Nuclide identification method, electronic equipment and storage medium

Abstract

The invention provides a nuclide identification method, electronic equipment and a storage medium. The nuclide identification method comprises the following steps: acquiring a to-be-identified nuclide energy spectrum diagram; the to-be-identified nuclide energy spectrum diagram is a single nuclide energy spectrum diagram or a mixed nuclide energy spectrum diagram; performing nuclide identification on the to-be-identified nuclide energy spectrum diagram by using a target nuclide identification model to obtain a corresponding target identification result; the target nuclide identification model is obtained by performing model optimization training on a nuclide identification model to be trained; the to-be-trained nuclide identification model is obtained by adding an attention mechanism module in the convolutional neural network model; the attention mechanism module is used for performing space and channel weighting processing on a feature map output by a convolutional layer of the convolutional neural network model; the target recognition result is used for representing the nuclide corresponding to the to-be-recognized nuclide energy spectrum diagram. According to the technical scheme of the invention, efficient and accurate nuclide identification can be carried out on the to-be-identified nuclide energy spectrogram, so that the nuclide corresponding to the to-be-identified nuclide energy spectrogram can be efficiently and accurately identified.

Description

Technical Field

[0001] The present application relates to the field of data processing, and in particular to a nuclide identification method, an electronic device and a storage medium. Background Art

[0002] In the fields of nuclear energy safety, environmental monitoring, medical diagnosis, etc., radiation detection is often performed on the radioactive source first to obtain the energy spectrum of the nuclide to be identified. Then, the nuclide energy spectrum is used to identify the nuclide to determine the nuclide corresponding to the energy spectrum of the nuclide to be identified, and then the nuclide present in the radioactive source is determined.

[0003] When performing nuclide identification on the energy spectrum of the nuclide to be identified, if the energy spectrum of the nuclide to be identified is a single nuclide energy spectrum, the nuclide corresponding to the energy spectrum of the nuclide to be identified can usually be determined by analyzing the characteristic peaks. However, if the energy spectrum of the nuclide to be identified is a mixed nuclide energy spectrum, since the energy spectrum peaks of multiple nuclides may overlap, the characteristic peaks of different nuclides are intertwined in the energy spectrum. At this time, it is often difficult to determine the nuclide corresponding to the energy spectrum of the nuclide to be identified by analyzing the characteristic peaks.

[0004] However, in practical applications, the radioactive source may contain a single nuclide or a mixed nuclide. That is to say, in practical applications, the energy spectrum of the nuclide to be identified may be a single nuclide energy spectrum or a mixed nuclide energy spectrum. Therefore, in order to be able to perform efficient and accurate nuclide identification on the energy spectrum of the nuclide to be identified, whether it is a single nuclide energy spectrum or a mixed nuclide energy spectrum, a nuclide identification method is needed that can perform efficient and accurate nuclide identification on both the single nuclide energy spectrum and the mixed nuclide energy spectrum. Summary of the invention

[0005] The embodiments of the present application provide a nuclide identification method, an electronic device, and a storage medium to solve one or more of the above-mentioned technical problems.

[0006] In a first aspect, an embodiment of the present application provides a nuclide identification method, comprising:

[0007] Obtaining an energy spectrum of a nuclide to be identified; the energy spectrum of the nuclide to be identified is a single nuclide energy spectrum or a mixed nuclide energy spectrum;

[0008] The target nuclide identification model is used to identify the nuclide energy spectrum to be identified, and the corresponding target identification result is obtained; the target nuclide identification model is obtained based on the model optimization training of the nuclide identification model to be trained; the nuclide identification model to be trained is obtained by adding an attention mechanism module to the convolutional neural network model; the attention mechanism module is used to perform spatial and channel weighted processing on the feature map output by the convolution layer of the convolutional neural network model; the target identification result is used to represent the nuclide corresponding to the nuclide energy spectrum to be identified.

[0009] Optionally, an attention mechanism module is added after the convolutional layer.

[0010] Optionally, the steps of performing model optimization training on the nuclide identification model to be trained to obtain the target nuclide identification model are as follows:

[0011] A training set and a test set are obtained; the training set and the test set respectively include multiple sample nuclide energy spectra and the annotation results corresponding to each sample nuclide energy spectra; the multiple sample nuclide energy spectra include single nuclide energy spectra and mixed nuclide energy spectra;

[0012] The training set is used to perform multiple rounds of iterative training on the nuclide recognition model to be trained until the set iteration stop condition is met;

[0013] When the iteration stop condition is met, the nuclide identification model to be trained is tested using the test set, and when the test result meets the preset standard, the nuclide identification model to be trained is used as the target nuclide identification model.

