Information processing apparatus, information processing method, and program

Abstract

To make it possible to reduce false detection without complicated processing.SOLUTION: An information processing apparatus detects, from an image, a first region including an object to be detected, and detects positions of a plurality of detection target points from the first region. The information processing apparatus determines a second region according to a distribution of the positions of the detection target points detected from the first region, and determines that the first region has been falsely detected when a difference between the first region and the second region is equal to or larger than a predetermined threshold.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to an information processing technique for detecting an area of ​​an object or the like from an image. [Background technology]

[0002] As an image recognition method for detecting the area of ​​a detection target object from an image, a method of performing detection in stages to improve detection accuracy is known. In staged detection, the area of ​​the entire object is first detected from the image, and then a part of the object is further detected within the area of ​​the entire object. However, if the detection result when the area of ​​the entire object is detected from the image is incorrect, the detection results in the subsequent stages will also be incorrect. Therefore, a technology to reduce false detection is required. Patent Document 1 discloses a technology for setting a reference rectangle that contains grouped points and calculating an identification area that contains the reference rectangle based on the horizontal dispersion degree of the points. Patent Document 2 discloses a technology for setting a predetermined part reference point in a specimen image and determining whether to detect the target object based on whether the positions identified as having a part are dispersed. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2014-135039 A [Patent Document 2] JP 2011-210181 A Summary of the Invention [Problem to be solved by the invention]

[0004] However, the technology disclosed in Patent Document 1 cannot determine whether the detected region is a false detection. In addition, the technology disclosed in Patent Document 2 can determine the accuracy of the detected region to a certain extent, but requires a complicated process of sequentially generating regions that include the region reference point and have different sizes.

[0005] Therefore, an object of the present invention is to make it possible to reduce false detections without resorting to complicated processing. [Means for solving the problem]

[0006] The information processing device of the present invention comprises a detection means for detecting a first area containing a detection target object from an image, a position detection means for detecting the positions of multiple detection target points from the first area detected by the detection means, and a judgment means for judging the accuracy of the first area detected by the detection means, wherein the judgment means determines a second area based on the distribution of the positions of the multiple detection target points detected by the position detection means for the first area detected by the detection means, and judges that the first area is a false detection if the difference between the first area and the second area is greater than or equal to a predetermined threshold. Effect of the Invention

[0007] According to the present invention, it is possible to reduce false detections without resorting to complicated processing. [Brief description of the drawings]

[0008] [Figure 1] FIG. 2 is a diagram illustrating a functional configuration of the object detection device. [Diagram 2] FIG. 1 is a diagram used to explain object detection. [Diagram 3] 10 is a flowchart showing the flow of information processing when an object is detected. [Figure 4] FIG. 2 is a diagram illustrating an example of a hardware configuration of an information processing device. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0009] Hereinafter, embodiments of the present invention will be described with reference to the drawings. The following embodiments do not limit the present invention, and not all of the combinations of features described in the present embodiments are necessarily essential to the solution of the present invention. The configurations of the embodiments may be appropriately modified or changed depending on the specifications of the device to which the present invention is applied and various conditions (such as conditions of use and environment of use). In the following embodiments, the same configurations are described with the same reference symbols.

[0010] Fig. 1 is a block diagram showing an example of the functional configuration of an object detection device 101, which is an example of an information processing device according to this embodiment. In this embodiment, an application example of the object detection device 101 is assumed to be a personal computer equipped with a camera (imaging device) not shown in Fig. 1, but is not limited to this. For example, the object detection device 101 may be other devices such as a smartphone or tablet equipped with a camera, a head-mounted display, or a digital camera. The object detection device 101 of this embodiment has, as its main functional units, a data acquisition unit 102, a system control unit 103, a non-volatile memory 104, a system memory 105, an object detection unit 110, and the like.

[0011] The data acquisition unit 102 acquires data such as an image captured by a camera (not shown in FIG. 1) or an image that has been previously captured and processed in some way. The system control unit 103 controls the entire object detection device 101. The functions of the system control unit 103 are realized by at least one processor or circuit, etc. The system control unit 103 realizes each process of the flowchart in Fig. 3, which will be described later, by executing a program stored in the non-volatile memory 104, which will be described later.

