JP2022131194A - Information processing apparatus and image detection method - Google Patents

Abstract

To reduce the probability of omissions of extraction while maintaining the accuracy of detecting an image including an object identical with an object to be detected.SOLUTION: An information processing apparatus includes: a first recognition unit 22 which detects an image including an object whose similarity to a detection object is equal to or higher than a first threshold; a second recognition unit 23 which detects an image including an object whose similarity to the detection object is equal to or higher than a second threshold which is higher than the first threshold; a classification unit 24 which classifies images on the basis of at least one of time and date information indicating shooting time and date and location information indicating a shooting location; a characteristic information acquisition unit 25 which acquires characteristic information from an image group classified into the same category by the classification unit 24; and a detection unit 26 which detects an image including an object which is considered as the detection object, using a result detected by the first recognition unit 22, a result detected by the second recognition unit 23, and the characteristic information.SELECTED DRAWING: Figure 2

Description

The present invention relates to an information processing device and an image detection method.

2. Description of the Related Art Conventionally, techniques have been proposed for classifying a large amount of image data acquired by an imaging device such as a camera, based on features of a subject.
For example, Patent Literature 1 discloses an image classification device that focuses on a person's face in image data and automatically classifies the image data based on the features of the face.

JP 2005-107885 A

When classifying image data by focusing on the face of a person in an image, there is a possibility that different people with similar facial shapes or similarities in the image may be mistakenly recognized as the same person. There is
Further, by setting a high similarity threshold used for determining that the person is the same person, it is possible to improve the accuracy of determining the same person. However, simply increasing the threshold increases the possibility that the same person will be identified as a different person, and there is a possibility that images of the same person will not be extracted. be. Since there are a wide variety of shooting environments in which subjects are shot, it may not be possible to sufficiently extract the same person simply by adjusting the matching similarity threshold.
Moreover, the above-described problem is not limited to the case of specifying a person, and is a problem that similarly occurs when a target detection object other than a person is used, such as an industrial product or agricultural products.

SUMMARY OF THE INVENTION The present invention has been made in view of such circumstances, and provides information processing capable of reducing the probability of occurrence of omission of extraction while maintaining the detection accuracy of an image containing the same object as the object to be detected. It is an object to provide an apparatus and an image detection method.

A first aspect of the present invention is an information processing device for detecting an image in which a detection target is captured from among a plurality of images captured by an imaging device, wherein the degree of similarity to the detection target is first. A first recognition means for detecting an image containing an object equal to or greater than one threshold; 2 recognition means, classification means for classifying the plurality of images based on at least one of date and time information indicating the date and time when the image was taken and position information indicating the position where the image was taken, and the classification means. using the feature information acquisition means for acquiring feature information from the group of images classified into the same classification in the above, the detection result of the first recognition means, the detection result of the second recognition means, and the feature information, and detecting means for detecting an image of an object that can be regarded as the detection object.

Here, the above-mentioned "similarity" is an evaluation value indicating the matching rate between the detection object and the object, and is also expressed as "matching degree".
Further, the "similarity" is, for example, information specific to the detection object, for example, when the detection object is a person, based on the position of the eyes, nose and mouth in the person's face, the color of the eyes, the contour of the face, etc. to get the similarity. Also, as the "feature information", it is preferable to use features that are not specific to the detection object, for example, features that change depending on the day or situation. In this way, it is preferable to use feature information that was not used when obtaining the similarity as the "feature information".

A second aspect of the present invention is an image detection method for detecting an image in which a detection target is captured from among a plurality of images captured by an imaging device, wherein the degree of similarity to the detection target is second. A first recognition step of detecting an image containing an object equal to or greater than one threshold; 2 a recognition step, a classification step of classifying the plurality of images based on at least one of date and time information indicating the date and time when the image was taken and position information indicating the position where the image was taken, and the classification step. using the feature information acquisition step of acquiring feature information from the group of images classified into the same category in the above, the detection result of the first recognition step, the detection result of the second recognition step, and the feature information, An image detection method in which a computer executes a detection step of detecting an image in which an object that can be regarded as the detection object is imaged.