[0014] Optionally, the steps for each round of iterative training are as follows:

[0015] Select the target sample nuclide energy spectrum in the training set;

[0016] Input the target sample nuclide energy spectrum into the nuclide identification model to be trained to obtain the corresponding sample identification result;

[0017] Calculate the target loss value between the sample recognition result and the nuclide labeling result corresponding to the target sample image;

[0018] The target loss value is used to update the internal model parameters in the nuclide identification model to be trained.

[0019] Optionally, the test result includes the accuracy rate corresponding to the nuclide identification model to be trained.

[0020] Optionally, the sample nuclide energy spectrum of the mixed nuclide energy spectrum among the multiple sample nuclide energy spectra is obtained by the following steps:

[0021] Obtain single nuclide energy spectra corresponding to multiple nuclides;

[0022] Among the single nuclide energy spectra corresponding to the plurality of nuclides, single nuclide energy spectra corresponding to at least two nuclides are randomly selected;

[0023] The single nuclide energy spectra corresponding to at least two nuclides are mixed to obtain a mixed nuclide energy spectrum, and the mixed nuclide energy spectrum is used as a sample nuclide energy spectrum.

[0024] Optionally, the single nuclide energy spectrum is a real nuclide energy spectrum obtained through radiation detection, or a simulated nuclide energy spectrum generated through computer simulation.

[0025] Optionally, the attention mechanism module is a convolutional block attention module.

[0026] In a second aspect, an embodiment of the present application provides an electronic device, including a memory, a processor, and a computer program stored in the memory, and the processor implements the method provided in any embodiment of the present application when executing the computer program.

[0027] In a third aspect, an embodiment of the present application provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the method provided in any embodiment of the present application is implemented.

[0028] Compared with the prior art, this application has the following advantages:

[0029] According to the technical solution of the present application, the target nuclide identification model is obtained based on model optimization training of the nuclide identification model to be trained, and the nuclide identification model to be trained is obtained by adding an attention mechanism module to the convolutional neural network model, and the attention mechanism module is used to perform spatial and channel weighted processing on the feature map output by the convolution layer of the convolutional neural network model.

[0030] By performing spatial weighting on the feature map, the target nuclide identification model can highlight the key areas in the energy spectrum to be identified, while suppressing irrelevant or redundant areas. In this way, the target nuclide identification model can more accurately extract the spatial features related to nuclide identification, improve the target nuclide identification model's ability to focus on the areas that are critical to nuclide identification, and reduce the interference of background noise and irrelevant information on the target nuclide identification model, so that the target nuclide identification model can better identify the characteristics of the nuclide.

[0031] At the same time, by performing channel weighting processing on the feature graph, the target nuclide identification model can highlight the most contributing feature channels while suppressing irrelevant or redundant channels. In this way, the target nuclide identification model can more accurately identify the key channel features in the nuclide energy spectrum, thereby improving the effect and accuracy of the target nuclide identification model.

[0032] Therefore, by performing spatial weighting and channel weighting processing on the feature map, the target nuclide identification model can more accurately identify the nuclide energy spectrum of the nuclide to be identified, thereby more accurately identifying the nuclide corresponding to the nuclide energy spectrum of the nuclide to be identified.

[0033] In addition, the technical solution of the present application can perform nuclide identification on the nuclide energy spectrum to be identified when the nuclide energy spectrum to be identified is a single nuclide energy spectrum or a mixed nuclide energy spectrum, thereby identifying the nuclide corresponding to the nuclide energy spectrum to be identified.

[0034] In addition, since the target nuclide identification model is used to perform nuclide identification on the nuclide energy spectrum to be identified to obtain the corresponding target identification result, thereby identifying the nuclide corresponding to the nuclide energy spectrum to be identified, the technical solution of the present application can more efficiently perform nuclide identification on the nuclide energy spectrum to be identified, thereby more efficiently identifying the nuclide corresponding to the nuclide energy spectrum to be identified.

[0035] It can be seen that the technical solution of the present application can efficiently and accurately identify the nuclides in the nuclide energy spectrum to be identified, regardless of whether the nuclide energy spectrum to be identified is a single nuclide energy spectrum or a mixed nuclide energy spectrum, thereby efficiently and accurately identifying the nuclide corresponding to the nuclide energy spectrum to be identified.

