Method and device for verifying identification accuracy of nuclear radiation bright spot

Abstract

The invention provides a nuclear radiation bright spot recognition accuracy verification method and device, and relates to the technical field of nuclear radiation detection. The method comprises the following steps: acquiring video image data about nuclear radiation bright spots; predicting nuclear radiation bright spots in the video image data by adopting a nuclear radiation bright spot identification model to obtain a prediction frame and prediction frame information of the nuclear radiation bright spots; starting an internal calculator to count and accumulate according to the prediction box information; when the counting accumulation reaches a preset number, drawing the prediction frame on a new blank image to obtain a bright spot accumulation image; calculating a spatial entropy value, a maximum possible entropy value and an average nearest neighbor index of the bright spot accumulation image; according to the spatial entropy, the maximum possible entropy and the average nearest neighbor index, the distribution condition of the nuclear radiation bright spots on the image is judged; and verifying the accuracy of identifying the nuclear radiation bright spots by the nuclear radiation bright spot identification model according to the distribution condition. By adopting the method, the identification effect of the model on the nuclear radiation bright spot in the dynamic radiation scene can be evaluated.

Description

Technical Field

[0001] The present invention relates to the technical field of nuclear radiation detection, and particularly to a method and device for verifying the accuracy of nuclear radiation hotspot recognition. Background Art

[0002] In nuclear radiation detection, using a CMOS sensor combined with a deep learning model to identify nuclear radiation hotspots has become an important research direction. However, currently, there is a lack of an effective method to verify the accuracy of the nuclear radiation hotspot recognition model in identifying nuclear radiation hotspots in complex dynamic radiation scenarios. Existing technologies are difficult to accurately determine whether the model is interfered by moving objects and environmental light changes, and cannot ensure the reliability of the model's recognition results. As a result, in practical applications, misjudgment or missed judgment of nuclear radiation hotspots may occur, affecting the accuracy of nuclear safety monitoring and posing a potential threat to personnel safety and the environment. Therefore, a reliable verification method and system are urgently needed. Summary of the Invention

[0003] To solve the technical problem that the existing nuclear radiation hotspot recognition model lacks an effective means for verifying accuracy in complex dynamic scenarios, making it difficult to judge the reliability of the recognition results, the embodiments of the present invention provide a method and device for verifying the accuracy of nuclear radiation hotspot recognition. The technical solutions are as follows:

[0004] On the one hand, a method for verifying the accuracy of nuclear radiation hotspot recognition is provided. This method is implemented by a device for verifying the accuracy of nuclear radiation hotspot recognition, and the method includes:

[0005] S1. Obtain video image data regarding nuclear radiation hotspots; use a nuclear radiation hotspot recognition model to predict the nuclear radiation hotspots in the video image data, and obtain the prediction boxes and prediction box information of the nuclear radiation hotspots;

[0006] S2. According to the prediction box information, start an internal calculator to perform counting and accumulation; when the counting and accumulation reach a preset quantity, draw the prediction boxes on a new blank image to obtain a hotspot accumulation image;

[0007] S3. Calculate the spatial entropy value, maximum possible entropy value, and average nearest neighbor index of the hotspot accumulation image;

[0008] S4. According to the spatial entropy value, maximum possible entropy value, and average nearest neighbor index, judge the distribution of nuclear radiation hotspots on the image; according to the distribution, verify the accuracy of the nuclear radiation hotspot recognition model in recognizing nuclear radiation hotspots.

[0009] Optionally, the calculation process of calculating the spatial entropy value of the hotspot accumulation image includes:

[0010] Divide the hotspot accumulation image into multiple small regions;

[0011] Calculate the probability of radiation events occurring in each small area;

[0012] Calculate the spatial entropy value of the highlight accumulation image according to the probability of radiation events occurring in each small area; among them, the calculation formula of the spatial entropy value of the highlight accumulation image is represented by the following formula (1):

[0013] (1)

[0014] Among them, represents the spatial entropy value; is the number of regions; is the region the probability of radiation events occurring inside, that is, the ratio of the number of radiation highlights in region to the total number of radiation highlights.

[0015] Optionally, the calculation formula of the maximum possible entropy value of the highlight accumulation image is represented by the following formula (2):

[0016] (2)

[0017] Among them, represents the maximum possible entropy value; n represents the number of regions.

[0018] Optionally, the process of calculating the average nearest neighbor index of the highlight accumulation image includes:

[0019] Obtain the set of center points of the prediction box according to the prediction box;

[0020] Calculate the actual average distance between each point in the set of center points of the prediction box and its nearest neighbor point;

[0021] Calculate the point density, and calculate the expected average nearest neighbor distance according to the point density; among them, the calculation formula of the expected average nearest neighbor distance is represented by the following formula (3):

[0022] (3)

[0023] Among them, represents the expected average nearest neighbor distance; represents the point density;

[0024] Obtain the average nearest neighbor index of the highlight accumulation image according to the actual average distance and the expected average nearest neighbor distance; among them, the average nearest neighbor index formula is represented by the following formula (4):

[0025] (4)

[0026] Among them, represents the average nearest neighbor index; Indicates the actual average distance.