A third aspect of the present invention is a program for causing a computer to function as the information processing apparatus.

Advantageous Effects of Invention According to the present invention, it is possible to reduce the probability of occurrence of omission of extraction while maintaining the detection accuracy of an image containing the same object as the detection object.

1 is a schematic configuration diagram showing an example of a hardware configuration of an information processing apparatus according to a first embodiment of the present invention; FIG. 1 is a functional block diagram showing an extracted image detection function among various functions provided in an information processing apparatus according to a first embodiment of the present invention; FIG. 4 is a flow chart showing an example of the procedure of an image detection method according to the first embodiment of the present invention; It is a figure for demonstrating the image detection method which concerns on 1st Embodiment of this invention. It is a figure for demonstrating the image detection method which concerns on 1st Embodiment of this invention. It is a figure for demonstrating the image detection method which concerns on 1st Embodiment of this invention. It is a figure for demonstrating the image detection method which concerns on 1st Embodiment of this invention. FIG. 7 is a functional block diagram showing an extracted image detection function among various functions provided in an information processing apparatus according to a second embodiment of the present invention; 9 is a flow chart showing an example of the procedure of an image detection method according to the second embodiment of the present invention; It is a figure for demonstrating the image detection method based on 2nd Embodiment of this invention. It is a figure for demonstrating the image detection method based on 2nd Embodiment of this invention.

[First embodiment]
An information processing apparatus and an image detection method according to a first embodiment of the present invention will be described below with reference to the drawings. Examples of information processing apparatuses include notebook PCs, desktop PCs, tablet terminals, smart phones, digital cameras, and the like.

FIG. 1 is a schematic configuration diagram showing an example of the hardware configuration of an information processing device 1 according to the first embodiment of the present invention. As shown in FIG. 1, the information processing apparatus 1 includes, for example, a CPU (Central Processing Unit) 11, a main memory 12, a storage section 13, an external interface 14, a communication interface 15, an input section 16, a display section 17, and the like. there is These units are connected to each other directly or indirectly via a bus 18 and cooperate with each other to perform various processes. The information processing device 1 may also include an imaging device such as a camera.

The CPU 11 controls the entire information processing apparatus 1 by, for example, an OS (Operating System) stored in a storage unit 13 connected via a bus 18, and executes various programs stored in the storage unit 13. Executes various processes.

The main memory 12 is composed of a writable memory such as cache memory and RAM (Random Access Memory), and is used as a work area for reading the execution program of the CPU 11 and writing processing data by the execution program.

The storage unit 13 is, for example, a ROM (Read Only Memory), a HDD (Hard Disk Drive), a flash memory, or the like. It stores an OS for controlling the entire apparatus 1, various device drivers for hardware operation of peripheral devices, various application software, various data and files, and the like. The storage unit 13 also stores programs for implementing various processes and various data required for implementing various processes.

The external interface 14 is an interface for connecting with an external device. Examples of external devices include an external monitor, a USB memory, an external HDD, and the like. Although only one external interface is shown in the example shown in FIG. 1, a plurality of external interfaces may be provided.

The communication interface 15 functions as an interface for connecting to a network, communicating with other devices, and transmitting and receiving information.
The input unit 16 is a user interface such as a keyboard, mouse, touch panel, etc., for the user to give instructions to the information processing apparatus 1 .
The display unit 17 has a display screen configured by, for example, an LCD (Liquid Crystal Display), an organic EL (Electro Luminescence), or the like, and displays the result of an application or the like executed by the information processing apparatus 1. .