[0012] The non-volatile memory 104 is a memory capable of recording data, and stores programs and variables for operating the system control unit 103 . The system memory 105 is composed of, for example, a RAM (random access memory) and stores programs, variables, and the like when the system control unit 103 executes processing. In addition, the system control unit 103 stores in the system memory 105 the images acquired by the data acquisition unit 102 and the processing results by the object detection unit 110 (described later).

[0013] The object detection unit 110 receives the image acquired by the data acquisition unit 102 via the system control unit 103, and performs processing to detect the area of ​​a predetermined detection target object from the image. In this embodiment, a human hand is taken as an example of the predetermined detection target object. Therefore, the object detection unit 110 detects the area of ​​the human hand from the image. The object detection unit 110 is configured to include a detection unit 111, a position detection unit 112, and a determination unit 113.

[0014] The detection unit 111 detects a minimum area (hereinafter referred to as a first area) of a predetermined shape that is estimated to include a specific detection target object by applying an object area detection process to an image acquired by the data acquisition unit 102 and input via the system control unit 103. In this embodiment, a square is given as the predetermined shape. In the object area detection process in the detection unit 111, only one first area estimated to be a detection target object may be detected, multiple first areas may be detected, or none may be detected, depending on how a subject or the like in the input image is depicted. In this embodiment, a human hand is assumed to be the detection target object, and the detection unit 111 detects a first area estimated to include a human hand from the image. In addition, the object area detection process in the detection unit 111 may be realized, for example, by using a trained deep learning model or a rule-based algorithm.

[0015] The position detection unit 112 detects the positions of a plurality of detection target points (hereinafter, referred to as targets) in the first region detected by the detection unit 111. When the detection unit 111 detects a plurality of first regions, the position detection unit 112 detects the positions of a plurality of detection target points (targets) in each of the plurality of first regions. When a human hand is the detection target object as in this embodiment, the target positions are each joint point position including the fingertips of each finger of the human hand and the joint point position of the wrist. That is, the position detection unit 112 detects each joint point position including the fingertips of the fingers and the joint point position of the wrist as a plurality of target positions (positions of detection target points) from within the first region estimated by the detection unit 111 as the region of the human hand. The detection target point position detection process (target position detection process) in the position detection unit 112 may be realized, for example, by using a learned deep learning model or may be realized by a rule-based algorithm.

[0016] The determination unit 113 determines a second region corresponding to the distribution of the multiple target positions detected by the position detection unit 112 for the first region detected by the detection unit 111. In this embodiment, a region of the smallest predetermined shape that contains the multiple target positions is determined as the second region corresponding to the distribution of the multiple target positions. In this embodiment, the predetermined shape of the second region is also a square like the first region. Note that when the detection unit 111 detects multiple first regions and the position detection unit 112 detects multiple target positions for each of the first regions, the determination unit 113 determines a second region for each of the first regions.

[0017] Then, the determination unit 113 determines whether the first area is correct based on the difference between the first area and the second area determined corresponding to the first area. In this embodiment, the correctness of the first area refers to whether the detection result by the object area detection process in the detection unit 111 is correct or not. That is, in the case of the example of this embodiment, it refers to whether the first area detected by the detection unit 111 is an area of ​​a human hand or not. When the difference between the first area and the second area is equal to or greater than a predetermined threshold, the determination unit 113 determines that the detection result by the object area detection process in the detection unit 111 is incorrect, that is, determines that the first area is an area detected in error.

[0018] FIG. 2 is a diagram used to explain object region detection processing by detection section 111 of object detection section 110, detection target point position detection processing (target position detection processing) by position detection section 112, and true / false determination processing by determination section 113. In FIG.