[0027] Optionally, the rule for judging the distribution of nuclear radiation highlights on the image according to the spatial entropy value includes:

[0028] If , it is judged that the nuclear radiation highlights are approximately two-dimensionally uniformly distributed;

[0029] If , it is judged that the nuclear radiation highlights are aggregated;

[0030] If , it is judged that the nuclear radiation highlights are dispersed.

[0031] Optionally, the rule for judging the distribution of nuclear radiation highlights on the image according to the spatial entropy value and the maximum possible entropy value includes:

[0032] Compare the difference between the spatial entropy value and the maximum possible entropy; where the smaller the difference, the closer the distribution of the nuclear radiation highlights is to two-dimensional uniform distribution.

[0033] On the other hand, a device for verifying the recognition accuracy of nuclear radiation highlights is provided. The device is applied to the method for verifying the recognition accuracy of nuclear radiation highlights, and the device includes:

[0034] A prediction box acquisition unit, configured to acquire video image data about nuclear radiation highlights; use a nuclear radiation highlight recognition model to predict the nuclear radiation highlights in the video image data, and obtain a prediction box and prediction box information of the nuclear radiation highlights;

[0035] An image drawing unit, configured to start an internal calculator for counting and accumulating according to the prediction box information; when the counting and accumulation reach a preset quantity, draw the prediction box on a new blank image to obtain a highlight accumulation image;

[0036] A data calculation unit, configured to calculate the spatial entropy value, the maximum possible entropy value, and the average nearest neighbor index of the highlight accumulation image;

[0037] A result judgment unit, configured to judge the distribution of nuclear radiation highlights on the image according to the spatial entropy value, the maximum possible entropy value, and the average nearest neighbor index; and verify the recognition accuracy of the nuclear radiation highlight recognition model according to the distribution.

[0038] Optionally, the calculation process of calculating the spatial entropy value of the highlight accumulation image includes:

[0039] Divide the highlight accumulation image into multiple small regions;

[0040] Calculate the probability of radiation events occurring in each small region;

[0041] Calculate the spatial entropy value of the highlight accumulation image according to the probability of radiation events occurring in each small area; among them, the calculation formula of the spatial entropy value of the highlight accumulation image is expressed by the following formula (1):

[0042] (1)

[0043] Among them, represents the spatial entropy value; is the number of regions; is the region the probability of a radiation event occurring within, that is, the ratio of the number of radiation highlights in region to the total number of radiation highlights.

[0044] Optionally, the calculation formula of the maximum possible entropy value of the highlight accumulation image is expressed by the following formula (2):

[0045] (2)

[0046] Among them, represents the maximum possible entropy value; n represents the number of regions.

[0047] Optionally, the process of calculating the average nearest neighbor index of the highlight accumulation image includes:

[0048] Obtain the set of center points of the prediction box according to the prediction box;

[0049] Calculate the actual average distance between each point in the set of center points of the prediction box and its nearest neighbor point;

[0050] Calculate the point density, and calculate the expected average nearest neighbor distance according to the point density; among them, the calculation formula of the expected average nearest neighbor distance is expressed by the following formula (3):

[0051] (3)

[0052] Among them, represents the expected average nearest neighbor distance; represents the point density;

[0053] Obtain the average nearest neighbor index of the highlight accumulation image according to the actual average distance and the expected average nearest neighbor distance; among them, the average nearest neighbor index formula is expressed by the following formula (4):

[0054] (4)

[0055] Among them, represents the average nearest neighbor index; represents the actual average distance.

[0056] Optionally, the rule for judging the distribution of nuclear radiation highlights on the image according to the spatial entropy value includes:

[0057] If , it is judged that the nuclear radiation highlights are approximately two-dimensionally uniformly distributed;

[0058] If , it is judged that the nuclear radiation highlights are aggregated;

[0059] If , it is judged that the nuclear radiation highlights are dispersed.

[0060] Optionally, the rule for judging the distribution of nuclear radiation highlights on the image according to the spatial entropy value and the maximum possible entropy value includes:

[0061] Compare the difference between the spatial entropy value and the maximum possible entropy; where the smaller the difference, the closer the distribution of the nuclear radiation highlights is to two-dimensional uniform distribution.

[0062] On the other hand, a device for verifying the recognition accuracy of nuclear radiation highlights is provided. The device for verifying the recognition accuracy of nuclear radiation highlights includes: a processor; a memory, and computer-readable instructions are stored on the memory. When the computer-readable instructions are executed by the processor, any one of the methods in the above-mentioned method for verifying the recognition accuracy of nuclear radiation highlights is implemented.

