A target detection method and system based on data fusion

Abstract

The application discloses a target detection method and system based on data fusion, which comprises the following steps: adopting all intelligent agents to perform target detection on historical point cloud data sets, and constructing a graph matching matrix based on the similarity features between all target detection results obtained; obtaining initial vertex features based on the graph matching matrix; processing the edge weight matrix of the graph matching matrix, the degree matrix of the graph matching matrix and the initial vertex features by using a first graph convolution network to obtain updated vertex features; performing matching analysis on each updated vertex feature and the remaining updated vertex features by using a second graph convolution network to obtain a double random matrix; constructing a loss function, and training a graph convolution neural network model by using the loss function; and processing target detection results of target intelligent agents by using the trained model, and fusing the target detection results. The method provided in the application embodiment improves the accuracy of target detection.

Description

Technical Field

[0001] This invention relates to the field of detection technology, and in particular to a target detection method and system based on data fusion. Background Technology

[0002] Object detection, a cornerstone of computer vision, aims to identify specific targets in images or videos and determine their location. It is an important technology in the field of computer vision, giving machines the ability to recognize and understand image content, and playing a vital role in fields such as security monitoring, autonomous driving, and smart retail.

[0003] However, the acquired images or videos may be affected by various factors, resulting in low image quality. Furthermore, the target to be detected may be obscured by other objects in the acquired images, leading to low accuracy in target detection.

[0004] Therefore, improving the accuracy of target detection has become a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] This invention provides a target detection method and system based on data fusion to solve the technical problem of low accuracy in target detection, thereby improving the accuracy of target detection.

[0006] To address the aforementioned technical problems, embodiments of the present invention provide a target detection method based on data fusion, comprising:

[0007] All agents are used to perform object detection on the historical point cloud dataset, and a graph matching matrix is ​​constructed based on the similarity features among all the obtained object detection results;

[0008] Initial vertex features are obtained based on the graph matching matrix; the edge weight matrix of the graph matching matrix, the degree matrix of the graph matching matrix, and the initial vertex features are processed by a first graph convolutional network to obtain updated vertex features;

[0009] A second graph convolutional network is used to perform matching analysis on each updated vertex feature and the remaining updated vertex features to obtain a non-negative fraction matrix; the non-negative fraction matrix is ​​normalized to obtain a double random matrix; a loss function is constructed based on the difference features between the double random matrix and the true matching matrix, and the graph convolutional neural network model is trained using the loss function;

[0010] The trained graph convolutional neural network model is used to process all real-time target detection results of the target agent to obtain the detection results of each associated target agent. All the associated target detection results are then weighted and fused to obtain the fused target detection result.

[0011] As one preferred embodiment, the construction of a graph matching matrix based on the similarity features among all obtained target detection results includes:

[0012] Each target detection result is treated as a node, each node is connected to the other nodes, and the line between any two nodes is treated as an edge.

[0013] A first feature vector for each node is constructed based on the label, detection box shape, and position of the center point of the detection box for each target detection result; a second feature vector for each edge is constructed based on the length and angle of each edge and the label of each corresponding node.

[0014] The comprehensive node similarity between the corresponding nodes is calculated based on the similarity features between the first feature vectors of the two nodes, and the comprehensive edge similarity between the corresponding edges is calculated based on the similarity features between the second feature vectors of the two edges.

[0015] The comprehensive node similarity is used as the element on the diagonal of the graph matching matrix, and the comprehensive edge similarity is used as the element on the off-diagonal of the graph matching matrix.

[0016] As one preferred embodiment, the step of calculating the comprehensive node similarity between corresponding nodes based on the similarity features between the first feature vectors of the two nodes includes:

[0017] The similarity of the detection box shapes of the two nodes is used as the first node similarity. The Euclidean distance between the center points of the detection boxes of the two nodes is inversely proportional to obtain the second node similarity. The similarity of the labels of the two nodes is used as the third node similarity. Based on the linear relationship between the first node similarity, the second node similarity and the third node similarity, the node similarity of the two nodes is calculated.