[0019] An image 200 is an example of an image acquired by the data acquisition unit 102 and input to the object detection unit 110 via the system control unit 103. Detection unit 111 of object detection unit 110 detects a first region including human hand 201 from image 200. In this embodiment, detection unit 111 detects the smallest square region that contains an image region estimated to be human hand 201 as a hand region (first region). In the example of FIG. 2, it is assumed that two image regions estimated to be human hand 201 from image 200, region 211 and region 221, are detected as first regions by the object region detection process in detection unit 111.

[0020] The position detection unit 112 detects multiple target positions for each first region estimated by the detection unit 111 to be a hand region. In the present embodiment, the position detection unit 112 detects each joint point position including the fingertip of each finger of the hand and the joint point position of the wrist as multiple target positions, as described above. In the example of FIG. 2, the position detection unit 112 detects multiple target positions 212 indicated by black circles in the figure from the first region 211, and also detects multiple target positions 222 indicated by black circles in the figure from the first region 221.

[0021] The determination unit 113 determines, as a second area, the smallest square area that contains the multiple target positions detected by the position detection unit 112 within the first area detected by the detection unit 111. Then, the determination unit 113 determines whether the first area is correct or not based on a comparison between the difference between the first area and the second area and a predetermined threshold value.

[0022] In the example of Fig. 2, the determination unit 113 determines the smallest area that includes the multiple target positions 212 detected by the position detection unit 112 with respect to the first area 211 detected by the detection unit 111 as the second area 213. Then, the determination unit 113 determines whether the first area 211 is correct or not based on a comparison between the difference between the first area 211 and the second area 213 and a predetermined threshold. That is, when the difference between the first area 211 and the second area 213 is equal to or greater than the predetermined threshold, the determination unit 113 determines that the first area 211 detected by the detection unit 111 as the hand area is a false detection.

[0023] Similarly, the determination unit 113 determines the smallest area that includes the multiple target positions 222 detected by the position detection unit 112 with respect to the first area 221 detected by the detection unit 111 as the second area 223. Then, the determination unit 113 determines whether the first area 221 is correct or not based on a comparison between the difference between the first area 221 and the second area 223 and a predetermined threshold. That is, when the difference between the first area 221 and the second area 223 is equal to or greater than the predetermined threshold, the determination unit 113 determines that the first area 221 detected by the detection unit 111 as the hand area is a false detection.

[0024] In this embodiment, the determination unit 113 calculates the width ratio, height ratio, and area ratio of the first region and the second region as the difference between the first region and the second region, and compares them with the thresholds set for the width ratio, height ratio, and area ratio. That is, when the width ratio, height ratio, and area ratio calculated as the difference between the first region and the second region are equal to or greater than the corresponding thresholds, the determination unit 113 determines that the first region detected by the detection unit 111 as a hand region is a false detection.

[0025] For example, assume that the vertical and horizontal widths of the first region 211 are each 400 pixels, and the vertical and horizontal widths of the second region 213 are each 360 pixels. In this case, the width ratio of the first region 211 to the second region 213 is 400 / 360=1.11, the vertical ratio is 400 / 360=1.11, and the area ratio is 160000 / 129600=1.23. For example, suppose that the vertical and horizontal widths of the first region 221 are each 200 pixels, and the vertical and horizontal widths of the second region 223 are each 60 pixels. In this case, the width ratio of the first region 221 to the second region 223 is 200 / 60=3.33, the vertical ratio is 200 / 60=3.33, and the area ratio is 40000 / 3600=11.11.

[0026] The predetermined threshold is set in advance based on the distribution of the joint points of the fingers and wrist in the smallest square area that contains the hand area. In general, the joint points of the fingers and wrist are not concentrated in a part of the smallest square area that contains the hand area, but are distributed throughout the square area. In general, the height ratio and width ratio of the joint points of the fingers and wrist are about 1.2, and the area ratio is about 1.44. For this reason, in the present embodiment, the threshold for the width ratio and height ratio is set to 1.2, and the threshold for the area ratio is set to 1.44 as the predetermined thresholds to be compared with the difference between the first area and the second area.

