresponding second block is calculated, and it is determined whether the block difference value is greater than a first critical value. In step 403, the number of difference blocks is calculated. The number of difference blocks indicates that the difference value of several blocks in the first image or the second image is greater than the first critical value. In step 404, it is determined whether the number of difference blocks is greater than a second threshold. If the result of step 404 is YES, it is determined in step 405 that the operation mode is a dynamic mode. If the result of step 404 is no, it is determined in step 406 that the operation mode is a static mode. However, the steps in FIG. 4 have been described in detail above, and will not be repeated here. It should be noted that each step in FIG. 4 can be implemented as multiple codes or circuits, and the present invention is not limited thereto. In addition, the method of FIG. 4 can be used in conjunction with the above embodiments, or it can be used alone. In other words, the steps of FIG. 4 are also used. Other steps can be added.

Although the present invention has been disclosed as above with the examples, it is not intended to limit the present invention. Any person with ordinary knowledge in the technical field can make some modifications and retouching without departing from the spirit and scope of the present invention. The protection scope of the present invention shall be determined by the scope of the attached patent application.

Claims (10)

An image motion detection method is executed by an electronic device. The motion detection method includes: acquiring a first image and a second image, dividing the first image into a plurality of first blocks, and dividing the first image into a plurality of first blocks. The two images are divided into a plurality of second blocks, where the first blocks are respectively corresponding to the second blocks; one between each of the first blocks and the corresponding second block is calculated Block difference value, and determine whether the block difference value is greater than a first critical value; calculate a number of difference blocks, and the number of difference blocks indicates how many of the block differences are in the first image or the second image The value is greater than the first critical value; and if the number of difference blocks is greater than a second critical value, it is judged that an action mode is a dynamic mode, otherwise the action mode is a static mode, where the first A critical value is less than the first critical value in the static mode, or the second critical value in the dynamic mode is less than the second critical value in the static mode. The motion detection method according to item 1 of the scope of patent application, wherein the step of calculating the block difference value between each of the first blocks and the corresponding second block includes: for each of the A first histogram is calculated for pixels in the first block; a second histogram is calculated for each pixel in the second blocks; and each of the first histograms and the corresponding first histogram are calculated. The difference between the two histograms is used as the difference value of the block. The motion detection method as described in item 2 of the scope of patent application, wherein the block difference value is an absolute difference sum. The motion detection method according to item 1 of the scope of patent application, further comprising: setting the first threshold value or the second threshold value according to a time distance between the first image and the second image, wherein the The time distance is positively correlated with the first critical value, or the time distance is positively correlated with the second critical value. The motion detection method according to item 1 of the scope of patent application, further comprising: determining the motion mode between the second image and a third image, wherein the time distance between the first image and the second image Less than the time distance between the second image and the third image. An electronic device includes: a permanently activated sensor for obtaining a first image and a second image; and an arithmetic circuit for dividing the first image into a plurality of first blocks and dividing the first image into a plurality of first blocks. The second image is divided into a plurality of second blocks, where the first blocks are respectively corresponding to the second blocks, and the operation circuit is used to calculate each of the first blocks and the corresponding first block. A block difference value between two blocks, and determining whether the block difference value is greater than a first critical value, the operation circuit is used to calculate a number of difference blocks, and the number of difference blocks indicates the first image or There are several difference values of the block in the second image that are greater than the first threshold value. If it is determined that the number of difference blocks is greater than a second threshold value, the arithmetic circuit determines that an action mode is a dynamic mode, otherwise the action The mode is a static mode, wherein the first critical value in the dynamic mode is smaller than the first critical value in the static mode, or the second critical value in the dynamic mode is smaller than the first critical value in the static mode. Two critical values. The electronic device according to item 6 of the scope of patent application, wherein the operation circuit is further configured to calculate a first histogram for each pixel in the first blocks, and for each of the second blocks Calculate a second histogram, and calculate the differences between each of the first histograms and the corresponding second histograms as the block difference value. The electronic device according to item 7 of the scope of patent application, wherein the block difference value is an absolute difference sum. The electronic device according to item 6 of the scope of patent application, wherein the arithmetic circuit is further configured to set the first threshold value or the second threshold value according to a time distance between the first image and the second image, The time distance is positively correlated with the first critical value, or the time distance is positively correlated with the second critical value. The electronic device according to item 6 of the scope of patent application, wherein the arithmetic circuit is further configured to determine the action mode between the second image and a third image, wherein the first image and the second image are between the first image and the second image. The time distance is less than the time distance between the second image and the third image.