JP6998922B2 - Notification judgment device, notification judgment method and notification judgment program - Google Patents

Description

The present invention relates to a technique for determining whether or not to notify that an object has been detected by a detection process.

There is a technique of detecting an object from the image data obtained by photographing the imaged area by the photographing device and notifying the administrator or the like when the object is detected. Patent Document 1 describes that an object is detected by comparing the image data obtained by photographing the imaged area by the photographing apparatus with the background image data.

JP-A-2018-133042

When an object is detected from image data, it may be difficult to prevent erroneous detection only by the detection processing logic due to factors such as sunshine change and shadow. If the administrator or the like is notified when an erroneous detection is made, unnecessary confirmation work or the like will occur.
As described in Patent Document 1, when the object is detected by comparison with the background image data, it is conceivable to prepare the background image data for each factor. However, it is necessary to acquire the background image data at a timing when people or the like are not reflected. Therefore, in reality, the background image data can be acquired only at a specific timing such as at night, and it is difficult to prepare the background image data for each factor.
It is an object of the present invention to reduce the notification in case of false detection.

The notification determination device according to the present invention is
The target area in the target image data in which the object is detected, and the false detection image data in which the false detection object determined to be false notification in the past is detected. The overlap calculation unit that calculates the overlap rate with the false detection area where there is a false positive object,
Similarity calculation unit for calculating the similarity between the target partial data which is the partial image data of the target area in the target image data and the false positive partial data which is the partial image data of the false detection area in the false detection image data. When,
Based on the duplication rate calculated by the duplication calculation unit and the similarity calculated by the similarity calculation unit, a determination unit for determining whether or not to notify that the object has been detected is provided. Be prepared.

The determination unit determines that the notification is not given when the overlap rate is equal to or higher than the reference rate and the similarity is equal to or higher than the reference degree.

The similarity calculation unit calculates the similarity between the target partial data and the false positive partial data when the overlap rate is equal to or higher than the reference rate.

The duplication calculation unit sets each of the plurality of false positive objects as the target false positive objects, and includes the target area and the false positive area where the false positive objects are present in the false positive image data in which the false positive objects of the target are detected. Calculate the duplication rate of
The similarity calculation unit includes false detection partial data and target partial data in false positive image data in which the overlap rate is equal to or higher than the reference rate among false positive image data in which each of the plurality of false positive objects is detected. Calculate the similarity.

The target image data is image data obtained by shooting a shooting area with a shooting device.
The notification determination device further
Detection to detect the object from the target image data by comparing the target image data with the background image data obtained by photographing the imaged area in a state where the object is not present in the imaged area. It has a part.

The notification determination method according to the present invention is
The duplicate calculation unit is erroneous image data in which the target area in the target image data, which is the image data in which the object is detected, and the false detection object determined to be erroneous notification in the past are detected. The overlap rate of the false positive object in the detected image data with the false positive area is calculated, and the overlap rate is calculated.
The similarity calculation unit determines the degree of similarity between the target partial data, which is the partial image data of the target region in the target image data, and the false positive partial data, which is the partial image data of the false detection region in the false detection image data. Calculate and
The determination unit determines whether or not to notify that the object has been detected based on the overlap rate and the similarity.

The notification determination program according to the present invention is
The target area in the target image data in which the object is detected, and the false detection image data in which the false detection object determined to be false notification in the past is detected. Duplicate calculation processing that calculates the overlap rate with the false detection area where there is a false positive object,
Similarity calculation process for calculating the similarity between the target partial data which is the partial image data of the target area in the target image data and the false positive partial data which is the partial image data of the false detection area in the false detection image data. When,
Based on the duplication rate calculated by the duplication calculation process and the similarity calculated by the similarity calculation process, a determination process for determining whether or not to notify that the object has been detected is performed. Make the computer function as a notification judgment device.

In the present invention, it is determined whether or not to notify that the object has been detected by using the false detection image data which is the image data in which the false positive object determined to be false notification in the past is detected. As a result, it is difficult to prevent erroneous detection only by the logic of the detection process, and it is possible to prevent notification in cases where erroneous detection occurs many times.

The block diagram of the notification determination apparatus 10 which concerns on Embodiment 1. FIG. The flowchart which shows the process flow of the notification determination apparatus 10 which concerns on Embodiment 1. The figure which shows the data stored in the false positive data storage unit 31 which concerns on Embodiment 1. FIG. An explanatory diagram of a specific example of processing of the notification determination device 10 according to the first embodiment. An explanatory diagram of a specific example of processing of the notification determination device 10 according to the first embodiment. The flowchart of the process of detecting an object from the object image data in step S12 which concerns on Embodiment 1. FIG. The block diagram of the notification determination apparatus 10 which concerns on modification 3. FIG.