[0018] The step of calculating the comprehensive edge similarity between corresponding edges based on the similarity features between the second feature vectors of the two edges includes:

[0019] The similarity of the labels of the starting nodes of the two edges is used as the first edge point similarity, and the similarity of the labels of the ending nodes of the two edges is used as the second edge point similarity. The third edge similarity is obtained based on the difference in the length of the two edges. The fourth edge similarity is obtained based on the difference in the angle of the two edges. The comprehensive edge similarity of the two nodes is calculated according to the linear relationship between the first edge point similarity, the second edge point similarity, the third edge similarity, and the fourth edge similarity.

[0020] As one preferred embodiment, obtaining the features of each initial vertex based on the graph matching matrix includes:

[0021] The elements at all diagonal positions of the graph matching matrix are used as the initial vertex features.

[0022] As one preferred embodiment, the first graph convolutional network includes two fully connected layers, and the activation function of the first graph convolutional network is the ReLU activation function;

[0023] The step of using a first graph convolutional network to process the edge weight matrix of the graph matching matrix, the degree matrix of the graph matching matrix, and the initial vertex features to obtain updated vertex features includes:

[0024] m (k) =D -1 Wf1(v (k-1) )+f1(v (k-1) )

[0025] Where, m (k) For the k-th updated vertex feature, D -1 Let W be the inverse of the degree matrix of the graph matching matrix, W be the edge weight matrix of the graph matching matrix, f1 be the fully connected layer of the first graph convolutional network, and v be the inverse of the degree matrix of the graph matching matrix. (k-1) This is the feature of the (k-1)th initial vertex.

[0026] As one preferred embodiment, the second graph convolutional network includes a fully connected layer, and the activation function of the second graph convolutional network is an exponential activation function.

[0027] As one preferred embodiment, the sum of the elements in each row of the double random matrix is ​​1, and the sum of the elements in each column of the double random matrix is ​​1.

[0028] As one preferred embodiment, the step of constructing a loss function based on the difference features between the double random matrix and the true matching matrix includes:

[0029]

[0030] Where L is the loss function, np To update the number of vertex features, log is the logarithmic function, P i,j Let X be the value of the element in the i-th row and j-th column of a double random matrix. i,j Let be the value of the element in the i-th row and j-th column of the true matching matrix.

[0031] As one preferred embodiment, the weighted fusion of all the associated target detection results to obtain the fused target detection result includes:

[0032] The classification confidence of each detection box of the target agent is used as the corresponding weight, and the associated target detection results of each detection box of the target agent are weighted and summed to obtain the fused target detection result.

[0033] Another embodiment of the present invention provides a target detection system based on data fusion, comprising:

[0034] A processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor, when executing the computer program, implements a data fusion-based target detection method as described in any of the preceding claims.

[0035] Compared with the prior art, the beneficial effects of the embodiments of the present invention are at least one of the following:

[0036] (1) All agents are used to perform target detection on the historical point cloud dataset. A graph matching matrix is ​​constructed based on the similarity features between all the target detection results. The graph matching matrix records the similarity between multiple detection results of different agents, thus preserving the most original correlation between the detection results.

[0037] (2) Based on the graph matching matrix, the initial vertex features are obtained. Since the initial features only record the original node information, when there is a position error in the detection result, the node similarity will decrease, which will weaken the correlation between different detection results of the same target. Therefore, in order to reduce the negative impact of the error, edge similarity information is introduced to correct the vertex features. The first graph convolutional network is used to process the edge weight matrix of the graph matching matrix, the degree matrix of the graph matching matrix and the initial vertex features to obtain updated vertex features. The features of adjacent vertices are learned through the graph convolutional network to update the features of itself, which improves the accuracy of subsequent correlation fusion analysis and enhances the robustness of target detection in complex environments, thereby improving the accuracy of target detection.

