Image analysis system, program, and image analysis method

Abstract

To provide an image analysis system with improved convenience.SOLUTION: An image analysis system is provided, comprising a plurality of image analysis units, a selection unit for selecting at least one of the plurality of image analysis units, an analysis control unit configured to cause the selected image analysis unit to analyze an image specified by a user, and an output unit for outputting a result of the image analysis.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present invention relates to an image analysis system and an image analysis method. [Background technology]

[0002] Patent Document 1 discloses a technique for analyzing an image and detecting a suspicious person, etc. Patent Documents 2 and 3 disclose an index generating device for generating an index in which a plurality of nodes are hierarchically organized. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2017-191621 A [Patent Document 2] International Publication No. 2014 / 109127 [Patent Document 3] JP 2015-49574 A Summary of the Invention [Problem to be solved by the invention]

[0004] In image analysis, there are various types of analysis, such as face analysis that extracts facial feature values, posture analysis that extracts human posture feature values, and vehicle analysis that extracts vehicle feature values. However, if one system supports only one type of image analysis and can only perform one type of image analysis, the system is inconvenient. Patent Documents 1 to 3 do not disclose the problem or a means for solving the problem. The problem of the present invention is to improve the convenience of an image analysis system. [Means for solving the problem]

[0005] According to the present invention, A plurality of image analysis means; A selection means for selecting at least one of the plurality of image analysis means; analysis control means for causing the selected image analysis means to analyze an image designated by a user; an output means for outputting a result of the image analysis; An image analysis system is provided having the following:

[0006] Further, according to the present invention, The computer Selecting at least one of a plurality of image analysis methods; causing the selected image analysis means to analyze an image designated by a user; An image analysis method is provided that outputs a result of the analysis of the image. Effect of the Invention

[0007] According to the present invention, the convenience of the image analysis system is improved. [Brief description of the drawings]

[0008] [Figure 1] FIG. 1 is a diagram illustrating an example of a hardware configuration of an image analysis system according to an embodiment of the present invention. [Diagram 2] FIG. 2 is an example of a functional block diagram of the image analysis system according to the present embodiment. [Diagram 3] 4 is a diagram for explaining an example of processing executed by the image analysis system of the present embodiment. FIG. [Figure 4] 1 is a configuration example of an image analysis system according to an embodiment of the present invention. [Diagram 5] 3 is a diagram illustrating an example of information processed by the image analysis system according to the present embodiment. FIG. [Figure 6] FIG. 2 is a diagram illustrating an example of information output by the image analysis system according to the present embodiment. [Figure 7] 4 is a flowchart showing an example of a processing flow of the image analysis system of the present embodiment. [Figure 8] FIG. 2 is a diagram illustrating an example of information output by the image analysis system according to the present embodiment. [Figure 9] FIG. 2 is a diagram illustrating an example of information output by the image analysis system according to the present embodiment. [Figure 10] FIG. 2 is a diagram illustrating an example of information output by the image analysis system according to the present embodiment. [Figure 11] FIG. 2 is a diagram illustrating an example of information output by the image analysis system according to the present embodiment. [Figure 12] FIG. 2 is a diagram illustrating an example of information output by the image analysis system according to the present embodiment. [Figure 13] FIG. 2 is a diagram illustrating an example of information output by the image analysis system according to the present embodiment. [Figure 14] FIG. 2 is a diagram illustrating an example of information output by the image analysis system according to the present embodiment. [Figure 15] FIG. 2 is a diagram illustrating an example of information output by the image analysis system according to the present embodiment. [Figure 16] FIG. 2 is a diagram illustrating an example of information output by the image analysis system according to the present embodiment. [Figure 17] FIG. 2 is a diagram illustrating an example of information output by the image analysis system according to the present embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0009] Hereinafter, embodiments of the present invention will be described with reference to the drawings. In all the drawings, the same components are given the same reference numerals and the description will be omitted as appropriate.