Embodiment 1.
*** Explanation of configuration ***
The configuration of the notification determination device 10 according to the first embodiment will be described with reference to FIG.
The notification determination device 10 is a computer.
The notification determination device 10 includes hardware for a processor 11, a memory 12, a storage 13, and a communication interface 14. The processor 11 is connected to other hardware via a signal line and controls these other hardware.

The processor 11 is an IC (Integrated Circuit) that performs processing. Specific examples of the processor 11 are a CPU (Central Processing Unit), a DSP (Digital Signal Processor), and a GPU (Graphics Processing Unit).

The memory 12 is a storage device that temporarily stores data. As a specific example, the memory 12 is a SRAM (Static Random Access Memory) or a DRAM (Dynamic Random Access Memory).

The storage 13 is a storage device for storing data. As a specific example, the storage 13 is an HDD (Hard Disk Drive). The storage 13 includes SD (registered trademark, Secure Digital) memory card, CF (CompactFlash, registered trademark), NAND flash, flexible disk, optical disk, compact disk, Blu-ray (registered trademark) disk, DVD (Digital Versaille Disk), and the like. It may be a portable recording medium.

The communication interface 14 is an interface for communicating with an external device. As a specific example, the communication interface 14 is a port of Ethernet (registered trademark), USB (Universal Serial Bus), HDMI (registered trademark, High-Definition Multimedia Interface).

The notification determination device 10 is connected to a photographing device 41 such as a surveillance camera via a communication interface 14.

The notification determination device 10 includes an image acquisition unit 21, a detection unit 22, an overlap calculation unit 23, a similarity calculation unit 24, a determination unit 25, and a notification unit 26 as functional components. The functions of each functional component of the notification determination device 10 are realized by software.
The storage 13 stores a program that realizes the functions of each functional component of the notification determination device 10. This program is read into the memory 12 by the processor 11 and executed by the processor 11. As a result, the functions of each functional component of the notification determination device 10 are realized.

Further, the storage 13 realizes a false positive data storage unit 31.

In FIG. 1, only one processor 11 is shown. However, the number of processors 11 may be plural, and the plurality of processors 11 may execute programs that realize each function in cooperation with each other.

*** Explanation of operation ***
The operation of the notification determination device 10 according to the first embodiment will be described with reference to FIGS. 2 to 6.
The operation procedure of the notification determination device 10 according to the first embodiment corresponds to the notification determination method according to the first embodiment. Further, the program that realizes the operation of the notification determination device 10 according to the first embodiment corresponds to the notification determination program according to the first embodiment.

With reference to FIG. 2, the processing flow of the notification determination device 10 according to the first embodiment will be described.
As a premise of the processing shown in FIG. 2, as shown in FIG. 3, false detection image data which is image data in which a false positive object determined to be false notification in the past is detected, and false detection in false positive image data. The coordinates of the false positive area where the object is located are stored in the false positive data storage unit 31 as false positive data.

(Step S11: Image acquisition process)
The image acquisition unit 21 acquires the image data of the latest frame obtained by photographing the photographing area by the photographing apparatus 41 as the target image data via the communication interface 14. The image acquisition unit 21 writes the target image data to the memory 12.

(Step S12: Detection process)
The detection unit 22 reads the target image data acquired in step S11 from the memory 12. The detection unit 22 detects an object from the target image data. Specific examples of the objects are a leftover object that has been left for a certain period of time or longer, and a resident person who has stayed for a certain period of time or longer. When the object is detected, the detection unit 22 specifies the coordinates of the target area where the object is located in the target image data.
As a process of detecting an object from the target image data, it is possible to use a process of taking a difference between the target image data and the background image data and determining whether or not a similar difference continues to appear at the same position for a certain period of time or longer. can.

(Step S13: Detection determination process)
When an object is detected in step S12, the detection unit 22 writes each detected object into the memory 12 as a notification candidate, and then proceeds to the process in step S14. On the other hand, if the object is not detected in step S12, the detection unit 22 returns the process to step S11.

The processing of steps S14 to S19 is executed with each object detected in step S13 as the processing target. Further, the processes of steps S14 to S18 are executed for each erroneous notification data stored in the erroneous detection data storage unit 31 as a processing target.