[0027] In the example of image 200 in Fig. 2 described above, the difference between first region 211 and second region 213 is, as described above, 400 / 360 = 1.11 in width ratio, 400 / 360 = 1.11 in height ratio, and 160000 / 129600 = 1.23 in area ratio. Therefore, determination unit 113 determines that the width ratio, height ratio, and area ratio calculated as the difference between first region 211 and second region 213 are all less than a predetermined threshold. Therefore, determination unit 113 determines that first region 211 detected by detection unit 111 is a correct detection result as a region including a human hand.

[0028] In contrast, the difference between the first region 221 and the second region 223 is, as described above, a width ratio of 200 / 60=3.33, a height ratio of 200 / 60=3.33, and an area ratio of 40000 / 3600=11.11. In this case, the determination unit 113 determines that the width ratio, height ratio, and area ratio calculated as the difference between the first region 221 and the second region 223 are all equal to or greater than a predetermined threshold. Therefore, the determination unit 113 determines that the first region 221 detected by the detection unit 111 is an erroneous detection result as a region including a human hand.

[0029] In the present embodiment, the width ratio, height ratio, and area ratio are calculated as the difference between the first region and the second region, and if any one of them is equal to or greater than a predetermined threshold, it is determined that the detection is erroneous, but the present invention is not limited to this. For example, the determination unit 113 may calculate one or two of the width ratio, height ratio, and area ratio as the difference between the first region and the second region, and if any one of them is equal to or greater than a predetermined threshold, it may determine that the detection is erroneous.

[0030] In this embodiment, the predetermined shape of the first region and the second region is a square, but the predetermined shape of the first region and the second region is not necessarily limited to a square, and may be, for example, a rectangle, a polygon, a circle, an ellipse, etc., as long as the first region and the second region have the same shape. Even in the case of a shape other than a square, at least one of the width ratio (horizontal region width), the height ratio (vertical region width), and the area ratio may be calculated as the difference between the first region and the second region, and the result may be compared with a predetermined threshold value set corresponding to the calculated difference to determine whether the result is correct or not.

[0031] FIG. 3 is a flowchart showing the flow of processing operations in the object detection device 101 of this embodiment described above. First, in the process of step S301, the system control unit 103 acquires an image to be subjected to object detection from the data acquisition unit 102. In the case of the example of this embodiment, the data acquisition unit 102 acquires, for example, an image 200 as shown in FIG.

[0032] Next, in the process of step S302, detection unit 111 of object detection unit 110 performs object region detection processing to detect, as a first region, a region including a detection target object, for the image acquired by data acquisition unit 102 and input from system control unit 103. In the case of this embodiment, detection unit 111 detects a region including a human hand from image 200.

[0033] Next, in the process of step S303, the detection unit 111 judges whether or not one or more first regions have been detected in the object region detection process of step S302. If no first region has been detected, the process of the object detection unit 110 proceeds to step S311. On the other hand, if one or more first regions have been detected, the process of the object detection unit 110 proceeds to the loop process of steps S304 to S310. In the case of the example of the image 200 in FIG. 2 described above, since two first regions, region 211 and region 221, were detected in step 302, the process of the object detection unit 110 proceeds to the loop process of steps S304 to S310. The loop process of steps S304 to S310 is performed for each first region detected by the detection unit 111. That is, in the case of the example of FIG. 2, the object detection unit 110 performs the process of steps S304 to S310 for the first region 211 and the process of steps S304 to S310 for the first region 221.

[0034] Here, in the loop processing of steps S304 to S310, it is explained that, of the two first regions shown in FIG. 2, processing is first performed on first region 211, and then processing is performed on first region 221. When the process proceeds to the loop process of steps S304 to S310, first, in the process of step S305, the position detection unit 112 detects multiple target positions in the first area 211 detected by the detection unit 111. When the first area 211 in Fig. 2 is the processing target, the position detection unit 112 detects multiple target positions 212, which are the positions of each joint point of the hand, from this first area 211.

[0035] Next, in the process of step S306, the determination unit 113 determines the smallest area that includes the multiple target positions detected by the position detection unit 112 relative to the first area 211 detected by the detection unit 111 as the second area 213. Next, in the process of step S307, the determination unit 113 calculates the difference between the first region 211 and the second region 213. In the case of this embodiment, as described above, the determination unit 113 calculates the width ratio, height ratio, and area ratio as the differences between the first region 211 and the second region 213.