[0063] On the other hand, a computer-readable storage medium is provided, and at least one instruction is stored in the storage medium. The at least one instruction is loaded and executed by a processor to implement any one of the methods in the above-mentioned method for verifying the recognition accuracy of nuclear radiation highlights.

[0064] The beneficial effects brought by the technical solution provided by the embodiments of the present invention at least include:

[0065] In the embodiments of the present invention, first, video image data about nuclear radiation highlights is obtained; a nuclear radiation highlight recognition model is used to predict the nuclear radiation highlights in the video image data to obtain the prediction boxes and prediction box information of the nuclear radiation highlights; secondly, according to the prediction box information, an internal calculator is started for counting and accumulation; when the counting and accumulation reach a preset quantity, the prediction boxes are drawn on a new blank image to obtain a highlight accumulation image; the spatial entropy value, the maximum possible entropy value, and the average nearest neighbor index of the highlight accumulation image are calculated; finally, according to the spatial entropy value, the maximum possible entropy value, and the average nearest neighbor index, the distribution of the nuclear radiation highlights on the image is judged; according to the distribution, the accuracy of the nuclear radiation highlight recognition model in recognizing nuclear radiation highlights is verified.

[0066] Embodiments of the present invention can intuitively and quantitatively evaluate the recognition effect of a model on nuclear radiation highlights in a dynamic radiation scenario through an innovative spatial distribution feature analysis method, effectively solving the problem of the lack of effective verification means in the prior art, and providing strong support for the development of nuclear radiation detection technology. The present invention can accurately determine whether the model is interfered by external factors, timely discover problems existing in the model, provide a strong basis for model optimization, thereby improving the accuracy and reliability of nuclear radiation detection, reducing the risk of nuclear radiation monitoring, and ensuring personnel safety and environmental safety; Embodiments of the present invention have good versatility and scalability, can be applied to a variety of nuclear radiation detection scenarios and model verifications, and promote the development of nuclear radiation detection technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0068] Figure 1 is a flowchart of a method for verifying the accuracy of nuclear radiation highlight recognition provided by an embodiment of the present invention;

[0069] Figure 2 is a distribution diagram of a scattered point set with an average nearest neighbor index ANNI = 1.91 provided by an embodiment of the present invention;

[0070] Figure 3 is a distribution diagram of an aggregated point set with an average nearest neighbor index ANNI = 0.12 provided by an embodiment of the present invention;

[0071] Figure 4 is a two-dimensional uniform point set distribution diagram with an average nearest neighbor index ANNI = 1.08 provided by an embodiment of the present invention;

[0072] Figure 5 is a highlight accumulation image without blocking the monitoring camera provided by an embodiment of the present invention;

[0073] Figure 6 is a highlight accumulation image with a blocked monitoring camera provided by an embodiment of the present invention;

[0074] Figure 7 is a block diagram of a device for verifying the accuracy of nuclear radiation highlight recognition provided by an embodiment of the present invention;

[0075] Figure 8 is a structural schematic diagram of a device for verifying the accuracy of nuclear radiation highlight recognition provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0076] The technical solutions in the present invention will be described below with reference to the accompanying drawings.

[0077] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to represent examples, illustrations or explanations. Any embodiment or design solution described as an "example" in the present invention should not be construed as being more preferred or more advantageous than other embodiments or design solutions. Rather, the use of the word "example" is intended to present concepts in a specific manner. In addition, in the embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one of the two.

[0078] In the embodiments of the present invention, "image" and "picture" can sometimes be used interchangeably. It should be noted that when the difference is not emphasized, the meanings they express are the same. "(of)", "corresponding", and "corresponding" can sometimes be used interchangeably. It should be noted that when the difference is not emphasized, the meanings they express are the same.

[0079] In the embodiments of the present invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meanings they express are the same.

[0080] To make the technical problems to be solved, technical solutions and advantages of the present invention clearer, the following will be described in detail with reference to the accompanying drawings and specific embodiments.

[0081] The embodiments of the present invention provide a method for verifying the accuracy of nuclear radiation highlight recognition. This method can be implemented by a device for verifying the accuracy of nuclear radiation highlight recognition, and this device for verifying the accuracy of nuclear radiation highlight recognition can be a terminal or a server. As Figure 1 shown in the flowchart of the method for verifying the accuracy of nuclear radiation highlight recognition, the processing flow of this method can include the following steps:

[0082] S1. Obtain video image data about nuclear radiation highlights; use a nuclear radiation highlight recognition model to predict the nuclear radiation highlights in the video image data, and obtain the prediction box and prediction box information of the nuclear radiation highlights.