(Step S14: Duplicate calculation process)
The duplication calculation unit 23 calculates the overlap rate between the target area having the object to be processed in the target image data and the false detection area having the false detection object in the false notification image data included in the false notification data of the processing target. ..
Specifically, the duplication calculation unit 23 reads out the coordinates of the erroneous detection area included in the erroneous notification data to be processed from the erroneous detection data storage unit 31. The overlap calculation unit 23 calculates the overlap rate between the target area and the false detection area from the coordinates of the target area specified in step S12 and the coordinates of the false detection area read from the false detection data storage unit 31. .. Here, the duplication calculation unit 23 calculates the ratio of the false detection area overlapping with the target area as the duplication rate.

(Step S15: Duplicate determination process)
If the overlap rate calculated in step S14 is equal to or greater than the reference rate, the determination unit 25 advances the process to step S16. On the other hand, when the duplication rate calculated in step S14 is less than the reference rate, the determination unit 25 ends the processing for the erroneous notification data to be processed.

(Step S16: Similarity calculation process)
The similarity calculation unit 24 is a portion of the target partial data which is the partial image data of the target area where the object to be processed in the target image data is present, and the portion of the false detection area in the false detection image data included in the false notification data of the processing target. The degree of similarity with the false positive partial data, which is image data, is calculated.
Specifically, the similarity calculation unit 24 reads out the erroneous detection image data included in the erroneous notification data to be processed and the coordinates of the erroneous detection area from the erroneous detection data storage unit 31. The similarity calculation unit 24 cuts out the partial image data of the region indicated by the coordinates of the false detection region as the false detection partial data from the false detection image data. Then, the similarity calculation unit 24 calculates the similarity between the target partial data and the false positive partial data.

At this time, for example, the similarity calculation unit 24 calculates the similarity by calculating the edge difference between the target partial data and the false positive partial data. That is, the similarity calculation unit 24 performs edge detection processing on each of the target portion data and the false detection portion data to extract edge features. The existing method may be used for the edge detection process. Then, the similarity calculation unit 24 identifies pixels having no difference between the image data representing the edge features extracted from the target partial data and the image data representing the edge features extracted from the false positive partial data. Then, the similarity calculation unit 24 sets the ratio of pixels having no difference to the number of pixels of the target partial data as the similarity.
The similarity calculation unit 24 performs expansion processing, smoothing processing, opening processing, and closing processing on the image data representing the edge characteristics, and then determines whether or not there is a difference for each pixel. You may. The expansion process, the opening process, and the closing process are morphology conversion processes. The smoothing process is a process for smoothing the luminance value of an image, and is a process executed by using, for example, a moving average filter or a Gaussian filter.

(Step S17: Similarity determination process)
If the similarity calculated in step S16 is equal to or higher than the reference degree, the determination unit 25 advances the process to step S18. On the other hand, when the similarity calculated in step S16 is less than the reference degree, the determination unit 25 ends the processing for the erroneous notification data to be processed.

(Step S18: Candidate deletion process)
The notification unit 26 excludes the object to be processed from the notification candidates.
That is, the determination unit 25 determines that the notification is not given when the overlap rate is equal to or higher than the reference rate and the similarity is equal to or higher than the reference degree. Then, if it is determined not to notify, the notification unit 26 excludes the object to be processed from the notification candidates. As a result, the object to be processed is not notified. If there is an object that has not been processed, processing for the next object is executed.

(Step S19: Notification processing)
The notification unit 26 notifies the object of the notification candidate.
That is, the determination unit 25 compares the object of the notification candidate with all the erroneous notification data stored in the erroneous detection data storage unit 31, and the degree of similarity is the degree of reference when the duplication rate is less than the reference rate. In at least one of the cases where it is less than, it is determined that the notification is made. If there is an object that has not been processed, processing for the next object is executed.

When the notification is given in step S19, the administrator or the like determines whether or not the notification is incorrect. An erroneous notification means that the detected object is not an object but an erroneous object.
When it is determined that the notification is erroneous, the notification unit 26 uses the target image data acquired in step S11 as false detection image data, and the coordinates of the area where the target object is detected in step S12 are false detection objects. It is written in the false positive data storage unit 31 as the coordinates of the false positive region.

A specific example of the processing of the notification determination device 10 according to the first embodiment will be described with reference to FIGS. 4 and 5.
It is assumed that the erroneous detection data storage unit 31 stores erroneous notification data including the erroneous detection image data 51 and the coordinates of the erroneous detection area 52. That is, the shadow 72 formed at the root of the pillar 71 is erroneously notified as an object, the image data at that time is the false detection image data 51, and the region of the shadow 72 formed at the root of the pillar 71 is the false detection region. 52.

