JP2005004564A - Image classifying and processing system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a database construction system for use in compiling image data such as a photographic image into a database. <P>SOLUTION: The image database construction system classifies and processes images in a short period of time by freely copying the basically arranged images while referring to the images themselves, then grouping the images, and giving a name to each group. New images displayed on a display are classified to attribute information similarly displayed on the display. The images to be classified can be classified to the appropriate attribute information and registered in the database by the mere operation of a mouse, etc. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an image classification processing system for classifying image data such as a photographic image for each attribute.
[0002]
[Prior art]
Today, with the spread of imaging devices such as digital cameras, a large amount of photographic images are generated, and these image data are classified and edited by a computer. In such classification and editing operations, the meanings extracted from photographic images (subjects) are diverse. For example, if a car, a building, and a person are shown, it can be classified into transportation technology, building technology, customs, folk, etc. according to the classification viewpoint. In addition, there is a case where a photographic image that has no meaning in one sheet is meaningful by being compared with an image of a completely different subject. Thus, image information is inherently ambiguous, and is premised on receiving multilateral evaluation.
[0003]
In this regard, Patent Document 1 discloses a system that classifies and searches a large number of images using place names and facility names as keywords. The present invention is an invention in which an image is stored in an image database in association with a keyword such as a position, the input keyword is input, and a corresponding image is read from the image database when searching for an image.
[0004]
Patent Document 2 discloses an invention of an album production system. In order to reduce the amount of work for an album editor and efficiently perform album production, a server and a user terminal are connected, and photographed images are classified into predetermined categories. The photographed image stored in the storage means is extracted based on the item of the album in response to a request from the user terminal.
[0005]
[Patent Document 1]
JP 2002-259389 A [Patent Document 2]
Japanese Patent Laid-Open No. 2002-8050
[Problems to be solved by the invention]
However, in the conventional image database classification method, since each image is taken into a computer and character attributes are input, excessive labor is required, and much time and cost are required.
[0007]
Therefore, according to the present invention, it is possible to perform image classification efficiently and perform image classification and image processing in a short time by copying, grouping, and giving names to groups while simultaneously referring to a large number of images. An image database construction system is provided. In addition, in the existing database that has been surveyed and organized, the attributes of the expanded subject and the image group are freely associated.
[0008]
[Means for Solving the Problems]
According to the first aspect of the present invention, the subject is grouped by the input means for inputting new image data, the classification means for classifying the image data input from the input means by visual information, and the classification means. This can be achieved by providing an image classification processing system having attribute information adding means for adding attribute information to an image and storage means for storing corresponding image information together with the attribute information.
[0009]
With this configuration, when a new image is classified, attribute information corresponding to the image can be added and recorded in the database.
According to the second aspect of the present invention, there is provided a storage means for storing an image based on attribute information, a selection means for selecting corresponding attribute information for new image data, and the selection This can be achieved by providing an image classification processing system having writing means for writing image data in the corresponding attribute information area of the storage means based on the result.
[0010]
With this configuration, when a new image is classified, the attributes of the already classified image data can be displayed on the display unit, and the new image can be classified through vision. In particular, since the image classification operation is performed on the basis of previously set attribute information, consistent attribute information can be set in a temporal flow.
[0011]
According to a third aspect of the present invention, in the first or second aspect of the invention, the selection process displays the new image data on a display unit, and displays an area corresponding to the attribute information on the display unit. By providing the new image data, the new image data is moved to an appropriate attribute information area.
[0012]
With this configuration, when a new photographic image is classified, the attributes already registered in the database can be displayed, and the new image can be classified through vision.
[0013]
According to a fourth aspect of the present invention, in the first, second, or third aspect, the new image data is included in a plurality of attribute information.
With this configuration, an attribute is added to image data that has been duplicated by copying, and a link code is added to the corresponding image file, so that new image classification processing across multiple attributes can be performed. it can. For example, when a certain photograph (image) is included in the attribute of “water storage dam” and further included in the attribute of “water gate”, it is registered in both attribute areas.
[0014]
According to a fifth aspect of the present invention, in the first, second, third, or fourth aspect, the storage unit is hierarchized.
For example, it is classified into upper and lower groups, and by configuring in this way, for example, classification based on commonality of photographic images can be constructed in a hierarchical structure.
[0015]
A sixth aspect of the present invention is a configuration in which the image information corresponding to the attribute information is edited in the first, second, third, fourth, or fifth aspect.
According to claim 7, in the description of claim 1, 2, 3, 4, 5, or 6, attribute information of the group is attributed as a classification name to a member image based on the group hierarchy. , Constituting one record of the storage means.
[0016]
According to the invention described in claim 8, there is provided an input process for inputting new image data, a classification process for classifying the input image data based on visual information, and attributes assigned to the images grouped by the classification process. This can be achieved by providing an image classification processing method that performs a process of adding information and a storage process of storing corresponding image information together with the attribute information.