FIG. 2 is a functional block diagram showing an extracted image detection function among various functions provided in the information processing apparatus 1 according to the present embodiment. As shown in FIG. 2 , the information processing apparatus 1 includes an image detection section (image detection function) 20 . The image detection unit 20 has a function of detecting an image in which a detection object is imaged from among a plurality of images imaged by an imaging device such as a camera. Here, the object to be detected is not particularly limited, but examples thereof include people, agricultural products, industrial products, and the like.

A series of processes for realizing various functions described later are stored in the storage unit 13 in the form of a program (for example, an image detection program), for example. Various functions are realized by executing the processing and arithmetic processing of . The program may be pre-installed in the storage unit 13, may be provided in a state stored in another computer-readable storage medium, or may be distributed via wired or wireless communication means. may apply. Computer-readable storage media include magnetic disks, magneto-optical disks, CD-ROMs, DVD-ROMs, semiconductor memories, and the like.

As shown in FIG. 2, the image detection section 20 includes, for example, a storage section 21, a first recognition section 22, a second recognition section 23, a classification section 24, a feature information acquisition section 25, and a detection section .

The storage unit 21 stores a plurality of images (image data) captured by an imaging device such as a camera.

The first recognizing unit 22 detects, for example, an image including an object whose degree of similarity to the detected object is equal to or greater than a first threshold from among the plurality of images stored in the storage unit 21 .
Here, various known techniques such as image matching have been proposed as specific techniques for obtaining the degree of similarity between an object included in a plurality of images and a detection target. should be adopted. For example, when the object to be detected is a person, for example, a known face recognition technology that discriminates the person based on the feature amount of the face is adopted, and the object to be detected and the person included in each image are identified. You may acquire the similarity with a person image.

The second recognition unit 23 detects, for example, an image including an object whose similarity to the detected object is equal to or higher than the second threshold from among the plurality of images stored in the storage unit 21 . Here, the second threshold is set to a value higher than the first threshold.
For example, the second recognizing unit 23 may detect an image including an object having a degree of similarity equal to or greater than the second threshold from the images detected by the first recognizing unit 22 .

The classification unit 24 classifies the plurality of images based on at least one of date and time information indicating the date and time when the images were captured and position information indicating the positions at which the images were captured. For example, the classification unit 24 classifies images detected by the first recognition unit 22 based on at least one of date and time information and position information.

When sorting based on date and time information, sorting may be performed by date, sorting based on a time period, or sorting by a time span of two days or more. Further, if the number of images belonging to the same classification is a predetermined number or more when classified based on dates, etc., the images belonging to that classification may be further classified based on time information. Further, in this fine classification, classification may be performed using position information. In this way, position information may be used as information supplementary to time information.

When classification is performed based on date and time information, for example, the duration may be determined according to feature information acquired by the feature information acquisition unit 25, which will be described later. For example, clothing characteristics are likely to change from day to day or depending on the time of day, whereas body shape and hair color do not change for a long time compared to clothing characteristics. For this reason, the time width of the classification unit 24 may be changed according to what kind of information is to be acquired as feature information.

Furthermore, if an application related to a calendar is installed, it may be linked with calendar information, and the time span may be dynamically changed based on the calendar information. For example, if an event such as a wedding or party is registered in the calendar, the clothing may change in units of time. In this way, by dynamically changing the classification time span using the information registered in the calendar, it is possible to effectively classify the images.

In the case of classification based on position information, for example, images within a predetermined distance range may be classified as one classification.

The feature information acquisition unit 25 acquires feature information from the images classified into the same classification by the classification unit 24 and having a degree of similarity to the detection object equal to or greater than a first threshold. The feature information includes the color of at least part of the object, the shape of at least part of the object, the color of the article photographed accompanying the object, and the shape of the article photographed accompanying the object. including any one of For example, when the object to be detected and the object to be detected are a person, clothing color, clothing shape, hairstyle, hair color, ornament color, ornament shape, body shape, and the like can be mentioned.