[0010] "overview" The image analysis system of this embodiment (hereinafter, "image analysis system 10") includes a plurality of image analysis means for performing different types of image analysis. The image analysis system 10 analyzes an image specified by a user with the image analysis means selected based on user input, and outputs the analysis result. Such an image analysis system 10 improves convenience.

[0011] "composition" Next, the configuration of the image analysis system 10 will be described. First, an example of the hardware configuration of the image analysis system 10 will be described. Each functional part of the image analysis system 10 is realized by any combination of hardware and software, centering on a CPU (Central Processing Unit) of any computer, memory, programs loaded into the memory, a storage unit such as a hard disk for storing the programs (programs stored in advance at the stage of shipping the device, as well as programs downloaded from storage media such as CDs (Compact Discs) or servers on the Internet, can be stored), and a network connection interface. Those skilled in the art will understand that there are various variations in the method of realizing the system and the device.

[0012] FIG. 1 is a block diagram illustrating a hardware configuration of an image analysis system 10. As shown in FIG. 1, the image analysis system 10 has a processor 1A, a memory 2A, an input / output interface 3A, a peripheral circuit 4A, and a bus 5A. The peripheral circuit 4A includes various modules. The image analysis system 10 does not need to have the peripheral circuit 4A. The image analysis system 10 may be composed of a plurality of physically and / or logically separated devices, or may be composed of a single device that is physically and / or logically integrated. When the image analysis system 10 is composed of a plurality of physically and / or logically separated devices, each of the plurality of devices can have the above hardware configuration.

[0013] The bus 5A is a data transmission path for the processor 1A, memory 2A, peripheral circuit 4A, and input / output interface 3A to transmit and receive data to each other. The processor 1A is an arithmetic processing device such as a CPU or a GPU (Graphics Processing Unit). The memory 2A is a memory such as a RAM (Random Access Memory) or a ROM (Read Only Memory). The input / output interface 3A includes an interface for acquiring information from an input device, an external device, an external server, an external sensor, a camera, etc., and an interface for outputting information to an output device, an external device, an external server, etc. Examples of the input device include a keyboard, a mouse, a microphone, a physical button, a touch panel, etc. Examples of the output device include a display, a speaker, a printer, a mailer, etc. The processor 1A can issue commands to each module and perform calculations based on the results of those calculations.

[0014] Next, the functional configuration of the image analysis system 10 will be described. Fig. 2 shows an example of a functional block diagram of the image analysis system 10. As shown in the figure, the image analysis system 10 has a plurality of image analysis units 11, an input unit 14, an analysis control unit 15, a storage unit 16, and an output unit 17. The input unit 14 has a selection unit 12 and an image designation unit 13. The configuration of each functional unit will be described in detail below.

[0015] The image analysis unit 11 analyzes an image. The multiple image analysis units 11 perform different types of image analysis. Examples of the types of image analysis include face recognition, human recognition, posture estimation, appearance attribute estimation, image gradient feature detection, image color feature detection, and object recognition. The image analysis performed by the multiple image analysis units 11 may be any of the types exemplified here.

[0016] In face recognition, facial features of a person are extracted. Furthermore, similarities between facial features may be compared and calculated (e.g., to determine whether the people are the same person). In human form recognition, human body features (for example, overall features such as body fatness or thinness, height, clothing, etc.) are extracted. Furthermore, similarities between human body features may be compared and calculated (e.g., to determine whether the people are the same person). In posture estimation, a person's joint points are recognized and connected to form a stick figure. Then, information about the stick figure is used to estimate the person's height and extract posture features. Furthermore, similarities between posture features may be compared and calculated (e.g., to determine whether the people are performing the same action).

[0017] Appearance attribute estimation recognizes appearance attributes associated with a person (e.g., clothing color, shoe color, hairstyle, wearing a hat, necktie, etc., for a total of more than 100 types of appearance attributes). In addition, the recognized appearance attributes may be matched and calculated for similarity (it is possible to determine whether they are the same attribute). Image gradient features include SIFT, SURF, RIFF, ORB, BRISK, CARD, HoG, etc. Image color features include color histograms, etc. Object recognition is realized using engines such as YoLo (which can extract general objects [e.g., cars, bicycles, chairs, etc.] and people).