[0017]
The present invention is an invention of a method, and by constructing an image classification processing method in this way, when a new image is classified, attribute information corresponding to the image is automatically added and recorded in the database. Can do.
[0018]
According to the invention described in claim 9, the object is to add new image data, to classify the input image data by visual information, and to add attribute information to the images grouped by the classification. It is a program that performs at least a function to perform the function to store the corresponding image information together with the attribute information, and can be achieved by providing a computer-executable program.
[0019]
The present invention is an invention of an image classification processing program. Even with this configuration, when a new image is classified, attribute information corresponding to the image can be automatically added and recorded in a database. it can.
[0020]
According to the tenth aspect of the present invention, there is provided a function for inputting new image data, a function for classifying input image data by visual information, and adding attribute information to images grouped by the classification. This can be achieved by providing a recording medium that records a program that performs at least the function of performing the function of storing the corresponding image information together with the attribute information.
[0021]
The present invention is an invention of a recording medium such as a CD-ROM or a floppy disk. By recording and distributing a program having the above functions on such a recording medium, the program is started on a computer, and a new one is created. Image classification processing can be performed automatically.
[0022]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments of the image database construction system of the present invention will be described in detail with reference to the drawings.
[0023]
FIG. 1 is a system configuration diagram of this example. In the figure, a computer 10 includes a CPU 11, a ROM 12, a RAM 13, and the like. The CPU 11 performs processing based on a program recorded in the ROM 12, and accesses an external storage device 14 connected to the computer 10 as necessary. , Send and receive data.
[0024]
In addition, necessary information is displayed on the display 15 of the computer 10, and information is exchanged via a communication line. The processing in this example is performed according to a program stored in the ROM 12, but a recording medium such as a flexible disk 17 or a CD-ROM 18 is attached to the media driver 16 provided in the computer 10 as shown in FIG. The program may be read from the recording medium and used.
[0025]
Next, the processing operation of this example will be described. FIG. 2 is a flowchart for explaining the processing operation of this example. First, image data is fetched (step (hereinafter referred to as S) 1). This capturing of image data is a process of opening a directory in which image information is recorded, specifying an image file to be read, and performing continuous reading. For example, in the case of image information recorded on a CD (compact disk), flash memory, SD card, etc., the driver loaded with the recording medium is accessed, and the directory in which the target image information is recorded is specified to display the image information. Batch import. Note that image information can be captured in any size, and up to 4,500 photographic images can be captured simultaneously.
[0026]
Next, the image data is taken into the virtual memory (S2) and displayed on the display (S3). The read image can be freely moved on the virtual memory. FIG. 3 shows a state in which the read image is freely moved on the display.
[0027]
Next, common images are grouped (S4). FIG. 4 shows an example in which images are grouped, and the images are classified by common items. At this time, the images are sequentially moved (drag and drop). Further, the image can be duplicated, and one image can be classified into a plurality of groups. Furthermore, images once grouped can be taken out and moved to another group.
[0028]
Next, a name is given to the grouped image group (S5). FIG. 5 shows a state in which a label describing a name is added to each group. For example, names common to the image group such as “water storage dam”, “sluice gate”, and “iron bridge” are described in this label. FIG. 6 shows an example in which the same image is included in a plurality of groups.
[0029]
In the image classification database, the grouped image data with the name added is recorded. FIG. 7 is a flowchart for explaining this process. In the figure, it is first determined whether or not a classification process has been performed on new image information (step (hereinafter referred to as ST) 1). If a classification process has been performed (ST1 is YES), the attribute information added is added. Information on the image is registered in the corresponding area (ST2).
[0030]
Thereafter, it is determined whether or not the classification process has been completed for all images (ST3), and the registration process is repeated until the classification process for all images is completed (ST3 is NO, ST1 to ST3). Thereafter, when the processing for all the images is completed, the classification processing is terminated (YES in ST3).
[0031]
On the other hand, when image information is newly added after the image classification database has already been constructed, processing is performed according to the flowchart shown in FIG. First, image data to be added is captured (step (hereinafter referred to as STP) 1) and recorded in the memory (STP2).
[0032]
Next, the captured image is overlaid on the display (STP3), and a common group is searched by referring to the group name displayed on the display and the image included in the group (STP4). For example, FIG. 9 shows an example in which an image 1 to be newly grouped is displayed in an overlapping manner, and already grouped images 2, 3, 4,. In addition, together with the image group 1 that needs to be newly classified on the display, for example, the “storage dam” group 2, the “water gate” group 3, the “iron bridge” group 4, and the “00”, “×” A large number of groups such as “XX” are displayed in an overlapping image together with the group name.
[0033]
First, a common group is searched for the image (1) shown in the figure, and the image (1) is moved to the area of the selected group name operated by the mouse (drag and drop) (STP5).
[0034]
Next, it is determined whether or not the classification process has been completed for all images (STP6). If the classification process has not been completed for all images (STP6 is NO), the process returns to the above process (STP4), and the next image (image Select the common group name for (2)).