[0083] Among them, the video image data of the nuclear radiation highlights is sourced from two groups of monitoring video data in the radiotherapy department of a certain hospital; among them, the first group of video image data is collected after covering the monitoring lens, and the image only contains nuclear radiation highlights; the second group of video image data is collected without covering the monitoring lens, and the image not only contains nuclear radiation highlights, but also includes dynamic elements such as walking doctors and rotating machines, which is a real complex radiation scene. Among them, the video resolution is 1920×1080p, and the frame rate is 30FPS.

[0084] Among them, the nuclear radiation highlight recognition model can be a recognition model based on deep learning, which is not limited in this application.

[0085] S2. According to the prediction box information, start the internal calculator for counting accumulation; when the counting accumulation reaches a preset quantity, draw the prediction box on a new blank image to obtain the highlight accumulation image.

[0086] Among them, when starting the internal calculator, each time a prediction box is received, the counter is incremented by 1; the preset quantity is 500; when the counting accumulation reaches 500, according to the information of the prediction box, draw the position and range of each highlight on a new blank image to obtain the highlight accumulation image. Among them, the information of the prediction box includes: the coordinate and size information of the prediction box.

[0087] S3. Calculate the spatial entropy value, maximum possible entropy value, and average nearest neighbor index of the highlight accumulation image.

[0088] Optionally, the calculation process of calculating the spatial entropy value of the highlight accumulation image includes:

[0089] Divide the highlight accumulation image into multiple small regions;

[0090] In a feasible implementation manner, divide the highlight accumulation image into 32×32 small regions, and the size of each region is 20×20; among them, dividing the highlight accumulation image into multiple small regions can more carefully analyze the distribution of highlights in the image and ensure the accuracy of the calculation.

[0091] Calculate the probability of radiation events occurring in each small region;

[0092] Among them, by calculating the regional probability, the possibility of highlights appearing in different regions can be reflected.

[0093] According to the probability of radiation events occurring in each small region, calculate the spatial entropy value of the highlight accumulation image; among them, the calculation formula of the spatial entropy value of the highlight accumulation image is represented by the following formula (1):

[0094] (1)

[0095] Among them, represents the spatial entropy value; is the number of regions; is the region the probability of radiation events occurring inside, that is, the ratio of the number of radiation highlights in region to the total number of radiation highlights.

[0096] In a feasible implementation, the degree of disorder of the system is measured according to entropy. When the distribution of points in each region is completely random, the maximum possible spatial entropy will occur. In this case, all nuclear radiation bright spots are evenly distributed in each small region, so that the probability of radiation events occurring in all regions is equal to .

[0097] Optionally, the calculation formula of the maximum possible entropy value of the bright spot accumulation image is expressed by the following formula (2):

[0098] (2)

[0099] wherein, represents the maximum possible entropy value; n represents the number of regions.

[0100] Optionally, the process of calculating the average nearest neighbor index of the bright spot accumulation image includes:[[]]

[0101] Obtaining a set of center points of the prediction box according to the prediction box;

[0102] Among them, the present application determines a set of 500 center points of the prediction box.

[0103] Calculating the actual average distance between each point in the set of center points of the prediction box and its nearest neighbor point;

[0104] Calculating the point density, and calculating the expected average nearest neighbor distance according to the point density; wherein, the calculation formula of the expected average nearest neighbor distance is expressed by the following formula (3):

[0105] (3)

[0106] wherein, represents the expected average nearest neighbor distance; represents the point density;

[0107] Obtaining the average nearest neighbor index of the bright spot accumulation image according to the actual average distance and the expected average nearest neighbor distance; wherein, the average nearest neighbor index formula is expressed by the following formula (4):

[0108] (4)

[0109] wherein, represents the average nearest neighbor index; represents the actual average distance.

[0110] S4. Judging the distribution of nuclear radiation bright spots on the image according to the spatial entropy value, the maximum possible entropy value and the average nearest neighbor index; and verifying the accuracy of the nuclear radiation bright spot recognition model for recognizing nuclear radiation bright spots according to the distribution situation.

[0111] Optionally, the rule for judging the distribution of the nuclear radiation bright spots on the image according to the spatial entropy value includes:

[0112] like , it is judged that the nuclear radiation bright spots are approximately uniformly distributed in two dimensions;

[0113] in, Figure 2 is a distribution diagram of a scattered point set with an average nearest neighbor index ANNI=1.91 provided by an embodiment of the present invention. , judging that the nuclear radiation bright spots are distributed in clusters;

[0114] in, Figure 3 This is a distribution diagram of clustered points with an average nearest neighbor index ANNI=0.12 provided by an embodiment of the present invention.

[0115] like , judging that the nuclear radiation bright spots are dispersed.

[0116] in, Figure 4 It is a two-dimensional uniform point set distribution diagram with an average nearest neighbor index ANNI=1.08 provided by an embodiment of the present invention.