[0018] The above is merely an example, and other image analyses may be performed. For example, any conventional technology may be used, such as vehicle analysis for extracting both feature amounts. The image analysis unit 11 stores the analysis results in the storage unit 16.

[0019] The image analysis unit 11 may store raw data of the analysis results generated by the image analysis in the storage unit 16. The raw data includes various feature amounts extracted from each image.

[0020] Additionally, the image analysis unit 11 may store the calculation results obtained by performing various processes such as compilation and editing based on the raw data in the storage unit 16. For example, the image analysis unit 11 may group multiple feature amounts extracted from multiple images (e.g., multiple frame images included in a moving image) into groups of similar features, and store information indicating the grouping results in the storage unit 16. By this grouping, it is possible to group feature amounts (such as facial feature amounts) of the same person extracted from multiple images, group feature amounts of similar postures extracted from multiple images, or group feature amounts of the same vehicle model extracted from multiple images.

[0021] Here, an example of a method for implementing the grouping will be described. For example, the similarity between a feature extracted from an image and all feature extracted previously may be calculated, and images with similarity equal to or greater than a reference value may be grouped together. However, in this process, the more the number of extracted feature values ​​increases, the greater the number of pairs for which similarity is calculated, which increases the processing load on the computer. Therefore, for example, the following method may be adopted.

[0022] For example, the extracted feature quantities may be indexed as shown in Fig. 3, and the above grouping may be realized using the index. By using the index, the processing speed can be increased. Details of the index and the method of generating it are disclosed in Patent Documents 2 and 3. Below, the structure of the index in Fig. 3 and the method of using it will be briefly described.

[0023] The extraction ID (identifier): "FXXX-XXX" shown in Figure 3 is identification information attached to each feature (human face feature, posture feature, vehicle feature, etc.) extracted from each image. FXXX is the identification information of the image, and the part after the hyphen is the identification information of each feature extracted from each image.

[0024] In the third layer, nodes corresponding to all the extracted IDs obtained from the images processed up to that point are placed. Then, the multiple nodes placed in the third layer are grouped together with those whose feature similarity is at or above the first level. A group ID is assigned to each group in the third layer. The grouping results in the third layer are the final grouping results.

[0025] In the second layer, one node (representative) selected from each of the multiple groups in the third layer is placed and linked to the corresponding group in the third layer. The multiple nodes placed in the second layer are grouped together with nodes having a similarity of a second level or higher. The second level of similarity is lower than the above-mentioned first level. In other words, nodes that would not be grouped when the first level is used as the standard may be grouped when the second level is used as the standard.

[0026] In the first layer, one node (representative) selected from each of the multiple groups in the second layer is placed and linked to the corresponding group in the second layer.

[0027] The image analysis unit 11, for example, indexes the obtained multiple extraction IDs as shown in Fig. 3, and when a new extraction ID is obtained, adds a node of the new extraction ID to the index. An example of the addition process will be described below.

[0028] First, the image analysis unit 11 compares a plurality of extraction IDs located in the first layer. The image analysis unit 11 generates pairs between the new extraction ID and each of the plurality of extraction IDs located in the first layer, and calculates the similarity of the feature amount for each pair. Then, the image analysis unit 11 determines whether the calculated similarity is equal to or greater than a first threshold value.

[0029] If there is no extraction ID in the first layer whose similarity is equal to or greater than the first threshold, the image analysis unit 11 adds a node corresponding to the new extraction ID to the first to third layers and links them together. In the second and third layers, a new group is generated by the added node of the new extraction ID. Then, a new group ID is issued corresponding to the new group in the third layer.

[0030] On the other hand, if an extraction ID whose similarity is equal to or greater than the first threshold exists in the first layer, the image analysis unit 11 moves the comparison target to the second layer. Specifically, the group in the second layer linked to the "extraction ID in the first layer whose similarity is determined to be equal to or greater than the first threshold" is set as the comparison target.