[0038] (3) A second graph convolutional network is used to perform matching analysis on the features of each updated vertex and the features of the remaining updated vertices to obtain a non-negative fraction matrix. The target association is transformed into a vertex classification problem and solved using a graph convolutional network. The non-negative fraction matrix is ​​normalized to obtain a double random matrix to satisfy the one-to-one constraint of the graph matching problem. The elements in the double random matrix represent the matching probability between nodes. A loss function is constructed based on the difference features between the double random matrix and the real matching matrix. The graph convolutional neural network model is trained with the loss function. Gradient descent is performed through backpropagation algorithm to optimize the graph matching result. By minimizing the loss function, the model learns how to generate a prediction probability matrix that is more consistent with the real matching label, thereby gradually improving the matching performance and ensuring the robustness of target detection in complex environments.

[0039] (4) The trained graph convolutional neural network model is used to process all real-time target detection results of the target agent to obtain the target agent's associated target detection results. All associated target detection results are weighted and fused to obtain fused target detection results. The results of multiple detection boxes are combined through the weighted fusion strategy, which further improves the accuracy of target detection. Attached Figure Description

[0040] Figure 1 This is a schematic diagram of the target detection method based on data fusion in one embodiment of the present invention;

[0041] Figure 2 This is a schematic diagram illustrating the acquisition of the loss function in one embodiment of the present invention. Detailed Implementation

[0042] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0043] In the description of this application, the terms "first," "second," "third," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined with "first," "second," "third," etc., may explicitly or implicitly include one or more of that feature. In the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0044] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections or electrical connections; they can refer to direct connections or indirect connections through an intermediate medium; and they can refer to the internal communication between two components. The terms "vertical," "horizontal," "left," "right," "upper," "lower," and similar expressions used herein are for illustrative purposes only and do not indicate or imply that the device or component referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as limiting the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0045] In the description of this application, it should be noted that, unless otherwise defined, all technical and scientific terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in this specification is for the purpose of describing specific embodiments only and is not intended to limit the invention. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0046] One embodiment of the present invention provides a target detection method based on data fusion. For details, please refer to [link to relevant documentation]. Figure 1 , Figure 1 The diagram shown is a schematic flowchart of a target detection method based on data fusion in one embodiment of the present invention, including:

[0047] S1: Use all agents to perform object detection on the historical point cloud dataset, and construct a graph matching matrix based on the similarity features between all the obtained object detection results;

[0048] S2: Obtain the initial vertex features based on the graph matching matrix; use a first graph convolutional network to process the edge weight matrix of the graph matching matrix, the degree matrix of the graph matching matrix, and the initial vertex features to obtain updated vertex features;

[0049] S3: A second graph convolutional network is used to perform matching analysis on each updated vertex feature and the remaining updated vertex features to obtain a non-negative fraction matrix; the non-negative fraction matrix is ​​normalized to obtain a double random matrix; a loss function is constructed based on the difference features between the double random matrix and the true matching matrix, and the graph convolutional neural network model is trained using the loss function;

[0050] S4: The trained graph convolutional neural network model is used to process all real-time target detection results of the target agent to obtain the detection results of each associated target agent. All associated target detection results are weighted and fused to obtain the fused target detection result.

[0051] In step S1, a graph matching matrix is ​​constructed based on the similarity features among all the obtained target detection results. Specifically, this includes treating each target detection result as a node, connecting each node to the other nodes, and treating the connection between any two nodes as an edge.

[0052] The first feature vector of each node is constructed based on the label, bounding box shape, and center point position of each target detection result. The second feature vector of each edge is constructed based on the length and angle of each edge and the corresponding label of each node.

[0053] The comprehensive node similarity between the corresponding nodes is calculated based on the similarity features between the first feature vectors of the two nodes, and the comprehensive edge similarity between the corresponding edges is calculated based on the similarity features between the second feature vectors of the two edges.

[0054] The overall node similarity is used as the element on the diagonal of the graph matching matrix, and the overall edge similarity is used as the element on the off-diagonal of the graph matching matrix.