As shown in FIG. 4, it is assumed that the target image data 61X is acquired in step S11, the difference between the target image data 61X and the background image data 62 is calculated in step S12, and the target object 63X is specified. Then, since the object 63X is specified in step S13, the process proceeds to step S14. In steps S14 to S15, the target area where the object 63X is located and the false detection area 52 are substantially at the same position, and the overlap rate is equal to or higher than the reference rate, so that the process proceeds to step S16. In steps S16 to S17, the target partial data which is the partial image data of the target region in the target image data 61X and the false positive partial data which is the partial image data of the false positive region 52 in the false positive image data 51 are similar. Since the similarity rate is less than the reference degree, the process of step S18 is skipped. Then, in step S19, the object 63X is notified.

As shown in FIG. 5, it is assumed that the target image data 61Y is acquired in step S11, the difference between the target image data 61Y and the background image data 62 is calculated in step S12, and the target object 63Y is specified. Then, since the object 63Y is specified in step S13, the process proceeds to step S14. In steps S14 to S15, the target area where the object 63Y is located and the false detection area 52 are substantially at the same position, and the overlap rate is equal to or higher than the reference rate, so that the process proceeds to step S16. In steps S16 to S17, the target partial data which is the partial image data of the target region in the target image data 61Y and the false positive partial data which is the partial image data of the false positive region 52 in the false positive image data 51 are similar. Since the similarity rate is equal to or higher than the reference degree, the process proceeds to step S18. Then, in step S18, the object 63Y is excluded from the notification candidates.

With reference to FIG. 6, a process of detecting an object from the target image data in step S12 according to the first embodiment will be described.
(Step S21: Image acquisition process)
The detection unit 22 reads the target image data from the memory 12.

(Step S22: Background subtraction calculation process)
The detection unit 22 calculates background difference information indicating a region having a difference between the target image data acquired in step S21 and the background image data.
The background subtraction information indicates a region where there is a difference in the texture features that quantitatively represent the properties of the texture between the target image data and the background image data. The properties of the texture are the roughness of the texture, the directionality, the regularity, and whether it is granular or linear. In the first embodiment, the texture feature is an edge feature extracted by edge detection.

(Step S23: Object detection process)
The detection unit 22 sets an object having an object in a region indicating that the background difference information calculated in step S22 has a difference by using image data equal to or larger than the reference number among the image data of a plurality of past frames as the target image data. Detect as an area.
The reference number is a value that is determined in advance from information such as the relationship between detection omission and false positives and the number of frames in the past.

(Step S24: Area determination process)
The detection unit 22 determines whether or not the target area detected in step S23 is a new target area.
If the detection unit 22 is a new target area, the detection unit 22 advances the process to step S25. On the other hand, if the detection unit 22 is not a new target area, the detection unit 22 proceeds to step S26.

(Step S25: Candidate addition process)
The detection unit 22 adds the target area detected in step S23 to the target candidate. Then, the detection unit 22 selects the target image data indicating that the background difference information has a difference in the target area detected in step S23 among the target image data of the past plurality of frames targeted in step S23. Extract as a template for.
In the first embodiment, the detection unit 22 extracts the target image data acquired in step S21 as a template. Here, since the target area is newly detected by acquiring the target image data acquired in step S21, the background difference information about the target image data acquired in step S21 has a difference with respect to the target area. Show that.

The processing of steps S26 to S29 is executed for each target candidate.

(Step S26: Similarity determination process)
The detection unit 22 determines whether or not the target image data acquired in step S21 and the background image data are similar to the region indicated by the target candidate. Further, the detection unit 22 determines whether or not the target image data acquired in step S21 and the template for the target target candidate are similar to the region indicated by the target target candidate.
Specifically, the detection unit 22 converts the area indicated by the target candidate of the target image data and the background image data into grayscale image data, performs edge detection, and extracts edge features. Then, the detection unit 22 calculates the correlation value between the edge feature of the target image data and the edge feature of the background image data. When the correlation value is equal to or greater than the threshold value, the detection unit 22 determines that the target image data and the background image data are similar.
Similarly, the detection unit 22 converts the target image data and the region indicated by the target candidate of the template into grayscale image data, performs edge detection, and extracts edge features. Then, the detection unit 22 calculates the correlation value between the edge feature of the target image data and the edge feature of the template. The detection unit 22 determines that the target image data and the template are similar when the correlation value is equal to or greater than the threshold value.