[0031] Then, the image analysis unit 11 generates pairs between the new extraction ID and each of the extraction IDs included in the group to be processed in the second layer, and calculates the similarity of the feature amount for each pair. Then, the image analysis unit 11 determines whether the calculated similarity is equal to or greater than a second threshold. Note that the second threshold is higher than the first threshold.

[0032] If there is no extraction ID with a similarity equal to or greater than the second threshold value in the group to be processed in the second layer, the image analysis unit 11 adds a node corresponding to the new extraction ID to the group to be processed in the second layer, and adds a node corresponding to the new extraction ID to the third layer and links them together. In the third layer, a new group is generated by the node of the added new extraction ID. Then, a new group ID is issued corresponding to the new group in the third layer.

[0033] On the other hand, if an extraction ID whose similarity is equal to or greater than the second threshold exists in the group to be processed in the second layer, the image analysis unit 11 adds a node corresponding to the new extraction ID to the group in the third layer to which the extraction ID whose similarity is equal to or greater than the second threshold belongs. This concludes the explanation of the configuration of the index in FIG. 3 and how to use it.

[0034] The image analysis system 10 may be configured so that the number of image analyzers 11 can be increased by installing a plug-in. By configuring in this way, the types of image analysis that the image analysis system 10 can perform can be increased by the relatively simple process of installing a plug-in. The more types of image analysis that the image analysis system 10 can perform, the more convenient it becomes.

[0035] Returning to FIG. 2, the input unit 14 accepts user input. The user input may be realized via an input device such as a touch panel, a keyboard, a mouse, a physical button, or a microphone connected to the image analysis system 10. Alternatively, the image analysis system 10 may be a server of a client-server system as shown in FIG. 4. In this case, the user input is realized via a user terminal 20 (client) communicatively connected to the image analysis system 10 via a communication network. Examples of the user terminal 20 include, but are not limited to, a personal computer, a smartphone, a tablet terminal, a smart watch, and a mobile phone.

[0036] The selection unit 12 selects at least one from among the multiple image analysis units 11. For example, the selection unit 12 may present multiple types of image analysis corresponding to each of the multiple image analysis units 11, such as "face analysis," "posture analysis," and "vehicle analysis," in a selectable manner, and accept a user input to select at least one from among them. Then, the selection unit 12 may select the image analysis unit 11 corresponding to the image analysis selected by the user.

[0037] Alternatively, as shown in FIG. 5, linking information linking an analysis purpose with at least one type of image analysis (at least one image analysis unit 11) to be performed for each analysis purpose may be generated in advance and stored in the image analysis system 10. The selection unit 12 may then refer to the linking information and select the image analysis unit 11 linked to the analysis purpose selected by the user. Examples of analysis purposes include, but are not limited to, "suspicious person detection," "customer trend analysis," and "vehicle trend analysis." The "report format" shown in FIG. 5 will be described later.

[0038] The image designation unit 13 designates an image to be analyzed based on a user operation. The designated image may be a moving image or a still image. The user operation may be an operation to designate one of images stored in a storage device accessible by the image analysis system 10, an operation to input a new image to the image analysis system 10, or other operations.

[0039] The input unit 14 can receive various other inputs related to image analysis. Examples of the various other inputs include, but are not limited to, input of a comment to be added to the image analysis, input of related information of the specified image (shooting date and time, shooting location, etc.), input of specifying a part of the specified video image to be analyzed (e.g., specifying the elapsed time from the beginning), etc.

[0040] The analysis control unit 15 causes at least one image analysis unit 11 selected by the selection unit 12 to analyze the image designated by the user.

[0041] Returning to FIG. 2, the output unit 17 outputs the result of the image analysis performed by the image analysis unit 11.

[0042] As an example, the output unit 17 can output an analysis report generated based on the result of the analysis of the image. The analysis report can include at least one of the following items.

[0043] A list of features extracted from the specified image. The result of grouping the features extracted from the specified images. For example, the number of groups and related information for each group (representative features, representative images, etc.) may be displayed. Analysis results for each group (number of occurrences, changes in number of occurrences over time, etc.).