[0117] Optionally, the rule for judging the distribution of the nuclear radiation bright spots on the image according to the spatial entropy value and the maximum possible entropy value includes:

[0118] Compare the difference between the spatial entropy value and the maximum possible entropy; the smaller the difference, the closer the distribution of the nuclear radiation bright spots is to a two-dimensional uniform distribution.

[0119] In a feasible implementation, the difference between the spatial entropy value and the maximum possible entropy is compared to evaluate the closeness of the distribution of the nuclear radiation bright spot to the two-dimensional uniform distribution. The smaller the difference, the closer the bright spot distribution is to the uniform distribution, and the more reliable the model recognition result is.

[0120] In the process of calculating spatial entropy, the accuracy of data statistics must be ensured to avoid deviations in calculation results due to unreasonable regional division or incorrect statistics on the number of bright spots.

[0121] Among them, Figure 5 is a bright spot accumulation image that does not block the monitoring lens provided by an embodiment of the present invention, The value is 0.863, and the difference between the spatial entropy and the maximum possible entropy is 1.656, indicating that there is a certain deviation in the model recognition in this scenario, which is due to the interference of residual information of the mobile device or the influence of high brightness areas during the imaging process, indicating that the nuclear radiation bright spot recognition model needs to be improved in a targeted manner; Figure 6 is an accumulated image of bright spots blocking the monitoring lens provided by an embodiment of the present invention; The value is 1.01, and the difference between the spatial entropy and the maximum possible entropy is 0.896, indicating that the nuclear radiation bright spot recognition model has a high recognition accuracy in this scenario.

[0122] In an embodiment of the present invention, first, video image data regarding nuclear radiation highlights is acquired; a nuclear radiation highlight recognition model is used to predict the nuclear radiation highlights in the video image data to obtain a prediction box and prediction box information of the nuclear radiation highlights; secondly, according to the prediction box information, an internal calculator is started for counting and accumulating; when the count accumulation reaches a preset quantity, the prediction box is drawn on a new blank image to obtain a highlight accumulation image; the spatial entropy value, the maximum possible entropy value, and the average nearest neighbor index of the highlight accumulation image are calculated; finally, according to the spatial entropy value, the maximum possible entropy value, and the average nearest neighbor index, the distribution of the nuclear radiation highlights on the image is judged; and according to the distribution, the accuracy of the nuclear radiation highlight recognition model in recognizing nuclear radiation highlights is verified.

[0123] Through an innovative spatial distribution feature analysis method, the embodiment of the present invention can intuitively and quantitatively evaluate the recognition effect of the model on nuclear radiation highlights in a dynamic radiation scenario, effectively solve the problem of the lack of effective verification means in the prior art, and provide strong support for the development of nuclear radiation detection technology. The present invention can accurately judge whether the model is interfered by external factors, timely discover problems existing in the model, provide a strong basis for model optimization, thereby improving the accuracy and reliability of nuclear radiation detection, reducing the risk of nuclear radiation monitoring, and ensuring personnel safety and environmental safety; the embodiment of the present invention has good versatility and scalability, can be applied to a variety of nuclear radiation detection scenarios and model verifications, and promotes the development of nuclear radiation detection technology.

[0124] Figure 7 is a block diagram of a device for verifying the recognition accuracy of nuclear radiation highlights shown according to an exemplary embodiment. This device is used for the method of verifying the recognition accuracy of nuclear radiation highlights. Refer to Figure 7 , this device includes a prediction box acquisition unit 710, an image drawing unit 720, a data calculation unit 730, and a result judgment unit 740. Among them:

[0125] The prediction box acquisition unit 710 is configured to acquire video image data regarding nuclear radiation highlights; use a nuclear radiation highlight recognition model to predict the nuclear radiation highlights in the video image data to obtain a prediction box and prediction box information of the nuclear radiation highlights;

[0126] The image drawing unit 720 is configured to, according to the prediction box information, start an internal calculator for counting and accumulating; when the count accumulation reaches a preset quantity, draw the prediction box on a new blank image to obtain a highlight accumulation image;

[0127] The data calculation unit 730 is configured to calculate the spatial entropy value, the maximum possible entropy value, and the average nearest neighbor index of the highlight accumulation image;

[0128] A result judgment unit 740 is configured to judge the distribution of nuclear radiation highlights on the image according to the spatial entropy value, the maximum possible entropy value, and the average nearest neighbor index; and verify the accuracy of the nuclear radiation highlight recognition model in recognizing nuclear radiation highlights according to the distribution.

[0129] Optionally, the calculation process of calculating the spatial entropy value of the highlight accumulation image includes:

[0130] Dividing the highlight accumulation image into multiple small regions;

[0131] Calculating the probability of radiation events occurring in each small region;

[0132] Calculating the spatial entropy value of the highlight accumulation image according to the probability of radiation events occurring in each small region; wherein, the calculation formula of the spatial entropy value of the highlight accumulation image is represented by the following formula (1):

[0133] (1)

[0134] Wherein, represents the spatial entropy value; is the number of regions; is the region the probability of radiation events occurring within, that is, the ratio of the number of nuclear radiation highlights in region to the total number of nuclear radiation highlights.