Using the detection result of the first recognition unit 22, the detection result of the second recognition unit 23, and the feature information acquired by the feature information acquisition unit 25, the detection unit 26 captures an image of an object that can be regarded as a detection object. Detect images.

The detection unit 26 detects, for example, the image detected by the second recognition unit 23 as an image of an object that can be regarded as a detection object. Further, the detection unit 26 detects an image that is not detected by the second recognition unit 23 among the images detected by the first recognition unit 22, in other words, an object whose similarity is greater than or equal to the first threshold and less than the second threshold. Among the images detected by the second recognition unit 23, an image containing a target object having feature information that matches the feature information of the target object included in the image detected by the second recognition unit 23 is detected as an image in which the target object that can be regarded as the detection target object is captured. do.

Next, an example of the image detection method according to this embodiment will be described with reference to the drawings. FIG. 3 is a flow chart showing an example of the procedure of the image detection method according to this embodiment. Also, in the following description, for convenience of description, a person is exemplified as a detection target.

A series of processes to be described later are stored in the storage unit 13 in the form of, for example, a program (for example, an image detection program). It is realized by executing

First, a plurality of images stored in the storage unit 13 are classified based on date/time information (SA1). At this time, the classification may be based on the position information, or the classification may be based on the date and time information and the position information. Since the details of the classification are as described above, the detailed description is omitted here.

Next, an image recognition process or the like is performed on a group of images classified into the same category, and the degree of similarity between the target person and the person included in each image is obtained (SA2). For example, images of persons (hereinafter referred to as “person images”) are extracted from a group of images belonging to the same category, and the degree of similarity between each person image and the target person is obtained by performing image recognition processing such as image matching. Further, the degree of similarity is stored in the storage unit 13 in association with, for example, the information of the person image or the image including the person image.
For example, as exemplified in FIG. 4, when human images a to g are extracted from a group of images belonging to the same classification, the degree of similarity between each of the human images a to g and the target person is obtained.

Next, a person image whose degree of similarity to the target person is greater than or equal to the first threshold is detected from the person image group belonging to the same classification (SA3).

Next, from among the human image groups classified into the same classification, human images whose similarity to the target person is greater than or equal to the second threshold are detected (SA4). For example, among the human images detected in step SA2, a human image having a degree of similarity to the target person equal to or higher than a second threshold is further detected.

Note that in the following description, a group of person images whose similarity to the target person is greater than or equal to the first threshold and less than the second threshold is referred to as the first person image group, and a person image group whose similarity to the target person is greater than or equal to the second threshold is referred to as the first person image group. This is called a two-person image group.

Next, feature information is acquired from a person image whose degree of similarity to the target person is equal to or greater than the first threshold (SA5). That is, the feature information is acquired from the person images of the first image group and the second image group.
For example, as shown in FIG. 5, human images c to e are detected as human images whose degree of similarity to the target person is greater than or equal to the first threshold and less than the second threshold, and the person whose similarity to the target person is greater than or equal to the second threshold is detected. Assume that human images a and b are detected as images. In this case, feature information is obtained from each of the person images a to e.

Next, the person images in the second person image group are identified as images in which the target person is captured (SA6). Furthermore, among the person images of the first person image group, person images having the same features as those of the person images of the second person image group are identified (SA7).

For example, when extracting the color of clothing of each person image as feature information, as shown in FIG. Assume that d has feature B (eg, blue) and person image e has feature C (eg, red).

In this case, first, in step SA6, the person images a and b, which are the second person image group, are identified. Further, in step SA7, among the person images c to e of the first person image group, the person image c having the same feature as the feature A of the person image a of the second person image group and the person image c having the same feature as the person image b A person image d having characteristics is specified. On the other hand, the human image c is excluded because there is no human image having the same feature C in the second human image group.

Subsequently, an image including the person image specified in steps SA6 and SA7, that is, an image from which the specified person image is extracted is detected as an image of the same person as the target person (SA8). As a result, for example, images from which human images a to d (see FIG. 7) obtained as final results are extracted are detected.