[0036] The above summary is for illustrative purposes only and is not intended to be limiting in any way. In addition to the illustrative aspects, embodiments and features described above, further aspects, embodiments and features of the present application will be readily apparent by reference to the accompanying drawings and the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In the accompanying drawings, unless otherwise specified, the same reference numerals throughout the multiple drawings represent the same or similar parts or elements. These drawings are not necessarily drawn to scale. It should be understood that these drawings only depict some embodiments disclosed in the present application and should not be regarded as limiting the scope of the present application.

[0038] Figure 1 A flowchart of a nuclide identification method provided in an embodiment of the present application is shown;

[0039] Figure 2 A schematic diagram of a nuclide identification device provided in an embodiment of the present application is shown; and

[0040] Figure 3 A block diagram of an electronic device used to implement an embodiment of the present application is shown. DETAILED DESCRIPTION

[0041] Many specific details are described in the following description to facilitate a full understanding of the present application. However, the present application can be implemented in many other forms different from those described herein, and those skilled in the art can make similar generalizations without violating the connotation of the present application. Therefore, the present application is not limited to the specific implementation disclosed below.

[0042] Figure 1 A flowchart of a nuclide identification method 100 provided in an embodiment of the present application is shown. The method may include steps S101-S102.

[0043] In step S101, an energy spectrum of a nuclide to be identified is obtained; the energy spectrum of the nuclide to be identified is a single nuclide energy spectrum or a mixed nuclide energy spectrum.

[0044] In step S102, the target nuclide identification model is used to perform nuclide identification on the nuclide energy spectrum to be identified, and the corresponding target identification result is obtained; the target nuclide identification model is obtained based on model optimization training of the nuclide identification model to be trained; the nuclide identification model to be trained is obtained by adding an attention mechanism module to the convolutional neural network model; the attention mechanism module is used to perform spatial and channel weighted processing on the feature map output by the convolution layer of the convolutional neural network model; the target identification result is used to represent the nuclide corresponding to the nuclide energy spectrum to be identified.

[0045] In the present application, the nuclide spectrum Figure 1 Generally, it is a relationship diagram between energy and counts obtained by detecting radiation from a radiation source through a detector, or it can be a relationship diagram between energy and counts generated by computer simulation, for example, an energy spectrum diagram generated by a Monte Carlo simulation method. The nuclide energy spectrum diagram reflects the radiation characteristics of the radioactive nuclides present in the radiation source.

[0046] In the embodiments of the present application, the nuclide energy spectrum obtained by radiation detection is referred to as a real nuclide energy spectrum, and the nuclide energy spectrum generated by computer simulation is referred to as a simulated nuclide energy spectrum.

[0047] In the embodiments of the present application, a radiation source generally refers to a substance that can emit gamma rays. A radiation source is usually composed of natural or artificial radioactive nuclides, and the nuclides of the radiation source will release specific types of rays (such as gamma rays) during the decay process. Common radiation sources include cobalt-60 (Co-60) and cesium-137 (Cs-137).

[0048] In the embodiments of the present application, a single nuclide energy spectrum refers to an energy spectrum that contains the radiation characteristics of only one nuclide. A mixed nuclide energy spectrum refers to an energy spectrum formed by superimposing the radiation characteristics of two or more different nuclides, or refers to an energy spectrum formed by superimposing two or more single nuclide energy spectra.

[0049] In general, the structure of a convolutional neural network (CNN) usually consists of an input layer, a convolutional layer, an activation layer, a pooling layer, a fully connected layer, an output layer, etc. Since the attention mechanism module is used to perform spatial and channel weighted processing on the feature map output by the convolutional layer of the convolutional neural network model, the attention mechanism module is generally added after the convolutional layer.

[0050] The attention mechanism module of the present application can be a CBAM (Convolutional Block Attention Module), a BAM (Bottleneck Attention Module), a Dual Attention Network (DAN), an Attention-based Feature Fusion (AFNet), and a Focal Attention Network (FAN). However, due to the advantages of the CBAM module, such as high efficiency, lightweight design, simple integration, and strong flexibility, in most cases, the attention mechanism module is configured as a CBAM module.

[0051] When the attention mechanism module is configured as a CBAM module, the CBAM module is often added after the convolutional block of the convolutional neural network model. In one example, the convolutional neural network model contains four convolutional blocks, and different numbers of CBAM modules are added after each convolutional block, specifically: 3 CBAM modules are added after the first convolutional block, 4 CBAM modules are added after the second convolutional block, 6 CBAM modules are added after the third convolutional block, and 3 CBAM modules are added after the fourth convolutional block. Through this configuration, the CBAM module performs spatial and channel weighted processing on the output feature map of each convolutional block.

[0052] The convolution block generally includes: convolution layer, activation function layer and pooling layer.