[0135] Optionally, the calculation formula of the maximum possible entropy value of the highlight accumulation image is represented by the following formula (2):

[0136] (2)

[0137] Wherein, represents the maximum possible entropy value; n represents the number of regions.

[0138] Optionally, the process of calculating the average nearest neighbor index of the highlight accumulation image includes:

[0139] Obtaining a set of center points of the prediction boxes according to the prediction boxes;

[0140] Calculating the actual average distance between each point in the set of center points of the prediction boxes and its nearest neighbor point;

[0141] Calculating the point density, and calculating the expected average nearest neighbor distance according to the point density; wherein, the calculation formula of the expected average nearest neighbor distance is represented by the following formula (3):

[0142] (3)

[0143] Wherein, represents the expected average nearest neighbor distance; Indicates the point density;

[0144] According to the actual average distance and the expected average nearest neighbor distance, obtain the average nearest neighbor index of the bright spot accumulation image; wherein, the average nearest neighbor index formula is expressed by the following formula (4):

[0145] (4)

[0146] Wherein, Indicates the average nearest neighbor index; Indicates the actual average distance.

[0147] Optionally, the rule for judging the distribution of nuclear radiation bright spots on the image according to the spatial entropy value includes:

[0148] If , judge that the nuclear radiation bright spots are approximately two-dimensionally uniformly distributed;

[0149] If , judge that the nuclear radiation bright spots are aggregated;

[0150] If , judge that the nuclear radiation bright spots are dispersed.

[0151] Optionally, the rule for judging the distribution of nuclear radiation bright spots on the image according to the spatial entropy value and the maximum possible entropy value includes:

[0152] Compare the difference between the spatial entropy value and the maximum possible entropy; wherein, the smaller the difference, the closer the distribution of the nuclear radiation bright spots is to two-dimensional uniform distribution.

[0153] In the embodiment of the present invention, first, video image data about nuclear radiation bright spots is obtained; a nuclear radiation bright spot recognition model is used to predict the nuclear radiation bright spots in the video image data to obtain the prediction boxes and prediction box information of the nuclear radiation bright spots; secondly, according to the prediction box information, an internal calculator is started for counting and accumulating; when the counting and accumulation reach a preset quantity, the prediction boxes are drawn on a new blank image to obtain a bright spot accumulation image; calculate the spatial entropy value, the maximum possible entropy value and the average nearest neighbor index of the bright spot accumulation image; finally, according to the spatial entropy value, the maximum possible entropy value and the average nearest neighbor index, judge the distribution of the nuclear radiation bright spots on the image; according to the distribution, verify the accuracy of the nuclear radiation bright spot recognition model in recognizing nuclear radiation bright spots.

[0154] Embodiments of the present invention can intuitively and quantitatively evaluate the recognition effect of a model on nuclear radiation hotspots in a dynamic radiation scenario through an innovative spatial distribution feature analysis method, effectively solving the problem of the lack of effective verification means in the prior art, and providing strong support for the development of nuclear radiation detection technology. The present invention can accurately determine whether the model is interfered by external factors, timely discover problems existing in the model, provide a strong basis for model optimization, thereby improving the accuracy and reliability of nuclear radiation detection, reducing the risk of nuclear radiation monitoring, and ensuring personnel safety and environmental safety; Embodiments of the present invention have good versatility and scalability, can be applied to a variety of nuclear radiation detection scenarios and model verifications, and promote the development of nuclear radiation detection technology.

[0155] Figure 8 FIG. 4 is a schematic structural diagram of an apparatus for verifying the accuracy of nuclear radiation hotspot recognition provided by an embodiment of the present invention. As Figure 8 shown, the apparatus for verifying the accuracy of nuclear radiation hotspot recognition may include the above Figure 7 shown apparatus for verifying the accuracy of nuclear radiation hotspot recognition. Optionally, the apparatus 810 for verifying the accuracy of nuclear radiation hotspot recognition may include a first processor 2001.

[0156] Optionally, the apparatus 810 for verifying the accuracy of nuclear radiation hotspot recognition may further include a memory 2002 and a transceiver 2003.

[0157] Wherein, the first processor 2001 is connected to the memory 2002 and the transceiver 2003, such as through a communication bus.

[0158] Next, in conjunction with Figure 8 each component of the apparatus 810 for verifying the accuracy of nuclear radiation hotspot recognition will be specifically introduced:

[0159] Among them, the first processor 2001 is the control center of the apparatus 810 for verifying the accuracy of nuclear radiation hotspot recognition, which may be a single processor or a collective term for multiple processing elements. For example, the first processor 2001 is one or more central processing units (CPUs), or may be a specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present invention, such as: one or more digital signal processors (DSPs), or one or more field programmable gate arrays (FPGAs).