[0055] The calculation of the comprehensive node similarity between the corresponding nodes based on the similarity features between the first feature vectors of the two nodes includes: taking the similarity of the shape of the detection boxes of the two nodes as the first node similarity; performing an inverse proportional processing on the Euclidean distance between the positions of the center points of the detection boxes of the two nodes to obtain the second node similarity; taking the similarity of the labels of the two nodes as the third node similarity; and calculating the node similarity between the two nodes based on the linear relationship between the first node similarity, the second node similarity, and the third node similarity.

[0056] The comprehensive edge similarity between corresponding edges is calculated based on the similarity features between the second feature vectors of the two edges. This includes: using the similarity of the labels of the starting nodes of the two edges as the first edge point similarity, using the similarity of the labels of the ending nodes of the two edges as the second edge point similarity, obtaining the third edge similarity based on the difference in the length of the two edges, obtaining the fourth edge similarity based on the difference in the angle of the two edges, and calculating the comprehensive edge similarity between the two nodes based on the linear relationship between the first edge point similarity, the second edge point similarity, the third edge similarity, and the fourth edge similarity.

[0057] Specifically, each node represents a target detected by the agent. Node features include the target's location coordinates, the shape of the detection box, the detection box skew angle, the predicted category, and the predicted probability distribution. All nodes are connected in a fully connected manner to construct edges between nodes. Edge features include the relative distance and angle between nodes, as well as the label information of the start and end points.

[0058] Detection results with the same label, similar bounding box shapes, and close proximity will have high node similarity. Based on this, node similarity and edge similarity are calculated, and the calculated node and edge similarity information is integrated into the graph matching matrix.

[0059] The elements at all diagonal positions of the graph matching matrix are used as initial vertex features. The first graph convolutional network is used to process the edge weight matrix, degree matrix, and initial vertex features of the graph matching matrix to obtain updated vertex features.

[0060] Specifically, the first convolutional network consists of two fully connected layers, and the activation function of the first convolutional network is the ReLU activation function.

[0061] The first graph convolutional network is used to process the edge weight matrix, the degree matrix, and the initial vertex features of the graph matching matrix to obtain updated vertex features, including:

[0062] m (k) =D -1 Wf1(v (k-1) )+f1(v (k-1) )

[0063] Where, m (k) For the k-th updated vertex feature, D -1 Let W be the inverse of the degree matrix of the graph matching matrix, W be the edge weight matrix of the graph matching matrix, f1 be the fully connected layer of the first graph convolutional network, and v be the inverse of the degree matrix of the graph matching matrix. (k-1) This is the feature of the (k-1)th initial vertex.

[0064] A second graph convolutional network is used to perform matching analysis on the features of each updated vertex and the features of the remaining updated vertices to obtain a non-negative fraction matrix. The non-negative fraction matrix is ​​then normalized to obtain a double random matrix. A loss function is constructed based on the difference between the double random matrix and the true matching matrix, and the graph convolutional neural network model is trained using the loss function.

[0065] The second convolutional network includes a fully connected layer, and its activation function is an exponential activation function.

[0066] Non-negative fractional matrices typically represent some form of similarity or association, where each element is non-negative. To transform them into double-random matrices, we need to perform normalization in the following two steps:

[0067] Normalize each row of the matrix so that the sum of the elements in each row is 1. The row-normalized matrix represents the relative correlation between each node and other nodes.

[0068] Based on row normalization, each column of the matrix is ​​then normalized so that the sum of the elements in each column is also 1. Column normalization ensures that the distribution of the correlation degree originating from each node is also uniform.

[0069] After these two normalization steps, the original non-negative fractional matrix is ​​transformed into a double random matrix.

[0070] The differences between the double random matrix and the true matching matrix form the basis for constructing the loss function. The true matching matrix is ​​typically a binary matrix that represents the actual matching relationships between nodes (e.g., 1 indicates a match, and 0 indicates a mismatch).

[0071] Specifically, it includes:

[0072]

[0073] Where L is the loss function, n p To update the number of vertex features, log is the logarithmic function, P i,j Let X be the value of the element in the i-th row and j-th column of a double random matrix. i,j Let be the value of the element in the i-th row and j-th column of the true matching matrix.