[0053] In an embodiment of the present application, the CBAM module includes a spatial attention submodule and a channel attention submodule, and the CBAM module is obtained by connecting the spatial attention submodule and the channel attention submodule in series.

[0054] The spatial attention submodule is used to perform spatial weighting on the feature map. In one example, the structure of the spatial attention submodule may include the following parts:

[0055] The global pooling layer is used to perform global maximum pooling and global average pooling operations in the channel dimension on the input feature map of size C×L, and obtain two feature maps of size 1×L. The global pooling layer extracts the global information of each spatial position, provides a basis for subsequent weighting, and helps the network focus on the global features of different positions.

[0056] The concatenation layer is used to concatenate the global maximum pooling and global average pooling results to generate a feature map of size 2 × L. The concatenation layer combines the two pooling results to enhance the model's ability to capture spatial features, allowing the model to utilize the information of maximum pooling and average pooling at the same time.

[0057] The convolution layer is used to perform one-dimensional convolution on the concatenated feature maps to obtain a feature map of size 1×L. The convolution operation extracts spatial features through a one-dimensional convolution kernel and further adjusts and optimizes the feature representation. Specifically, the size of the convolution kernel can be set to 7, the input channel can be set to 2 (corresponding to the concatenated feature map), and the output channel is 1.

[0058] The activation function layer is used to activate the convolution result using the Sigmoid activation function to generate a spatial attention weight matrix. This matrix is ​​used to perform weighted processing on the input feature map in the spatial dimension, thereby highlighting the key areas in the feature map and suppressing irrelevant or redundant areas, so that the model can focus more on important information areas.

[0059] The channel attention submodule is used to perform channel weighting processing on the feature map. In one example, the structure of the channel attention submodule may include the following parts:

[0060] The global pooling layer is used to perform global average pooling and maximum pooling operations on the data of each channel of the feature map of size C×L to obtain two feature maps of size C×1. The global pooling layer helps the model capture the global features of the channel level by extracting the global information of each channel, providing a basis for subsequent weighting and strengthening the model's attention to important channels.

[0061] The convolution layer is used to convolve the two C×1 feature maps into feature maps of size (C / 4)×1, and then perform a nonlinear transformation on the convolution result through the ReLU activation function to obtain an activated feature map. Then, the activated feature map is restored to a feature map of C×1 dimension through the convolution operation. It should be noted that when performing convolution operations on the two C×1 feature maps respectively, it is necessary to ensure that the two convolution operations use the same convolution kernel parameters to ensure the consistency of channel weighting and avoid differences in weighting methods between different channels, thereby maintaining the effectiveness and stability of weighting and ensuring the consistency and optimization effect of feature weighting.

[0062] The addition and activation function layer is used to add the two C×1 feature maps after convolution to obtain a feature map of size C×1. Then, the Sigmoid activation function is used to activate the feature map to generate a channel attention weight matrix. This matrix represents the weight of each channel and is used to perform weighted processing on the input feature map in the channel dimension, highlighting the key channels and suppressing irrelevant channels.

[0063] In the embodiment of the present application, before optimizing the training of the nuclide identification model to be trained to obtain the target nuclide identification model, it is often necessary to first construct the nuclide identification model to be trained. When the nuclide identification model to be trained is constructed, the training of the nuclide identification model can be optimized according to the following steps to obtain the target nuclide identification model:

[0064] First, a training set and a test set are obtained. Then, the training set is used to perform multiple rounds of iterative training on the nuclide recognition model to be trained until the set iteration stop condition is met. After that, when the iteration stop condition is met, the test set is used to test the nuclide recognition model to be trained, and if the test result meets the preset standard, the nuclide recognition model to be trained is used as the target nuclide recognition model.

[0065] In the embodiment of the present application, the training set and the test set respectively include multiple sample nuclide energy spectra and the annotation results corresponding to each sample nuclide energy spectra. Among them, the multiple sample nuclide energy spectra include single nuclide energy spectra and mixed nuclide energy spectra. In the multiple sample nuclide energy spectra, configuring the single nuclide energy spectra and the mixed nuclide energy spectra at the same time can help the target nuclide recognition model learn the characteristics of different types of nuclides, so that the model can perform accurate nuclide recognition when performing nuclide recognition, regardless of whether the nuclide energy spectra to be identified are single nuclide energy spectra or mixed nuclide energy spectra.

[0066] In the embodiment of the present application, in the process of collecting sample nuclide energy spectra to construct training sets and test sets, single nuclide energy spectra corresponding to multiple nuclides are often obtained, and part or all of the single nuclide energy spectra corresponding to the multiple nuclides are added to the training set and the test set.