[0044] The format of the analysis report may be defined in advance. For example, the type of information to be displayed in the analysis report (e.g., any of the above-mentioned examples), the display method (graph type, graph format, etc.), the layout of the information in the analysis report, the character size, the character format, the image size, etc. may be defined. Then, the output unit 17 may generate and output the analysis report according to the definition.

[0045] 5, the format of the analysis report may be defined for each analysis purpose. As described above, examples of the analysis purpose include, but are not limited to, "suspicious person detection," "customer trend analysis," and "vehicle trend analysis." The output unit 17 may generate and output an analysis report according to the definition corresponding to the analysis purpose selected by the user.

[0046] The above-mentioned definition of the analysis report format may be customized by the user (modifying an existing definition, adding a new definition, deleting an existing definition, etc.). The output unit 17 may add or change the definition of the analysis report format based on a user input.

[0047] Also, the user may be able to add, change, or delete combinations of "analysis purpose," "image analysis method," and "report format" shown in Fig. 5. In other words, the image analysis system 10 may update the information shown in Fig. 5 based on user input.

[0048] Furthermore, the output unit 17 may output raw data generated by the image analysis by the image analysis unit 11. The output unit 17 may be capable of exporting the raw data, downloading it, or moving it to another storage device.

[0049] The output unit 17 may also include a means for notifying the user when the image analysis by the image analysis unit 11, which has been started in response to a user operation, is completed. For example, as shown in FIG. 6, the completion of the analysis may be notified by email to a preregistered email address of the user. This email may include an analysis report attached thereto, or may include a URL of the analysis report published on the web. Note that instead of / in addition to the email notification, other means such as a push notification of an app may be used.

[0050] Next, an example of the flow of processing by the image analysis system 10 will be shown with reference to the flowchart of FIG.

[0051] First, the image designation unit 13 designates an image to be analyzed based on a user operation (S10). The user operation may be an operation to designate one of images stored in a storage device accessible to the image analysis system 10, an operation to input a new image to the image analysis system 10, or any other operation.

[0052] Next, the input unit 14 sets various items including the image analysis method based on a user operation (S11). Setting the image analysis method means selection of the image analysis unit 11.

[0053] For example, the selection unit 12 of the input unit 14 may present multiple types of image analyses (multiple types of image analysis units 11) that the image analysis system 10 can perform, such as “face analysis,” “posture analysis,” “vehicle analysis,” etc., in a selectable manner, and accept user input to select at least one of them.

[0054] 5, linking information linking an analysis purpose with at least one type of image analysis (at least one image analysis unit 11) to be executed for each analysis purpose may be generated in advance and stored in the image analysis system 10. Then, the selection unit 12 may refer to the linking information and select the image analysis unit 11 linked to the analysis purpose selected by the user.

[0055] In addition, the input unit 14 may accept input of comments to be added to the image analysis, input of related information for the specified image (such as the shooting date and time, the shooting location, etc.), and input specifying a part of the specified video to be analyzed (such as specifying the elapsed time from the beginning).

[0056] The processing order of S10 and S11 is not limited to that shown in the figure.

[0057] Next, the analysis control unit 15 causes the selected image analysis unit 11 to analyze the designated image (S12). After that, the output unit 17 outputs the result of the image analysis (S13).

[0058] Next, an example of a UI (user interface) screen output by the output unit 17 will be described.

[0059] 8 shows a list of image analysis requests. After logging in to the image analysis system 10, the user can call up a UI screen showing a list of image analysis requests that the user has made.

[0060] In the illustrated image analysis request list, the analysis ID (identifier), request ID, image file name, analysis method, comment, status, registration date, execution date, and result are linked. Note that some of these items may not be included, and other items may be included.

[0061] The analysis ID is information that identifies the analysis from other analyses. The request ID is information that identifies the analysis request from other requests. The image file name is the file name of the image specified in the analysis. The analysis method is the analysis method selected for that analysis. The comment is a comment registered by the user in association with the analysis. The status indicates the current state of the analysis, and may be, for example, completed, incomplete, or an error. The registration date is the date the request for the analysis was registered. The execution date is the date the analysis was performed. The results include a link to the analysis results. Selecting "PDF" as shown opens a PDF file showing the analysis report. Selecting "Web Page" opens a web page showing the analysis report.