[0036] Next, in the process of step S308, the determination unit 113 determines whether the difference between the first region 211 and the second region 213 obtained in step S307 is equal to or greater than a predetermined threshold. In the case of this embodiment, as described above, it is determined whether the width ratio, height ratio, and area ratio calculated as the difference between the first region 211 and the second region 213 are equal to or greater than the corresponding thresholds. Then, if any one of the width ratio, height ratio, and area ratio between the first region 211 and the second region 213 is equal to or greater than the threshold, the determination unit 113 determines that the first region 211 detected by the detection unit 111 as a hand region is a false detection. In the case of the example of FIG. 2, since the difference between the first region 211 and the second region 213 obtained in step S307 is less than the predetermined threshold as described above, the determination unit 113 determines that the first region 211 is a region including a human hand (a correctly detected object region). Thereafter, the loop processing of object detection unit 110 proceeds to processing of second region 221, which is the next object region.

[0037] 2 is the target of the loop processing of steps S304 to S310, in step S305, position detection unit 112 detects multiple target positions for first region 221. When first region 221 in Fig. 2 is the processing target, position detection unit 112 detects multiple target positions 222, which are the positions of each joint point of the hand, from this first region 221.

[0038] Next, in the process of step S306, the determination unit 113 determines the smallest area that includes the multiple target positions detected by the position detection unit 112 relative to the first area 221 as the second area 223. Next, in the process of step S307, the determination unit 113 calculates the width ratio, height ratio, and area ratio as the differences between the first region 221 and the second region 223, respectively.

[0039] Next, in the process of step S308, the determination unit 113 determines whether the difference between the first region 221 and the second region 223 obtained in step S307 is equal to or greater than a predetermined threshold. In the case of this embodiment, as described above, the width ratio, height ratio, and area ratio between the first region 221 and the second region 223 are all equal to or greater than the threshold, so the determination unit 113 determines that the first region 221 detected as a hand region by the detection unit 111 is a false detection.

[0040] Thereafter, the loop process of object detection unit 110 proceeds to processing for the next object region, but in the example of Fig. 2, there are no first regions detected as object regions other than the two first regions described above, so object detection unit 110 exits the loop process of steps S304 to S310.

[0041] Then, after exiting the loop process of steps S304 to S310, system control unit 103 determines in step S311 whether or not to end the process of the flowchart in Fig. 3. For example, if the user inputs an end instruction via an operation device (not shown) or if data acquisition unit 102 fails to acquire an image, system control unit 103 determines to end the process of Fig. 3. On the other hand, if it is determined not to end in step S311, system control unit 103 returns the process to step S301.

[0042] As described above, in this embodiment, the object detection unit 110 detects a first region including a detection target object from an image, and further detects multiple target positions within the first region. Then, the object detection unit 110 determines a second region according to the distribution of the multiple target positions detected within the first region, and judges whether the first region is correct or not based on the difference between the first region and the second region. As a result, according to this embodiment, it is possible to reduce erroneous detection of an object region including a detection target object.

[0043] <Hardware configuration example> FIG. 4 is a diagram showing an example of the hardware configuration of an information processing device such as a personal computer, which is an application example of the object detection device 101 according to the embodiment. In FIG. 4, a CPU 402 comprehensively controls each device connected via a bus 401. The CPU 402 reads out and executes a program stored in a ROM (read-only memory) 403. An operating system (OS), each information processing program according to this embodiment, a device driver, and the like are stored in a ROM 403 included in a non-volatile memory 104 in FIG. 1. These programs are read out from the ROM 403, expanded in a RAM (random access memory) 404 included in a system memory 105 in FIG. 1, and appropriately executed by the CPU 402. A camera 407 is an imaging device that captures a subject such as an object. An input I / F 405 inputs a signal from an external device in a format that can be processed by the information processing device of this embodiment. An output I / F 406 outputs a signal in a format that can be processed by an external device.