When the detection unit 22 determines that the target image data and the background image data are similar, the detection unit 22 proceeds to the process in step S27. If the detection unit 22 determines that the target image data and the template are similar, the detection unit 22 proceeds to step S28. Further, when it is determined that the target image data is not similar to either the background image data or the template, the detection unit 22 ends the processing for the target target candidate.

(Step S27: Candidate exclusion process)
The detection unit 22 excludes the area indicated by the target candidate from the target candidate. That is, the fact that the target image data and the background image data are similar in the area indicated by the target target candidate means that there is no object in the area indicated by the target target candidate, so that the detection unit 22 may use the detection unit 22. , Exclude the area indicated by the target candidate from the target candidate.

(Step S28: Time determination process)
The detection unit 22 determines whether or not a reference period or more has elapsed since the target candidate was added. The reference period is a period determined according to how long it has passed since the item was placed to be treated as a leftover item.
When the reference period or more has elapsed, the detection unit 22 advances the process to step S29. On the other hand, if the reference period or more has not elapsed, the detection unit 22 ends the processing for the target candidate.

(Step S29: Object identification process)
The detection unit 22 identifies that the object is in the region indicated by the target candidate. The object specified here becomes the object detected in step S12.

*** Effect of Embodiment 1 ***
As described above, the notification determination device 10 according to the first embodiment detects an object by using false detection image data which is image data in which a false detection object determined to be false notification in the past is detected. It is determined whether or not to notify that the data has been sent. As a result, it is difficult to prevent erroneous detection only by the logic of the detection process, and it is possible to prevent notification in cases where erroneous detection occurs many times.
For example, there are cases where you want to be notified of shadows that appear at one position, but you do not want to be notified of shadows that appear at another position. In this case, it is difficult to deal with it by logic, but it can be easily dealt with by the notification determination device 10 according to the first embodiment.

In particular, in the notification determination device 10 according to the first embodiment, the overlap rate between the target area with the object and the false detection area with the false detection object is equal to or higher than the reference rate, and the notification determination device 10 is partial image data of the target area. If the degree of similarity between the target partial data and the false positive partial data, which is the partial image data of the false detection area, is equal to or higher than the reference degree, it is determined that the notification is not performed. This makes it possible not to notify false detections that constantly occur due to factors such as changes in sunshine and shadows.

The notification determination device 10 according to the first embodiment can store a plurality of erroneous notification data in the erroneous detection data storage unit 31. The tendency of false detection differs depending on factors such as time zone or weather, but by storing false notification data corresponding to each tendency in the false detection data storage unit 31, false detection in each tendency is not notified. It is possible.

The notification determination device 10 according to the first embodiment includes the target partial data and the false positive partial data only when the overlap rate between the target area containing the object and the false detection area containing the false detection object is equal to or higher than the reference rate. Calculate the similarity of. It is possible to reduce the processing load as a whole by suppressing the number of executions of the calculation of similarity, which has a higher processing load than the calculation of the duplication rate.

*** Other configurations ***
<Modification 1>
In the first embodiment, the similarity calculation unit 24 calculates the similarity using the edge difference. However, the similarity may be calculated using not only the edge difference but also another difference method such as a color difference.

<Modification 2>
In the first embodiment, the similarity calculation unit 24 calculates the similarity using the edge difference. In the edge difference, the amount of deviation between image data can be calculated from the peak value of the correlation. Therefore, when the deviation amount between the target portion data and the false positive portion data is equal to or greater than the threshold value, the similarity calculation unit 24 shifts one of the image data by the deviation amount and recalculates the similarity. May be good. This makes it possible to calculate the similarity with high accuracy.

<Modification 3>
In the first embodiment, each functional component is realized by software. However, as a modification 3, each functional component may be realized by hardware. The difference between the third modification and the first embodiment will be described.

The configuration of the notification determination device 10 according to the modification 3 will be described with reference to FIG. 7.
When each functional component is realized by hardware, the notification determination device 10 includes an electronic circuit 15 in place of the processor 11, the memory 12, and the storage 13. The electronic circuit 15 is a dedicated circuit that realizes the functions of each functional component, the memory 12, and the storage 13.

Examples of the electronic circuit 15 include a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, a logic IC, a GA (Gate Array), an ASIC (Application Specific Integrated Circuit), and an FPGA (Field-Programmable Gate Array). is assumed.
Each functional component may be realized by one electronic circuit 15, or each functional component may be distributed and realized by a plurality of electronic circuits 15.