[0062] Note that the administrator of the image analysis system 10 may also be able to call up a UI screen showing a list of image analysis requests as shown in Fig. 8 after logging in to the image analysis system 10. While the UI screen of each user displays a list of image analysis requests made by each user, the UI screen of the administrator displays a list of all image analysis requests made by all users.

[0063] 9 shows a list of images stored in a storage device accessible to the image analysis system 10. A user can store images in the storage device by any means, such as uploading or inputting images. After logging in to the image analysis system 10, the user can call up a UI screen showing a list of images that the user has stored in the storage device.

[0064] In the illustrated image file list, the file ID, thumbnail image, image file name, shooting date and time, playback time, registration date, file size, analysis history, and a button for new analysis are linked. Note that some of these items may not be included, and other items may be included.

[0065] The file ID is information that identifies the image from other images. The thumbnail is a thumbnail image of the image. The image file name is the file name of the image. The photographing date and time is the date and time when the image was photographed. The playback time is the playback time of the image (movie image). The registration date is the date on which the image was stored in the storage device. The file size is the file size of the image. The analysis history is the history of the analysis methods that have been executed on the image. When the New Analysis button is operated, a UI screen for requesting a new analysis of the image is called up. By performing a specific operation via the UI screen, a new analysis of the image can be executed.

[0066] Note that the administrator of the image analysis system 10 may also be able to call up a UI screen showing a list of image files as shown in Fig. 9 after logging in to the image analysis system 10. On the UI screen of each user, a list of images stored by each user in the storage device is displayed, whereas on the UI screen of the administrator, a list of all images stored by all users in the storage device is displayed.

[0067] Fig. 10 shows the browsing history of the analysis results. The user can browse the analysis results from, for example, the UI screens shown in Fig. 8 and Fig. 10. The browsing history can be confirmed on the UI screen. After logging in to the image analysis system 10, the user can call up a UI screen showing the user's browsing history.

[0068] In the analysis result browsing history shown in the figure, the analysis ID, the analysis method, the last browsing date, and the result are linked. Note that some of these items may not be included, or other items may be included.

[0069] The analysis ID is information that identifies the analysis from other analyses. The analysis method is the analysis method selected for that analysis. The last viewed date is the date the result was last viewed. The results include a link to the analysis results. Selecting "PDF" as shown opens a PDF file showing the analysis report. Selecting "Web Page" opens a web page showing the analysis report.

[0070] Note that the administrator of the image analysis system 10 may also be able to call up a UI screen showing the browsing history of the analysis results as shown in Fig. 10 after logging in to the image analysis system 10. While the UI screen for each user displays the browsing history of each user, the UI screen for the administrator displays the browsing history of all users.

[0071] Figure 11 shows an example of the results of the analysis presented in a PDF file.

[0072] The "Video Data" column shows detailed information about the analyzed image. In the "Thumbnail" column, a thumbnail image of the analyzed image is shown. The "List of people grouped by features" column shows the results of grouping the features extracted from the analyzed images. The "Frequency of GIDs Appearing" column shows the number of times that each group's feature appears in the analyzed image. For example, a video may be divided into multiple windows of a predetermined length, and the number of windows in which the GID appears may be used as the number of times that the GID appears. The "Proportion of appearance in video (%)" column indicates the proportion of time that the feature quantity of each group appears relative to the duration of the analyzed image (video).

[0073] FIG. 12 shows an example of the results of the analysis presented on a web page.

[0074] The "List of people grouped by features" column shows the results of grouping the features extracted from the analyzed images. The "Frequency of GIDs Appearing" column shows the number of times that each group's feature appears in the analyzed image. For example, a video may be divided into multiple windows of a predetermined length, and the number of windows in which the GID appears may be used as the number of times that the GID appears. The "Proportion of appearance in video (%)" column indicates the proportion of time that the feature quantity of each group appears relative to the duration of the analyzed image (video).