[0067] The sample nuclide energy spectrum that is a mixed nuclide energy spectrum among the multiple sample nuclide energy spectrums is obtained by the following steps: first, single nuclide energy spectra corresponding to the multiple nuclides are obtained. Then, single nuclide energy spectra corresponding to at least two nuclides are randomly selected from the single nuclide energy spectra corresponding to the multiple nuclides. After that, the single nuclide energy spectra corresponding to the at least two nuclides are mixed to obtain a mixed nuclide energy spectrum, and the mixed nuclide energy spectrum is used as the sample nuclide energy spectrum.

[0068] In general, in order to simplify the training process and improve the diversity of data, when randomly selecting single nuclide energy spectra of at least two nuclides, usually only single nuclide energy spectra of two nuclides are selected for mixing. That is to say, among multiple sample nuclide energy spectra, the sample nuclide energy spectrum used as the mixed nuclide energy spectrum is usually a mixture of single nuclide energy spectra corresponding to two nuclides.

[0069] It should be noted that when the sample nuclide energy spectrum serving as a mixed nuclide energy spectrum is a mixture of single nuclide energy spectra corresponding to two nuclides, in order to enable the target nuclide identification model to still efficiently and accurately identify nuclides when the nuclide energy spectrum to be identified is a mixed nuclide energy spectrum of three or more nuclides, it is necessary to ensure that the trained target nuclide identification model has a strong generalization ability.

[0070] In addition, in order to ensure that the target nuclide identification model has a strong generalization ability, the diversity of the mixed nuclide energy spectrum can be appropriately increased during the model training process. To this end, when randomly selecting single nuclide energy spectra corresponding to at least two nuclides, it is not limited to randomly selecting single nuclide energy spectra corresponding to two nuclides. In this way, the sample nuclide energy spectrum as a mixed nuclide energy spectrum can be a mixture of single nuclide energy spectra of two nuclides, or a mixture of single nuclide energy spectra of three or more nuclides. In this way, the diversity of the mixed nuclide energy spectrum can be effectively increased.

[0071] In the embodiment of the present application, multiple nuclides may refer to 48 nuclides. Among these nuclides, some nuclides can obtain their corresponding single nuclide energy spectra by radiation detection. For these nuclides, their single nuclide energy spectra can be obtained by actual radiation detection. However, due to objective conditions, other nuclides are difficult to prepare or cannot be accessed, so the corresponding single nuclide energy spectra cannot be obtained by radiation detection. In this case, it is necessary to simulate the single nuclide energy spectra of these nuclides by computer simulation, for example, by using the Monte Carlo simulation method to simulate the single nuclide energy spectra corresponding to these parts of the nuclides. That is to say, in the embodiment of the present application, the single nuclide energy spectra can be either the real nuclide energy spectra obtained by radiation detection or the simulated nuclide energy spectra generated by computer simulation.

[0072] In an embodiment of the present application, in order to improve the robustness of the target nuclide identification model, before adding the single nuclide energy spectrum and / or the mixed nuclide energy spectrum to the training set, these energy spectra can be subjected to image enhancement processing, for example, adding noise to the single nuclide energy spectrum and / or the mixed nuclide energy spectrum. Through this enhancement process, the model can better adapt to noise interference and improve its robustness in practical applications. However, it should be noted that in order to avoid the adverse effects of noise on energy spectrum data, when the noise at each data point is generated by a random function, the cumulative error of the noise on the energy spectrum count should be controlled, the unidirectional error should be effectively reduced, and the counting error of the entire energy spectrum should be ensured to eventually tend to zero. This processing method can ensure that the added noise does not destroy the statistical characteristics of the energy spectrum, thereby improving the generalization ability of the model and its adaptability to noise.

[0073] In the embodiment of the present application, in the process of multiple rounds of iterative training of the nuclide identification model to be trained, the steps of each round of iterative training are as follows: First, select the target sample nuclide energy spectrum in the training set. Then, input the target sample nuclide energy spectrum into the nuclide identification model to be trained to obtain the corresponding sample identification result. After that, calculate the target loss value between the sample identification result and the nuclide labeling result corresponding to the target sample image. Then, use the target loss value to update the internal model parameters in the nuclide identification model to be trained.

[0074] In an embodiment of the present application, the iteration stopping condition may be: the model target loss value converges, or the number of iterations has reached a set maximum number of iterations, or other set conditions.

[0075] In the embodiment of the present application, the internal model parameters of the model include: convolutional layer weights, bias parameters, and fully connected layer weights, etc.

[0076] In the embodiment of the present application, the nuclide labeling result refers to the actual nuclide corresponding to the pre-labeled sample nuclide energy spectrum.