By performing the processing of steps SA2 to SA8 for each classification, an image in which the same person as the target person is captured can be detected from among the plurality of images stored in the storage unit 13. .

Note that the processing procedure of the image detection method described above is an example, and the processing order may be changed or new processing may be added.
For example, in the above processing procedure, images are first classified (SA1), and then human images whose similarity to the target person is equal to or higher than the first threshold are detected for each classification (SA2 to SA4). It is also possible to change the order of these processes. For example, similarity acquisition processing and image detection processing (SA2 to SA4) may be performed first, and then image classification (SA1) may be performed. Further, first, a process of acquiring a degree of similarity with a target person and detecting a person image whose degree of similarity is greater than or equal to a first threshold (SA2, SA3), and then classifying the detected person image (SA1), Further, for each classification, a person image whose degree of similarity to the target person is equal to or greater than the second threshold may be detected (SA4). Also, the processing procedures of steps SA6 and SA7 can be interchanged.

As described above, according to the information processing apparatus 1 and the image detection method according to the present embodiment, a plurality of images are classified based on at least one of date and time information and position information, and images classified into the same classification are classified into the same classification. The image is specified using the feature information included in the captured image. In this way, not only the degree of similarity with the object to be detected, but also the feature information that was not taken into account in acquiring the degree of similarity is used to detect the image of the object to be detected. It is possible to improve the prevention of leakage of extraction while maintaining the same.

More specifically, according to the information processing apparatus 1 and the image detection method according to the present embodiment, for an image having a degree of similarity to the detection object equal to or higher than the second threshold, an object that matches the detection object is An image having the same features as the identified object is identified from among the images that are identified as captured images and whose degree of similarity to the detected object is greater than or equal to the first threshold and less than the second threshold. Then, these specified images are specified as images in which an object that matches the detection object is captured. In this way, not only an image with a degree of similarity equal to or higher than the second threshold, but also an image in which an object that matches the detection object is captured from among images having a degree of similarity equal to or higher than the first threshold, which is lower than the second threshold. is specified, it is possible to reduce the probability of omission of extraction.

[Second embodiment]
Next, an information processing apparatus and an image detection method according to a second embodiment of the present invention will be described with reference to the drawings. In the following description, the same reference numerals are given to the same configurations as in the first embodiment described above, and the description thereof is omitted, and mainly the different points will be described.

FIG. 8 is a functional block diagram showing an extracted image detection function among various functions provided in the information processing apparatus according to the present embodiment. As shown in FIG. 8, the information processing apparatus according to this embodiment includes an image detection unit 20'. The image detection unit 20 ′ includes a storage unit 21 , a first recognition unit 22 , a second recognition unit 23 , a classification unit 24 , a feature information acquisition unit 25 , a detection unit 26 and a re-recognized image extraction unit 27 . Thus, the image detection unit 20' according to the present embodiment is different in functional configuration in that the re-recognized image extraction unit 27 is further provided.

The re-recognized image extraction unit 27 extracts a re-recognized image using the feature information of the image detected by the first recognition unit 22 . For example, the re-recognized image extraction unit 27 classifies a plurality of images detected by the first recognition unit 22 based on the feature information, and extracts re-recognized images based on the number of images belonging to each classification.

The second recognition unit 23 detects an image having a degree of similarity to the detection object that is higher than the first threshold and is equal to or higher than the second threshold, among the images extracted as the re-recognized images.

The detection unit 26 identifies the images detected by the second recognition unit 23 among the re-recognized images, and also identifies the images detected by the first recognition unit 22 that are not extracted as re-recognized images. Then, the specified image is detected as an image of an object that can be regarded as a detection object.

An example of the image detection method according to this embodiment will be described below with reference to the drawings. FIG. 9 is a flow chart showing an example of the procedure of the image detection method according to this embodiment. Also, in the following description, for convenience of description, a person is exemplified as a detection target.