[0160] Optionally, the first processor 2001 can execute various functions of the nuclear radiation highlight recognition accuracy verification device 810 by running or executing software programs stored in the memory 2002 and calling data stored in the memory 2002.

[0161] In a specific implementation, as an embodiment, the first processor 2001 may include one or more CPUs, such as Figure 8 CPU0 and CPU1 shown in

[0162] In a specific implementation, as an embodiment, the nuclear radiation highlight recognition accuracy verification device 810 may also include multiple processors, such as Figure 8 the first processor 2001 and the second processor 2004 shown in

[0163] Among them, the memory 2002 is used to store software programs for implementing the solution of the present invention and is controlled by the first processor 2001 for execution. The specific implementation manner can refer to the above method embodiments and will not be elaborated here.

[0164] Optionally, the memory 2002 can be a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, a random access memory (RAM) or other types of dynamic storage devices that can store information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM), or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital versatile discs, Blu-ray discs, etc.), magnetic disk storage media, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory 2002 can be integrated with the first processor 2001 or exist independently and is coupled to the first processor 2001 through an interface circuit ( Figure 8 not shown in

[0165] The transceiver 2003 is used to communicate with a network device or a terminal device.

[0166] Optionally, the transceiver 2003 may include a receiver and a transmitter ( Figure 8 not shown separately). Among them, the receiver is used to implement the receiving function, and the transmitter is used to implement the transmitting function.

[0167] Optionally, the transceiver 2003 may be integrated with the first processor 2001 or exist independently, and is coupled to the first processor 2001 through an interface circuit ( Figure 8 not shown) of the nuclear radiation hot spot recognition accuracy verification device 810. The embodiments of the present invention do not make specific limitations on this.

[0168] It should be noted that Figure 8 the structure of the nuclear radiation hot spot recognition accuracy verification device 810 shown in

[0169] does not constitute a limitation on the router. The actual knowledge structure recognition device may include more or fewer components than shown in the figure, or combine some components, or have different component arrangements.

[0170] It should be understood that the first processor 2001 in the embodiments of the present invention may be a central processing unit (CPU), and this processor may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or this processor may also be any conventional processor, etc.

[0171] It should also be understood that the memory in the embodiments of the present invention can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memories. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), or a flash memory. The volatile memory can be a random access memory (RAM), which is used as an external cache. By way of example but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchlink DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0172] The above embodiments can be implemented in whole or in part by software, hardware (such as circuits), firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present invention are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wired (such as infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that contains one or more collections of available media. The available medium can be a magnetic medium (such as a floppy disk, a hard disk, a magnetic tape), an optical medium (such as a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0173] It should be understood that the term "and / or" in this document is merely a description of the association relationship between associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. Here, A and B can be singular or plural. In addition, the character " / " in this document generally represents an "or" relationship between the associated objects before and after, but it may also represent an "and / or" relationship, which can be specifically understood with reference to the context before and after.

[0174] In the present invention, "at least one" means one or more, and "a plurality" means two or more. "At least one of the following" or its similar expressions refer to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b, or c can represent: a, b, c, a - b, a - c, b - c, or a - b - c, where a, b, and c can be single or multiple.

[0175] It should be understood that in various embodiments of the present invention, the magnitudes of the sequence numbers of the above processes do not mean the order of execution. The order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

[0176] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

[0177] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the above-described devices, apparatuses, and units can refer to the corresponding processes in the foregoing method embodiments, and will not be described herein again.

[0178] In several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces. The indirect couplings or communication connections of the devices or units can be electrical, mechanical, or other forms.

[0179] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place, or can be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0180] In addition, the functional units in each embodiment of the present invention can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit.

[0181] When the above-mentioned functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or a part of this technical solution can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0182] As described above, the above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A method for verifying the accuracy of nuclear radiation bright spot identification, characterized in that: The method comprises: S1. Acquire video image data about nuclear radiation bright spots; use a nuclear radiation bright spot recognition model to predict the nuclear radiation bright spots in the video image data to obtain a prediction frame and prediction frame information of the nuclear radiation bright spots; S2, according to the prediction frame information, start the internal calculator to count and accumulate; when the count accumulation reaches a preset number, draw the prediction frame on a new blank image to obtain a bright spot accumulation image; S3, calculating the spatial entropy value, the maximum possible entropy value and the average nearest neighbor index of the bright spot accumulation image; S4. Determine the distribution of the nuclear radiation bright spots on the image based on the spatial entropy value, the maximum possible entropy value and the average nearest neighbor index; and verify the accuracy of the nuclear radiation bright spot recognition model in identifying the nuclear radiation bright spots based on the distribution.