[0035]
Thereafter, the above processing is performed sequentially. For example, in the example shown in FIG. 9, the same processing as image (2) → image (3) →... Is repeated, and when the selection processing for image (6) is completed, all processing is completed. Is completed, and the new image classification process is completed (YES in STP6). The image classification database is registered with new image data classified based on the attribute information.
[0036]
As described above, according to the present embodiment, it is possible to perform an appropriate classification process simply by dragging and dropping an image desired to be classified into the corresponding attribute information area. Therefore, it is possible to perform image classification and editing processing more efficiently and in a shorter time than in the past. In addition, consistent attribute information can be set for new image data.
[0037]
Next, the configuration of the database constructed in this example will be described. Before that, first, the relationship between the groups is defined. An inclusion relation is defined based on an operation of overlapping groups on a two-dimensional plane. The group of the group is set as a meta group, and a name is given to the meta group, and the name of the meta group becomes a higher classification. Then, as shown in FIG. 10, a plurality of groups are aggregated to form an upper group, and the aforementioned label is assigned to each group.
[0038]
Next, grouping and group nesting are important functions responsible for creating the above-mentioned classification. By grouping images and further grouping the groups, for example, the group inclusion relationship shown in FIG. Can be defined. This inclusion relationship is a nested structure of the group classification concept, the name representing the group is a classification name, and the nested structure is a classification thesaurus structure. Then, by outputting this structure to the database using the image as a key, a plurality of classifications can be assigned to the image.
[0039]
FIG. 12 shows the data structure of the image classification database. In the figure, field names BunruIDxx and Bunruxx can correspond to the tenth group. Note that the layout information in the figure stores the memory image as it is, so the layout data file is not included in the database.
[0040]
【The invention's effect】
As described above, according to the present invention, images such as photographic images can be classified into appropriate classifications in a short time, and consistent attribute information can be set.
[Brief description of the drawings]
FIG. 1 is a system configuration diagram of an image database construction system according to the present invention.
FIG. 2 is a flowchart illustrating processing of the present invention.
FIG. 3 is a diagram illustrating a state in which a read image is freely moved on a display.
FIG. 4 is an example in which images are grouped, and is classified by common items.
FIG. 5 is a diagram showing a state in which a label describing a name is added to each group.
FIG. 6 is a diagram illustrating an example when the same image is included in a plurality of groups.
FIG. 7 is a flowchart illustrating registration processing in an image classification database.
FIG. 8 is a flowchart illustrating a classification process for an additional image.
FIG. 9 is a diagram showing a processing screen for image classification processing;
FIG. 10 is a diagram illustrating an aggregation concept of a plurality of groups.
FIG. 11 is a diagram showing an inclusion relationship of groups.
FIG. 12 is a diagram illustrating an example of a data structure of an image classification database according to the present example.
[Explanation of symbols]
1 New image data 2 “Water storage dam” attribute information 3 “Sluice” attribute information 4 “Tetsubashi” attribute information 10 Computer 11 CPU
12 ROM
13 RAM
14 External storage device 15 Display 16 Media driver 17 Flexible disk 18 CD-ROM

Claims (10)

Input means for inputting new image data;
Classification means for classifying image data input from the input means according to visual information;
Attribute information adding means for adding attribute information to the images grouped by the classification means;
Storage means for storing corresponding image information together with the attribute information;
An image classification processing system comprising: Storage means for storing an image based on the attribute information;
Selecting means for selecting corresponding attribute information for new image data;
A writing means for writing image data in the corresponding attribute information area of the storage means based on the selection result;
An image classification processing system comprising: The selection process is performed by displaying the new image data on a display unit and moving the new image data to an appropriate attribute information area by providing an area corresponding to the attribute information in the display unit. The image classification processing system according to claim 1 or 2, wherein 4. The image classification processing system according to claim 1, wherein the new image data is included in a plurality of attribute information. 5. The image classification processing system according to claim 1, wherein the storage means is hierarchized. 6. The image classification processing system according to claim 1, further comprising means for editing image information corresponding to the attribute information. The attribute information of the group is attributed as a classification name to a member image based on the group hierarchy, and one record of the storage unit is configured. Or the image classification processing system according to 6. Input processing to input new image data;
A classification process for classifying input image data according to visual information;
A process of adding attribute information to the images grouped by the classification process;
Storage processing for storing corresponding image information together with the attribute information;
An image classification processing method. A function to input new image data;
A function to classify input image data by visual information;
A function of adding attribute information to images grouped by the classification;
A function of storing corresponding image information together with the attribute information;
A program that executes at least the program and is executable by a computer. A function to input new image data;
A function to classify input image data by visual information;
A function of adding attribute information to images grouped by the classification;
A function of storing corresponding image information together with the attribute information;
A recording medium for recording a program for performing at least.

Images