First, a plurality of images stored in the storage unit 21 are classified based on date/time information (SB1). At this time, the classification may be based on the position information, or the classification may be based on the date and time information and the position information. The details of the classification are the same as those of the above-described first embodiment, and detailed description thereof will be omitted here.

Next, image recognition processing or the like is performed on the group of images classified into the same classification to obtain the degree of similarity with the target person (SB2). (SB3). For example, as in the above-described first embodiment, as shown in FIG. 4, in a group of images belonging to the same classification, person images a to g included in each image are extracted, and each person image a to g and the purpose Obtain the degree of similarity with a person. Then, a person image whose degree of similarity to the target person is greater than or equal to the first threshold is detected.

Next, feature information is acquired from a person image whose degree of similarity to the target person is equal to or greater than the first threshold (SB4). For example, as shown in FIG. 10, it is assumed that human images a to e are detected as human images whose degree of similarity to the target person is greater than or equal to the first threshold in step SB3. In this case, feature information is obtained from each of the person images a to e. Here, for convenience of explanation, a case where the color of clothes of each person is extracted as feature information will be exemplified and explained.

As a result, for example, as shown in FIG. 11, the person images a and c have the feature A (for example, black), the person images b and d have the feature B (for example, blue), Assume that the person image e has a feature C (for example, red).

Subsequently, a re-recognized image is extracted using the feature information of the person image (SB5). For example, based on the feature information of each person image whose similarity is greater than or equal to a first threshold, the person images whose similarity is greater than or equal to a first threshold are classified. Then, re-recognized images are extracted based on the number of images belonging to each classification. For example, the number of images belonging to each classification is compared, and an image belonging to a classification having a smaller number of images than other classifications is extracted as a re-recognized image. More specifically, the threshold is determined by performing predetermined statistical processing on the number of human images belonging to each category. Then, re-recognized images are extracted based on the relationship between the number of human images belonging to each category and the threshold. For example, when the average value of the number of images belonging to each category is used as the threshold value, the person images of categories in which the number of person images belonging to each category is equal to or less than the threshold value are extracted as re-recognized images.

For example, in the example shown in FIG. 11, the number of images classified into feature A is "2", the number of images classified into feature B is "2", and the number of images classified into feature C is "1". , the number of human images belonging to feature C is less than or equal to the threshold when the threshold is the average value (about 1.7) of the number of human images belonging to each category. Therefore, in this case, the person image e belonging to the classification of feature C is extracted as the re-recognized image.

Next, an image having a degree of similarity with the target person equal to or higher than the second threshold is detected from the re-recognized images (SB6).

Next, a person image whose similarity is equal to or higher than the second threshold is specified in the re-recognized images (SB7). Identify the image (SB8). As a result, among the person images a to e shown in FIG. 11, the person images a to d that have not been extracted as re-recognized images are identified. Whether or not the person image e extracted as the re-recognized image is identified is determined depending on whether or not the degree of similarity with the target person is greater than or equal to the second threshold.

Then, the image from which the person image specified in steps SB7 and SB8 is extracted is detected as an image of the same person as the target person (SB9).

Then, by performing the processing from steps SB2 to SB9 for each classification, it becomes possible to detect an image in which the same person as the target person is captured from among the plurality of images stored in the storage unit 13. .

Note that the processing procedure of the image detection method described above is an example, and the processing order may be changed or new processing may be added.
For example, in the above processing procedure, images are first classified (SB1), after that, the degree of similarity with the target person is obtained for each classification, and human images whose similarity is equal to or greater than the first threshold are detected ( SB2, SB3), it is also possible to change the order of these processes. For example, steps SB2 and SB3 may be performed first, and then image classification (SB1) may be performed.

As described above, according to the information processing apparatus and the image detection method according to the present embodiment, a plurality of Images are classified, and images in which the same object as the detection target is captured are detected using feature information contained in the images classified into the same classification.