<Modification example 4>
As a modification 4, some functional components may be realized by hardware, and other functional components may be realized by software.

The processor 11, the memory 12, the storage 13, and the electronic circuit 15 are referred to as a processing circuit. That is, the function of each functional component is realized by the processing circuit.

10 Notification Judgment Device, 11 Processor, 12 Memory, 13 Storage, 14 Communication Interface, 15 Electronic Circuit, 21 Image Acquisition Unit, 22 Detection Unit, 23 Duplicate Calculation Unit, 24 Similarity Calculation Unit, 25 Judgment Unit, 26 Notification Unit, 31 False positive data storage unit, 41 Imaging device, 51 False positive image data, 52 False positive area, 61 Target image data, 62 Background image data, 63 Target, 71 Pillars, 72 Shadows.

Claims (7)

The target area in the target image data in which the target is detected and the false detection image data stored in the false detection data storage unit, which are determined to be false notifications in the past. A duplicate calculation unit that calculates the overlap rate of the false positive image data, which is the image data in which the false positive is detected, with the false positive area where the false positive is found.
Similarity calculation unit for calculating the similarity between the target partial data which is the partial image data of the target area in the target image data and the false positive partial data which is the partial image data of the false detection area in the false detection image data. When,
Based on the duplication rate calculated by the duplication calculation unit and the similarity calculated by the similarity calculation unit, a determination unit for determining whether or not to notify that the object has been detected, and a determination unit.
When it is determined by the determination unit to notify the object, and as a result of the notification, if the notification is found to be erroneous, the target image data is used as the false detection image data and the false detection data is used. A notification determination device including a notification unit to be added to the storage unit. The notification determination device according to claim 1, wherein the determination unit determines not to notify when the overlap rate is equal to or higher than the reference rate and the similarity is equal to or higher than the reference degree. The notification determination device according to claim 2, wherein the similarity calculation unit calculates the similarity between the target partial data and the false positive partial data when the overlap rate is equal to or higher than the reference rate. The duplication calculation unit sets each of the plurality of false positive objects as the target false positive objects, and includes the target area and the false positive area where the false positive objects are present in the false positive image data in which the false positive objects of the target are detected. Calculate the duplication rate of
The similarity calculation unit includes false detection partial data in false positive image data in which the overlap rate is equal to or higher than the reference rate among false positive image data in which each of the plurality of false positive objects is detected, and the target partial data. The notification determination device according to any one of claims 1 to 3, wherein the similarity is calculated. The target image data is image data obtained by shooting a shooting area with a shooting device.
The notification determination device further
Detection to detect the object from the target image data by comparing the target image data with the background image data obtained by photographing the imaged area in a state where the object is not present in the imaged area. The notification determination device according to any one of claims 1 to 4, further comprising a unit. The duplicate calculation unit is the target area in the target image data, which is the image data in which the object is detected, and the false detection image data stored in the false detection data storage unit. The overlap rate of the false positive object with the false positive area in the false positive image data, which is the image data in which the false positive object determined to be present is detected, is calculated.
The similarity calculation unit determines the degree of similarity between the target partial data, which is the partial image data of the target region in the target image data, and the false positive partial data, which is the partial image data of the false detection region in the false detection image data. Calculate and
The determination unit determines whether or not to notify that the object has been detected based on the overlap rate and the similarity.
When it is determined to notify, the object is notified, and when the notification is found to be erroneous as a result of the notification, the target image data is added to the false detection data storage unit as the false detection image data. Notification judgment method to be performed. The target area in the target image data in which the target is detected and the false detection image data stored in the false detection data storage unit, which are determined to be false notifications in the past. Duplicate calculation processing for calculating the overlap rate of the false detection object with the false detection area in the false detection image data, which is the image data in which the false detection object is detected, and
Similarity calculation process for calculating the similarity between the target partial data which is the partial image data of the target area in the target image data and the false positive partial data which is the partial image data of the false detection area in the false detection image data. When,
A determination process for determining whether or not to notify that the object has been detected based on the duplication rate calculated by the duplication calculation process and the similarity calculated by the similarity calculation process.
When it is determined to be notified by the determination process, the object is notified, and when the notification is found to be erroneous as a result of the notification, the target image data is used as the false detection image data and the false detection data is used. A notification judgment program that makes a computer function as a notification judgment device that performs notification processing to be added to the storage unit.

Images