[0074] In step S4, the trained graph convolutional neural network model is used to process all real-time target detection results of the target agent to obtain the detection results of each associated target agent. The weighted fusion of all associated target detection results is then performed to obtain the fused target detection result. The specific steps include: using the classification confidence of each detection box of the target agent as the corresponding weight, and weighted summing of the associated target detection results of each detection box of the target agent to obtain the fused target detection result.

[0075] Another embodiment of the present invention provides a target detection system based on data fusion. For details, please refer to [link to relevant documentation]. Figure 2 , Figure 2 The diagram illustrates the acquisition of a loss function in one embodiment of the present invention, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements a target detection method based on data fusion as described in any of the preceding claims.

[0076] Compared with the prior art, the beneficial effects of the embodiments of the present invention are at least one of the following:

[0077] (1) All agents are used to perform target detection on the historical point cloud dataset. A graph matching matrix is ​​constructed based on the similarity features between all the target detection results. The graph matching matrix records the similarity between multiple detection results of different agents, thus preserving the most original correlation between the detection results.

[0078] (2) Based on the graph matching matrix, the initial vertex features are obtained. Since the initial features only record the original node information, when there is a position error in the detection result, the node similarity will decrease, which will weaken the correlation between different detection results of the same target. Therefore, in order to reduce the negative impact of the error, edge similarity information is introduced to correct the vertex features. The first graph convolutional network is used to process the edge weight matrix of the graph matching matrix, the degree matrix of the graph matching matrix and the initial vertex features to obtain updated vertex features. The features of adjacent vertices are learned through the graph convolutional network to update the features of itself, which improves the accuracy of subsequent correlation fusion analysis and enhances the robustness of target detection in complex environments, thereby improving the accuracy of target detection.

[0079] (3) A second graph convolutional network is used to perform matching analysis on the features of each updated vertex and the features of the remaining updated vertices to obtain a non-negative fraction matrix. The target association is transformed into a vertex classification problem and solved using a graph convolutional network. The non-negative fraction matrix is ​​normalized to obtain a double random matrix to satisfy the one-to-one constraint of the graph matching problem. The elements in the double random matrix represent the matching probability between nodes. A loss function is constructed based on the difference features between the double random matrix and the real matching matrix. The graph convolutional neural network model is trained with the loss function. Gradient descent is performed through backpropagation algorithm to optimize the graph matching result. By minimizing the loss function, the model learns how to generate a prediction probability matrix that is more consistent with the real matching label, thereby gradually improving the matching performance and ensuring the robustness of target detection in complex environments.

[0080] (4) The trained graph convolutional neural network model is used to process all real-time target detection results of the target agent to obtain the target agent's associated target detection results. All associated target detection results are weighted and fused to obtain fused target detection results. The results of multiple detection boxes are combined through the weighted fusion strategy, which further improves the accuracy of target detection.

[0081] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A target detection method based on data fusion, characterized in that, include: All agents are used to perform object detection on the historical point cloud dataset, and a graph matching matrix is ​​constructed based on the similarity features among all the obtained object detection results; Initial vertex features are obtained based on the graph matching matrix; the edge weight matrix of the graph matching matrix, the degree matrix of the graph matching matrix, and the initial vertex features are processed by a first graph convolutional network to obtain updated vertex features; A second graph convolutional network is used to perform matching analysis on each updated vertex feature and the remaining updated vertex features to obtain a non-negative fraction matrix; the non-negative fraction matrix is ​​then normalized to obtain a double random matrix. A loss function is constructed based on the difference features between the double random matrix and the true matching matrix, and the graph convolutional neural network model is trained using the loss function; The trained graph convolutional neural network model is used to process all real-time target detection results of the target agent to obtain the detection results of each associated target agent. All the associated target detection results are then weighted and fused to obtain the fused target detection result.