2. The method for verifying the accuracy of nuclear radiation bright spot identification according to claim 1, characterized in that: The process of calculating the spatial entropy value of the bright spot accumulation image includes: Divide the bright spot accumulation image into multiple small areas; Calculate the probability of radiation events occurring in each small area; According to the probability of occurrence of radiation events in each small area, the spatial entropy value of the bright spot accumulation image is calculated; wherein the calculation formula of the spatial entropy value of the bright spot accumulation image is expressed by the following formula (1): （1） in, Represents the spatial entropy value; is the number of regions; For Region The probability of internal radiation events, i.e. the area The ratio of the number of internal radiation bright spots to the number of total radiation bright spots.

3. The method for verifying the accuracy of nuclear radiation bright spot identification according to claim 1, characterized in that: The calculation formula of the maximum possible entropy value of the bright spot accumulation image is expressed by the following formula (2): （2） in, represents the maximum possible entropy value; n represents the number of regions.

4. The method for verifying the accuracy of nuclear radiation bright spot identification according to claim 1, characterized in that: The process of calculating the average nearest neighbor index of the bright spot accumulation image includes: According to the prediction box, a set of prediction box center points is obtained; Calculate the actual average distance between each point in the prediction box center point set and its nearest neighbor point; Calculate the point density and calculate the expected average nearest neighbor distance based on the point density; the calculation formula of the expected average nearest neighbor distance is expressed by the following formula (3): （3） in, represents the expected average nearest neighbor distance; Indicates point density; According to the actual average distance and the expected average nearest neighbor distance, the average nearest neighbor index of the bright spot accumulation image is obtained; wherein the average nearest neighbor index formula is expressed by the following formula (4): （4） in, represents the average nearest neighbor index; Indicates the actual average distance.

5. The method for verifying the accuracy of nuclear radiation bright spot identification according to claim 1, characterized in that: The rule for judging the distribution of nuclear radiation bright spots on the image according to the spatial entropy value includes: like , it is judged that the nuclear radiation bright spots are approximately uniformly distributed in two dimensions; like It is determined that the nuclear radiation bright spots are distributed in clusters; like , judging that the nuclear radiation bright spots are dispersed.

6. The method for verifying the accuracy of nuclear radiation bright spot identification according to claim 1, characterized in that: The rule for judging the distribution of nuclear radiation bright spots on the image according to the spatial entropy value and the maximum possible entropy value includes: Compare the difference between the spatial entropy value and the maximum possible entropy; the smaller the difference, the closer the distribution of the nuclear radiation bright spots is to a two-dimensional uniform distribution.

7. A nuclear radiation bright spot recognition accuracy verification device, the nuclear radiation bright spot recognition accuracy verification device is used to implement the nuclear radiation bright spot recognition accuracy verification method as claimed in any one of claims 1 to 6, characterized in that: The device comprises: A prediction frame acquisition unit is used to acquire video image data about nuclear radiation bright spots; use a nuclear radiation bright spot recognition model to predict the nuclear radiation bright spots in the video image data to obtain a prediction frame and prediction frame information of the nuclear radiation bright spots; An image drawing unit is used to start an internal calculator to count and accumulate according to the prediction frame information; when the count accumulation reaches a preset number, the prediction frame is drawn on a new blank image to obtain a bright spot accumulation image; A data calculation unit, used for calculating the spatial entropy value, the maximum possible entropy value and the average nearest neighbor index of the bright spot accumulation image; The result judgment unit is used to judge the distribution of nuclear radiation bright spots on the image according to the spatial entropy value, the maximum possible entropy value and the average nearest neighbor index; and verify the accuracy of the nuclear radiation bright spot recognition model in identifying the nuclear radiation bright spots according to the distribution.

8. The nuclear radiation bright spot recognition accuracy verification device according to claim 1, characterized in that: The process of calculating the spatial entropy value of the bright spot accumulation image includes: Divide the bright spot accumulation image into multiple small areas; Calculate the probability of radiation events occurring in each small area; According to the probability of occurrence of radiation events in each small area, the spatial entropy value of the bright spot accumulation image is calculated; wherein the calculation formula of the spatial entropy value of the bright spot accumulation image is expressed by the following formula (1): （1） in, Represents the spatial entropy value; is the number of regions; For Region The probability of internal radiation events, i.e. the area The ratio of the number of internal radiation bright spots to the number of total radiation bright spots.

9. A nuclear radiation bright spot recognition accuracy verification device, characterized in that: The nuclear radiation bright spot recognition accuracy verification device comprises: processor; A memory having computer-readable instructions stored thereon, wherein when the computer-readable instructions are executed by the processor, the method according to any one of claims 1 to 6 is implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores program codes, which can be called by a processor to execute the method according to any one of claims 1 to 6.

Images