[0075] Fig. 13 shows an example of the analysis result presented on a web page. For example, in response to an operation of selecting one group (any one of Gid1 to 23) on the web page of Fig. 12, the screen shown in Fig. 13 is displayed.

[0076] At the top of the screen, a representative thumbnail image of the selected group and a number of other thumbnail images are displayed.

[0077] The column "Time series detection for analysis ID=182" shows the change over time in the number of times the features of the selected group appear in one analyzed image (one video image). The "Time series of all analysis IDs" column shows the time change in the number of times the feature of the selected group appears in each of the analyzed images (video images). Note that while the figure shows the analysis results of only one image, if multiple images are analyzed, broken lines corresponding to each of the images will be displayed, as shown in Figure 14.

[0078] Returning to Fig. 13, the "Nearby video" column shows a scene from an analyzed image. For example, in the "Time series detection for analysis ID=182" column or the "Time series of all analysis IDs," playback starts from the date and time specified in the "Nearby video" column in response to an operation to select an arbitrary date and time on the graph.

[0079] FIG. 15 shows an example of the analysis results presented on a web page. In FIG. 15, the content displayed in the "Nearby Images" column differs from that in FIG. 13. The other configurations are the same as in FIG. 13. FIG. 15 shows the results of posture analysis. In this case, as shown in the figure, the skeletal structure of the detected person may be displayed in the "Nearby Images" column.

[0080] 16 shows an example of the analysis results presented on a web page. On this screen, the results of a plurality of analyses on a plurality of images (a plurality of video images) are displayed together.

[0081] The "List of people grouped by features" column shows the results of grouping features extracted from the analyzed images (video images). The "Frequency of GIDs and Analysis IDs" column shows the number of times each group's feature appeared across all the analyzed images (videos). It also shows which analysis ID each feature was extracted with (which is equivalent to "which video analysis it was extracted with"). The "Proportion of appearance in video (%)" column indicates the proportion of time that each group's feature appears relative to the total duration of the analyzed images (video images). The "Proportion of appearance in multiple images" column shows the proportion of the number of times the feature was detected in each image to the number of times the feature was detected across all the analyzed images (multiple videos). In the "List of determined faces" field, a representative thumbnail image of the selected group and a plurality of other thumbnail images are shown.

[0082] When displaying the results of multiple analyses on multiple images (multiple video images) together, a section may be provided in which multiple images are simultaneously played back and displayed in time synchronization, as shown in Fig. 17. For example, such images may be displayed in the "Nearby Images" section in Fig. 13 or Fig. 15.

[0083] "Action and effect" The image analysis system 10 of the present embodiment described above includes a plurality of image analysis means that perform different types of image analysis, and analyzes an image specified by a user with an image analysis means selected based on user input. Such an image analysis system 10 improves convenience.

[0084] In addition, the image analysis system 10 of the present embodiment can generate and output one analysis report that summarizes the results of multiple types of image analysis. Such an analysis report allows the user to analyze while comparing the results of multiple types of image analysis or checking them in parallel. As a result, it becomes possible to find information hidden in the image.

[0085] Furthermore, the image analysis system 10 of this embodiment can be configured to allow the user to customize the format of the analysis report. In this case, the user can obtain an analysis report that displays desired information in a desired layout, for example. As a result, the efficiency of checking and analyzing the analysis report is improved.

[0086] Furthermore, according to the image analysis system 10 of this embodiment, it is possible to define the format of the analysis report for each analysis purpose, and generate and output an analysis report in a format corresponding to the analysis purpose selected by the user. In this case, the user can obtain an analysis report in a format suitable for the analysis purpose. As a result, the efficiency of checking and analyzing the analysis report is improved.

[0087] Furthermore, the image analysis system 10 can be configured so that the number of image analyzers 11 can be increased by installing a plug-in. By configuring in this way, the types of image analysis that the image analysis system 10 can perform can be increased by the relatively simple process of installing a plug-in. The more types of image analysis that the image analysis system 10 can perform, the more convenient it becomes.