[0044] The CPU 402 loads the information processing program according to this embodiment from the ROM 403 into the RAM 404 and executes it, thereby implementing the processing of each functional unit such as the system control unit 103, the data acquisition unit 102, and the object detection unit 110 in FIG. 1 and the flowchart in FIG. 3. When configuring hardware as an alternative to software processing using the CPU 402, it is sufficient to configure a calculation unit or circuit corresponding to each functional unit shown in FIG. 1. That is, each functional unit in the information processing device of this embodiment may be entirely implemented by hardware, or may be partially executed by software. When executed by software, the software is executed using the CPU 402, ROM 403, RAM 404, etc. in FIG. 4.

[0045] The present invention can also be realized by supplying a program for implementing one or more of the functions of the above-described embodiments to a system or device via a network or a storage medium, and having one or more processors in the computer of the system or device read and execute the program. It can also be realized by a circuit (e.g., ASIC) that implements one or more of the functions. The above-mentioned embodiments are merely examples of the implementation of the present invention, and the technical scope of the present invention should not be interpreted as being limited by these. In other words, the present invention can be implemented in various forms without departing from its technical concept or main characteristics.

[0046] The disclosure of this embodiment includes the following configuration, method, and program. (Configuration 1) A detection means for detecting a first area including a detection target object from an image; a position detection means for detecting positions of a plurality of detection target points from the first area detected by the detection means; a determination means for determining whether the first area detected by the detection means is correct; having The information processing device is characterized in that the judgment means determines a second area corresponding to the distribution of positions of multiple detection target points detected by the position detection means for a first area detected by the detection means, and determines that the first area is a false detection if the difference between the first area and the second area is greater than or equal to a predetermined threshold. (Configuration 2) When the detection means detects a plurality of the first regions, the position detection means detects positions of the plurality of detection target points for each of the first regions; 2. The information processing apparatus according to configuration 1, wherein the determination means determines the second area for each of the first areas and determines whether each of the first areas is correct or not. (Configuration 3) The information processing device according to configuration 1 or 2, characterized in that the determination means determines, as the second area, the smallest area that contains the positions of the multiple detection target points detected by the position detection means with respect to the first area. (Configuration 4) The detection means detects, as the first region, a minimum region of a predetermined shape that includes the detection target object, 4. The information processing apparatus according to configuration 3, wherein the determination means determines, as the second region, a region of a minimum predetermined shape that includes the positions of the plurality of detection target points. (Configuration 5) 5. The information processing device according to configuration 4, wherein the predetermined shape is a square. (Configuration 6) The information processing device described in any one of configurations 1 to 5, characterized in that the determination means calculates a width ratio, a height ratio, and an area ratio between the first region and the second region as the difference between the first region and the second region, and determines that the first region is a false detection if any of the width ratio, the height ratio, and the area ratio is equal to or greater than a predetermined threshold value corresponding to each of the first region, the height ratio, and the area ratio. (Configuration 7) The information processing device described in any one of configurations 1 to 5, characterized in that the determination means calculates at least one of the values ​​of the width ratio, height ratio, and area ratio between the first region and the second region as the difference between the first region and the second region, and if the calculated value is equal to or greater than a predetermined threshold value set in advance, determines that the first region is a false detection. (Configuration 8) 8. The information processing device according to any one of configurations 1 to 7, wherein the predetermined threshold value is a value set in accordance with a distribution of possible positions of the plurality of detection target points. (Configuration 9) the detection means detects a region of a human hand as the first region including the detection target object, 9. The information processing device according to any one of configurations 1 to 8, wherein the position detection means detects the positions of each joint point of the fingers and wrist of the person's hand as the positions of the multiple detection target points. (Method 1) a detection step of detecting a first region including a detection target object from the image; a position detection step of detecting positions of a plurality of detection target points from the first area detected in the detection step; a determination step of determining whether the first region detected in the detection step is correct or not; having The information processing method is characterized in that the judgment process determines a second area corresponding to the distribution of positions of multiple detection target points detected by the position detection process for a first area detected by the detection process, and judges that the first area is a false detection if the difference between the first area and the second area is greater than or equal to a predetermined threshold. (Program 1) A program for causing a computer to function as the information processing device according to any one of configurations 1 to 9. [Explanation of symbols]