[0077] In the embodiment of the present application, the nuclide identification model to be trained is tested using a test set, and only when the test result meets the preset standard, the nuclide identification model to be trained is used as the target nuclide identification model, which can ensure that the recognition performance of the model meets the requirements.

[0078] In one example, the test result includes the accuracy rate corresponding to the nuclide identification model to be trained.

[0079] In another example, the test result further includes the recall rate and / or the average precision mean corresponding to the nuclide identification model to be trained.

[0080] When the test result includes precision, recall and average precision, the preset standard includes precision threshold, recall threshold and average precision threshold. In this case, the test result meets the preset standard when: the precision corresponding to the nuclide identification model to be trained reaches the precision threshold, the recall corresponding to the nuclide identification model to be trained reaches the recall threshold, and the average precision corresponding to the nuclide identification model to be trained reaches the average precision threshold.

[0081] In the radionuclide identification method provided in the embodiment of the present application, the target radionuclide identification model is obtained based on model optimization training of the radionuclide identification model to be trained, and the radionuclide identification model to be trained is obtained by adding an attention mechanism module to the convolutional neural network model, and the attention mechanism module is used to perform spatial and channel weighted processing on the feature map output by the convolution layer of the convolutional neural network model.

[0082] By performing spatial weighting on the feature map, the target nuclide identification model can highlight the key areas in the energy spectrum to be identified, while suppressing irrelevant or redundant areas. In this way, the target nuclide identification model can more accurately extract the spatial features related to nuclide identification, improve the target nuclide identification model's ability to focus on the areas that are critical to nuclide identification, and reduce the interference of background noise and irrelevant information on the target nuclide identification model, so that the target nuclide identification model can better identify the characteristics of the nuclide.

[0083] At the same time, by performing channel weighting processing on the feature graph, the target nuclide identification model can highlight the most contributing feature channels while suppressing irrelevant or redundant channels. In this way, the target nuclide identification model can more accurately identify the key channel features in the nuclide energy spectrum, thereby improving the effect and accuracy of the target nuclide identification model.

[0084] Therefore, by performing spatial weighting and channel weighting processing on the feature map, the target nuclide identification model can more accurately identify the nuclide energy spectrum of the nuclide to be identified, thereby more accurately identifying the nuclide corresponding to the nuclide energy spectrum of the nuclide to be identified.

[0085] In addition, the nuclide identification method provided in the embodiments of the present application can perform nuclide identification on the nuclide energy spectrum to be identified when the nuclide energy spectrum to be identified is a single nuclide energy spectrum or a mixed nuclide energy spectrum, thereby identifying the nuclide corresponding to the nuclide energy spectrum to be identified.

[0086] In addition, since the target nuclide identification model is used to perform nuclide identification on the nuclide energy spectrum to be identified to obtain the corresponding target identification result, thereby identifying the nuclide corresponding to the nuclide energy spectrum to be identified, the nuclide identification method provided in the embodiment of the present application can more efficiently perform nuclide identification on the nuclide energy spectrum to be identified, thereby more efficiently identifying the nuclide corresponding to the nuclide energy spectrum to be identified.

[0087] It can be seen that the nuclide identification method provided in the embodiments of the present application can efficiently and accurately identify the nuclide energy spectrum to be identified, regardless of whether the nuclide energy spectrum to be identified is a single nuclide energy spectrum or a mixed nuclide energy spectrum, thereby efficiently and accurately identifying the nuclide corresponding to the nuclide energy spectrum to be identified.

[0088] Corresponding to the nuclide identification method provided in the embodiment of the present application, the embodiment of the present application also provides a nuclide identification device. Figure 2 As shown, it is a structural block diagram of a nuclide identification device 200 according to an embodiment of the present application. The device 200 may include:

[0089] The module 201 for acquiring the energy spectrum of the nuclide to be identified is used to acquire the energy spectrum of the nuclide to be identified; the energy spectrum of the nuclide to be identified is a single nuclide energy spectrum or a mixed nuclide energy spectrum;

[0090] The nuclide identification module 202 is used to use the target nuclide identification model to perform nuclide identification on the nuclide energy spectrum to be identified, and obtain the corresponding target identification result; the target nuclide identification model is obtained based on the model optimization training of the nuclide identification model to be trained; the nuclide identification model to be trained is obtained by adding an attention mechanism module to the convolutional neural network model; the attention mechanism module is used to perform spatial and channel weighted processing on the feature map output by the convolution layer of the convolutional neural network model; the target identification result is used to represent the nuclide corresponding to the nuclide energy spectrum to be identified.