More specifically, images of objects whose similarity to the detected object is equal to or greater than a first threshold are classified based on their feature information, and re-recognized images are extracted based on the number of images belonging to each classification. . Then, among the re-recognized images, images whose similarity is equal to or higher than the second threshold are specified, and among the images whose similarity is equal to or higher than the first threshold, images that are not extracted as re-recognized images are specified. Then, these specified images are detected as images in which an object that matches the detection object is captured.

As a result, it is possible to reduce the probability of occurrence of omission of extraction while maintaining the detection accuracy of an image containing the same object as the detection object.

Although the present invention has been described above using the embodiments, the technical scope of the present invention is not limited to the scope described in the above embodiments. Various changes or improvements can be made to the above-described embodiments without departing from the gist of the invention, and forms with such changes or improvements are also included in the technical scope of the present invention.

1: information processing device 11: CPU
12: main memory 13: storage unit 14: external interface 15: communication interface 16: input unit 17: display unit 18: buses 20, 20': image detection unit 21: storage unit 22: first recognition unit 23: second recognition Unit 24: Classifying unit 25: Feature information acquiring unit 26: Detecting unit 27: Re-recognized image extracting unit

Claims (7)

An information processing device for detecting an image in which a detection object is captured from among a plurality of images captured by an imaging device,
a first recognition means for detecting an image including an object having a degree of similarity equal to or higher than a first threshold with the detected object;
a second recognition means for detecting an image including an object whose degree of similarity to the detection target object is equal to or higher than a second threshold, which is higher than the first threshold;
Classifying means for classifying the plurality of images based on at least one of date and time information indicating the date and time when the image was taken and position information indicating the position where the image was taken;
a feature information obtaining means for obtaining feature information from the group of images classified into the same classification by the classification means;
a detection means for detecting an image of an object that can be regarded as the detection target using the detection result of the first recognition means, the detection result of the second recognition means, and the feature information;
An information processing device comprising: The detection means has the image detected by the first recognition means and the image detected by the second recognition means having the feature information that matches the feature information of the image detected by the second recognition means. 2. The information processing apparatus according to claim 1, wherein and are detected as an image of an object that can be regarded as the detection object. further comprising re-recognized image extraction means for extracting a re-recognized image to be processed by the second recognition means using the characteristic information of the image detected by the first recognition means;
The detection means detects an image detected by the second recognition means in the re-recognized image and an image detected by the first recognition means that is not extracted as the re-recognized image. 2. The information processing apparatus according to claim 1, wherein an object that can be regarded as a detection object is detected as an image captured. 4. The re-recognized image extraction means classifies the plurality of images detected by the first recognition means based on the feature information, and extracts the re-recognized image based on the number of images belonging to each classification. The information processing device according to . The feature information includes the color of at least part of the object, the shape of at least part of the object, the color of an article photographed with the object, and the object photographed with the object. 5. The information processing apparatus according to any one of claims 1 to 4, wherein the information processing apparatus includes any one of the shape of the article that is being held. An image detection method for detecting an image in which a detection object is captured from among a plurality of images captured by an imaging device, comprising:
a first recognition step of detecting an image including an object whose similarity to the detection object is equal to or greater than a first threshold;
a second recognition step of detecting an image containing an object whose similarity to the detection object is equal to or higher than a second threshold higher than the first threshold;
a classification step of classifying the plurality of images based on at least one of date and time information indicating the date and time when the image was taken and position information indicating the position where the image was taken;
a feature information acquisition step of acquiring feature information from the group of images classified into the same category in the classification step;
a detection step of detecting an image in which an object that can be regarded as the detection object is captured using the detection result of the first recognition step, the detection result of the second recognition step, and the feature information;
A computer-implemented image detection method. A program for causing a computer to function as the information processing apparatus according to any one of claims 1 to 5.

Images