2. The target detection method based on data fusion according to claim 1, characterized in that, The construction of a graph matching matrix based on the similarity features among all obtained target detection results includes: Each target detection result is treated as a node, each node is connected to the other nodes, and the line between any two nodes is treated as an edge. A first feature vector for each node is constructed based on the label, detection box shape, and position of the center point of the detection box for each target detection result; a second feature vector for each edge is constructed based on the length and angle of each edge and the label of each corresponding node. The comprehensive node similarity between the corresponding nodes is calculated based on the similarity features between the first feature vectors of the two nodes, and the comprehensive edge similarity between the corresponding edges is calculated based on the similarity features between the second feature vectors of the two edges. The comprehensive node similarity is used as the element on the diagonal of the graph matching matrix, and the comprehensive edge similarity is used as the element on the off-diagonal of the graph matching matrix.

3. The target detection method based on data fusion according to claim 2, characterized in that, The step of calculating the comprehensive node similarity between corresponding nodes based on the similarity features between the first feature vectors of the two nodes includes: The similarity of the detection box shapes of the two nodes is used as the first node similarity. The Euclidean distance between the center points of the detection boxes of the two nodes is inversely proportional to obtain the second node similarity. The similarity of the labels of the two nodes is used as the third node similarity. Based on the linear relationship between the first node similarity, the second node similarity and the third node similarity, the node similarity of the two nodes is calculated. The step of calculating the comprehensive edge similarity between corresponding edges based on the similarity features between the second feature vectors of the two edges includes: The similarity of the labels of the starting nodes of the two edges is used as the first edge point similarity, and the similarity of the labels of the ending nodes of the two edges is used as the second edge point similarity. The third edge similarity is obtained based on the difference in the length of the two edges. The fourth edge similarity is obtained based on the difference in the angle of the two edges. The comprehensive edge similarity of the two nodes is calculated according to the linear relationship between the first edge point similarity, the second edge point similarity, the third edge similarity, and the fourth edge similarity.

4. The target detection method based on data fusion according to claim 2, characterized in that, The step of obtaining the features of each initial vertex based on the graph matching matrix includes: The elements at all diagonal positions of the graph matching matrix are used as the initial vertex features.

5. The target detection method based on data fusion according to claim 1, characterized in that, The first graph convolutional network includes two fully connected layers, and the activation function of the first graph convolutional network is the ReLU activation function; The step of using a first graph convolutional network to process the edge weight matrix of the graph matching matrix, the degree matrix of the graph matching matrix, and the initial vertex features to obtain updated vertex features includes: m (k) =D -1 Wf1(v (k-1) )+f1(v (k-1) ) Where, m (k) For the k-th updated vertex feature, D -1 Let W be the inverse of the degree matrix of the graph matching matrix, W be the edge weight matrix of the graph matching matrix, f1 be the fully connected layer of the first graph convolutional network, and v be the inverse of the degree matrix of the graph matching matrix. (k-1) This is the feature of the (k-1)th initial vertex.

6. The target detection method based on data fusion according to claim 1, characterized in that, The second graph convolutional network includes a fully connected layer, and the activation function of the second graph convolutional network is an exponential activation function.

7. The target detection method based on data fusion according to claim 1, characterized in that, The sum of the elements in each row of the double random matrix is ​​1, and the sum of the elements in each column of the double random matrix is ​​1.

8. The target detection method based on data fusion according to claim 1, characterized in that, The loss function is constructed based on the difference features between the double random matrix and the true matching matrix. include: Where L is the loss function, n p To update the number of vertex features, log is the logarithmic function, P i,j Let X be the value of the element in the i-th row and j-th column of a double random matrix. i,j Let be the value of the element in the i-th row and j-th column of the true matching matrix.

9. The target detection method based on data fusion according to claim 1, characterized in that, The weighted fusion of all the associated target detection results to obtain the fused target detection result includes: The classification confidence of each detection box of the target agent is used as the corresponding weight, and the associated target detection results of each detection box of the target agent are weighted and summed to obtain the fused target detection result.

10. A target detection system based on data fusion, characterized in that, The system includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to implement a data fusion-based target detection method as described in any one of claims 1 to 9.

Images