[0088] Furthermore, the image analysis system 10 may include a means for notifying the user when the image analysis is complete. In this case, the user does not need to wait until the image analysis is complete, and can do other things during that time. As a result, convenience is improved.

[0089] Furthermore, the image analysis system 10 may include a means for outputting raw data of the image analysis. In this case, the user can process or edit the raw data in his / her own way. As a result, the user's convenience is improved.

[0090] Furthermore, the wide variety of UI screens described above enable users and administrators to easily grasp various situations and check the results of image analysis.

[0091] A part or all of the above-described embodiments can be described as follows, but is not limited to the following. 1. Multiple image analysis methods; A selection means for selecting at least one of the plurality of image analysis means; analysis control means for causing the selected image analysis means to analyze an image designated by a user; an output means for outputting a result of the image analysis; An image analysis system having the above configuration. 2. The image analysis system described in 1, wherein the output means outputs an analysis report generated based on the results of the analysis of the image. 3. The format of the analysis report is defined for each analysis purpose; 3. The image analysis system according to claim 2, wherein the output means generates and outputs the analysis report in the format defined in accordance with the analysis purpose selected by the user. 4. The image analysis system according to claim 3, wherein the output means adds or changes the definition of the format of the analysis report based on user input. 5. An image analysis system according to any one of 1 to 4, wherein the output means comprises means for notifying the user when analysis of the image by the selected image analysis means has been completed. 6. An image analysis system according to any one of 1 to 5, wherein the output means outputs raw data generated by the analysis of the image. 7. An image analysis system according to any one of 1 to 6, in which the image analysis means can be increased by installing a plug-in. 8. An image analysis system according to any one of 1 to 7, wherein the selection means selects the image analysis means selected by a user. 9. An image analysis system described in any one of 1 to 8, wherein the selection means refers to linking information linking an analysis purpose with at least one of the image analysis means, and selects at least one of the image analysis means linked to the analysis purpose selected by the user. 10. The computer: Selecting at least one of a plurality of image analysis methods; causing the selected image analysis means to analyze an image designated by a user; The image analysis method further comprises outputting a result of the analysis of the image. [Explanation of symbols]

[0092] 10. Image Analysis System 11 Image Analysis Section 12 Selection section 13 Image specification section 14 Input section 15 Analysis control section 16 Memory section 17 Output section 1A Processor 2A Memory 3A input / output I / F 4A Peripheral circuit 5A Bus

Claims

1. A plurality of image analysis means; A selection means for selecting at least one of the plurality of image analysis means; analysis control means for causing the selected image analysis means to analyze an image designated by a user; an output means for outputting a result of the image analysis; An image analysis system having the above configuration.

2. The image analysis system according to claim 1 , wherein the output means outputs an analysis report generated based on a result of the analysis of the image.

3. The format of the analysis report is defined for each analysis purpose; 3. The image analysis system according to claim 2, wherein the output means generates and outputs the analysis report in the format defined in accordance with the analysis purpose selected by the user.

4. The image analysis system according to claim 3 , wherein the output means adds or changes a definition of a format of the analysis report based on a user input.

5. 5. The image analysis system according to claim 1, wherein the output means comprises means for notifying the user when analysis of the image by the selected image analysis means is completed.

6. The image analysis system according to claim 1 , wherein the output means outputs raw data generated by the analysis of the image.

7. 7. The image analysis system according to claim 1, wherein the image analysis means can be increased by installing a plug-in.

8. The image analysis system according to claim 1 , wherein the selection means selects the image analysis means selected by a user.

9. The image analysis system according to any one of claims 1 to 8, wherein the selection means refers to linking information linking an analysis purpose with at least one of the image analysis means, and selects at least one of the image analysis means linked to the analysis purpose selected by the user.

10. The computer Selecting at least one of a plurality of image analysis means; causing the selected image analysis means to analyze an image designated by a user; The image analysis method further comprises outputting a result of the analysis of the image.

Images