[0091] In one possible implementation, the attention mechanism module is added after the convolutional layer.

[0092] In a possible implementation, the steps of performing model optimization training on the nuclide identification model to be trained to obtain the target nuclide identification model are as follows:

[0093] A training set and a test set are obtained; the training set and the test set respectively include multiple sample nuclide energy spectra and the annotation results corresponding to each sample nuclide energy spectra; the multiple sample nuclide energy spectra include single nuclide energy spectra and mixed nuclide energy spectra;

[0094] The training set is used to perform multiple rounds of iterative training on the nuclide recognition model to be trained until the set iteration stop condition is met;

[0095] When the iteration stop condition is met, the nuclide identification model to be trained is tested using the test set, and when the test result meets the preset standard, the nuclide identification model to be trained is used as the target nuclide identification model.

[0096] In one possible implementation, the steps of each round of iterative training are as follows:

[0097] Select the target sample nuclide energy spectrum in the training set;

[0098] Input the target sample nuclide energy spectrum into the nuclide identification model to be trained to obtain the corresponding sample identification result;

[0099] Calculate the target loss value between the sample recognition result and the nuclide labeling result corresponding to the target sample image;

[0100] The target loss value is used to update the internal model parameters in the nuclide identification model to be trained.

[0101] In a possible implementation, the test result includes the accuracy rate corresponding to the nuclide identification model to be trained.

[0102] In a possible implementation, the module 201 for acquiring the energy spectrum of the nuclide to be identified is specifically used to obtain single nuclide energy spectra corresponding to a plurality of nuclides; among the single nuclide energy spectra corresponding to a plurality of nuclides, single nuclide energy spectra corresponding to at least two nuclides are randomly selected; the single nuclide energy spectra corresponding to at least two nuclides are mixed to obtain a mixed nuclide energy spectrum, and the mixed nuclide energy spectrum is used as a sample nuclide energy spectrum.

[0103] In a possible implementation, the single nuclide energy spectrum is a real nuclide energy spectrum obtained through radiation detection, or a simulated nuclide energy spectrum generated through computer simulation.

[0104] In one possible implementation, the attention mechanism module is a convolutional block attention module.

[0105] The functions of each module in each device in the embodiments of the present application can be found in the corresponding description in the above method, and have corresponding beneficial effects, which will not be repeated here.

[0106] Figure 3FIG. 1 is a block diagram of an electronic device used to implement an embodiment of the present application. Figure 3 As shown, the electronic device includes: a memory 301 and a processor 302. The memory 301 stores a computer program that can be run on the processor 302. When the processor 302 executes the computer program, the method in the above embodiment is implemented. The number of the memory 301 and the processor 302 can be one or more.

[0107] The electronic device also includes:

[0108] The communication interface 303 is used to communicate with external devices and perform data exchange transmission.

[0109] If the memory 301, the processor 302 and the communication interface 303 are implemented independently, the memory 301, the processor 302 and the communication interface 303 can be connected to each other through a bus and communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 3 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0110] Optionally, in a specific implementation, if the memory 301, the processor 302 and the communication interface 303 are integrated on a chip, the memory 301, the processor 302 and the communication interface 303 can communicate with each other through an internal interface.

[0111] An embodiment of the present application provides a computer-readable storage medium storing a computer program, which implements the method provided in the embodiment of the present application when the program is executed by a processor.

[0112] An embodiment of the present application also provides a chip, which includes a processor for calling and executing instructions stored in the memory from the memory, so that a communication device equipped with the chip executes the method provided by the embodiment of the present application.

[0113] An embodiment of the present application also provides a chip, including: an input interface, an output interface, a processor and a memory, wherein the input interface, the output interface, the processor and the memory are connected via an internal connection path, and the processor is used to execute the code in the memory. When the code is executed, the processor is used to execute the method provided in the embodiment of the application.

[0114] It should be understood that the processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc. It is worth noting that the processor may be a processor supporting the Advanced RISC Machines (ARM) architecture.

[0115] Further, optionally, the above-mentioned memory may include a read-only memory and a random access memory, and may also include a non-volatile random access memory. The memory may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may include a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may include a random access memory (RAM), which is used as an external cache. By way of example but not limitation, many forms of RAM are available. For example, static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM) and direct memory bus random access memory (DR RAM).

[0116] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it 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. When the computer program instructions are loaded and executed on a computer, the process or function according to the present application is 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 computer-readable storage medium.

[0117] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine different embodiments or examples described in this specification and the features of different embodiments or examples, unless they are contradictory.