[0047] 101: object detection device, 102: data acquisition unit, 103: system control unit, 104: non-volatile memory, 105: system memory, 110: object detection unit, 111: detection unit, 112: position detection unit, 113: determination unit

Claims

1. A detection means for detecting a first region containing the object to be detected from an image, A position detection means for detecting the positions of multiple target points from the first region detected by the detection means, A determination means for determining whether the first region detected by the detection means is correct or incorrect, It has, The information processing apparatus is characterized in that the determination means determines a second region corresponding to the distribution of positions of a plurality of detection target points detected by the position detection means with respect to a first region detected by the detection means, and determines that the first region is a false detection if the difference between the first region and the second region is greater than or equal to a predetermined threshold.

2. When the detection means detects a plurality of the first regions, The position detection means detects the positions of the plurality of detection target points for each of the first regions, The information processing apparatus according to claim 1, characterized in that the determination means determines the second region for each of the first regions and determines correctness for each of the first regions.

3. The information processing apparatus according to claim 1 or 2, characterized in that the determination means determines the smallest region that encompasses the positions of the plurality of detection target points detected by the position detection means in relation to the first region as the second region.

4. The detection means detects the smallest area of ​​a predetermined shape that includes the object to be detected as the first region, The information processing apparatus according to claim 3, characterized in that the determination means determines the smallest predetermined shape of region that encompasses the positions of the plurality of detection target points as the second region.

5. The information processing apparatus according to claim 4, characterized in that the predetermined shape is a square.

6. The determination means calculates the width ratio obtained by dividing the width of the first region by the width of the second region, the height ratio obtained by dividing the height of the first region by the height of the second region, and the area ratio obtained by dividing the area of ​​the first region by the area of ​​the second region as the difference between the first region and the second region, and determines that the first region is a false detection if any of the width ratio, height ratio, or area ratio is greater than or equal to a predetermined threshold set in advance corresponding to each of them. This is the information processing apparatus according to claim 1.

7. The determination means calculates at least one of the following values ​​as the difference between the first region and the second region: the width ratio obtained by dividing the width of the first region by the width of the second region, the height ratio obtained by dividing the height of the first region by the height of the second region, and the area ratio obtained by dividing the area of ​​the first region by the area of ​​the second region, and if the calculated value is greater than or equal to a predetermined threshold set in advance, it determines that the first region is a false detection, as described in claim 1.

8. The information processing apparatus according to claim 1, characterized in that the predetermined threshold is a value set according to the distribution in which the positions of the plurality of detection target points can be.

9. The detection means detects the area of ​​a human hand as the first area which includes the object to be detected, The information processing device according to claim 1, characterized in that the position detection means detects the positions of the joint points of the fingers and wrist of the person's hand as the positions of the plurality of detection target points.

10. A detection step involves detecting a first region containing the object to be detected from the image, A position detection step for detecting the positions of multiple target points from the first region detected by the detection step, A determination step that determines whether the first region detected by the detection step is correct or incorrect, It has, The determination step is characterized by determining a second region corresponding to the distribution of positions of a plurality of detection target points detected by the position detection step with respect to the first region detected by the detection step, and determining that the first region is a false detection if the difference between the first region and the second region is greater than or equal to a predetermined threshold.

11. Computers, A detection means for detecting a first region containing the object to be detected from an image, A position detection means for detecting the positions of multiple target points from the first region detected by the detection means, A determination means for determining whether the first region detected by the detection means is correct or incorrect, It has, The determination means is a program that causes the determination means to function as an information processing device that determines a second region corresponding to the distribution of positions of a plurality of detection target points detected by the position detection means relative to a first region detected by the detection means, and determines that the first region is a false detection if the difference between the first region and the second region is greater than or equal to a predetermined threshold.

Images