[0118] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0119] Any process or method description in the flow chart or otherwise described herein can be understood to represent a module, fragment or portion of a code including one or more executable instructions for implementing the steps of a specific logical function or process. And the scope of the preferred embodiment of the present application includes other implementations, in which the functions may not be performed in the order shown or discussed, including in a substantially simultaneous manner or in a reverse order according to the functions involved.

[0120] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, which can be embodied in any computer-readable medium for use by an instruction execution system, apparatus or device (such as a computer-based system, a system including a processor or other system that can fetch instructions from an instruction execution system, apparatus or device and execute instructions), or used in combination with these instruction execution systems, apparatuses or devices.

[0121] It should be understood that the various parts of the present application can be implemented with hardware, software, firmware or a combination thereof. In the above embodiments, multiple steps or methods can be implemented with software or firmware stored in a memory and executed by a suitable instruction execution system. All or part of the steps of the above embodiment method can be completed by instructing the relevant hardware through a program, which can be stored in a computer-readable storage medium, and when the program is executed, it includes one of the steps of the method embodiment or a combination thereof.

[0122] In addition, each functional unit in each embodiment of the present application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into one module. The above-mentioned integrated module can be implemented in the form of hardware or in the form of a software functional module. If the above-mentioned integrated module is implemented in the form of a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. The storage medium can be a read-only memory, a disk or an optical disk, etc.

[0123] The above are only specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of various changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for nuclide identification, comprising: Obtain the energy spectrum of the nuclide to be identified; The nuclide energy spectrum to be identified is a single nuclide energy spectrum or a mixed nuclide energy spectrum; Using the target nuclide identification model to perform nuclide identification on the nuclide energy spectrum to be identified, and obtain a corresponding target identification result; The target nuclide identification model is obtained based on model optimization training of the nuclide identification model to be trained; The nuclide identification model to be trained is obtained by adding an attention mechanism module to the convolutional neural network model; the attention mechanism module is used to perform spatial and channel weighted processing on the feature map output by the convolution layer of the convolutional neural network model; the target recognition result is used to represent the nuclide corresponding to the nuclide energy spectrum to be identified.

2. The method according to claim 1, wherein: The attention mechanism module is added after the convolutional layer.

3. The method according to claim 1, wherein: The steps of performing model optimization training on the nuclide identification model to be trained to obtain the target nuclide identification model are as follows: Obtaining a training set and a test set; the training set and the test set respectively include a plurality of sample nuclide energy spectra and annotation results corresponding to each sample nuclide energy spectra; The plurality of sample nuclide energy spectra include the single nuclide energy spectrum and the mixed nuclide energy spectrum; Performing multiple rounds of iterative training on the nuclide identification model to be trained using the training set until a set iteration stop condition is met; When the iteration stop condition is met, the nuclide identification model to be trained is tested using the test set, and when the test result meets the preset standard, the nuclide identification model to be trained is used as the target nuclide identification model.

4. The method according to claim 3, wherein: The steps for each round of iterative training are as follows: Selecting a target sample nuclide energy spectrum in the training set; Inputting the target sample nuclide energy spectrum into the nuclide identification model to be trained to obtain a corresponding sample identification result; Calculating a target loss value between the sample recognition result and the nuclide labeling result corresponding to the target sample image; The target loss value is used to update internal model parameters in the nuclide identification model to be trained.

5. The method according to claim 3, wherein: The test result includes the accuracy rate corresponding to the nuclide identification model to be trained.

6. The method according to claim 3, wherein the sample nuclide energy spectrum of the mixed nuclide energy spectrum among the plurality of sample nuclide energy spectra is obtained by the following steps: Obtain single nuclide energy spectra corresponding to multiple nuclides; Randomly selecting single nuclide energy spectra corresponding to at least two nuclides from the single nuclide energy spectra corresponding to the multiple nuclides; The single nuclide energy spectra corresponding to the at least two nuclides are mixed to obtain the mixed nuclide energy spectrum, and the mixed nuclide energy spectrum is used as the sample nuclide energy spectrum.

7. The method according to claim 1, wherein: The single nuclide energy spectrum is a real nuclide energy spectrum obtained through radiation detection, or a simulated nuclide energy spectrum generated through computer simulation.

8. The method according to claim 1, wherein: The attention mechanism module is a convolutional block attention module.

9. An electronic device comprising a memory, a processor and a computer program stored in the memory, wherein the processor implements the method according to any one of claims 1 to 8 when executing the computer program.

10. A computer-readable storage medium, wherein a computer program is stored in the computer-readable storage medium, and when the computer program is executed by a processor, the method according to any one of claims 1 to 8 is implemented.

Images