Intelligent target tracking data analysis system and method based on DSP

Abstract

The invention discloses an intelligent target tracking data analysis system and method based on a DSP, and belongs to the technical field of target tracking. Image data are collected, visual feature extraction is carried out on an image to generate a visual feature vector, motion feature extraction is carried out on the image to generate a motion feature vector; recognizing an interference object, generating a visual feature vector and a motion feature vector of the interference object, constructing a target similar interference model, and importing the visual feature vectors and the motion feature vectors of the target tracking object and the interference object into the target similar interference model to evaluate the similarity between the target tracking object and the interference object. The method comprises the following steps: acquiring image environment data, constructing an environment interference model, importing the acquired image environment data into the environment interference model to evaluate the environment interference degree, constructing a target tracking effect evaluation model, importing target similarity interference and environment interference into the target tracking effect evaluation model, and evaluating the target tracking effect in combination with the continuity and stability of target tracking. And the target tracking effect evaluation capability is improved.

Description

Technical Field

[0001] This application belongs to the technical field of target tracking, specifically an intelligent target tracking data analysis system and method based on DSP. Background Art

[0002] With the rapid development of computer vision technology, target tracking has become one of the core technologies in systems such as automated monitoring, intelligent transportation, and driverless driving. Most of the existing target tracking methods rely on single visual features. When dealing with complex interferences in a dynamic environment, it is difficult to ensure stable and high-precision tracking effects, and it is easy to have tracking drift or lose the target. Most systems only focus on whether the tracking is successful and lack quantitative analysis means for factors such as continuity, stability, and interference effects during the tracking process. The prior art does not perform quantitative modeling on environmental changes and the feature intensity of interfering objects, resulting in the tracking evaluation being unable to accurately reflect the actual performance in an interference environment.

[0003] This application provides an intelligent target tracking data analysis system and method based on DSP, which can effectively analyze the target tracking effect in a complex interference environment, comprehensively consider the visual and motion feature similarities of interfering objects, as well as environmental factors such as sudden light changes, occlusion, and scene complexity. By feature extraction, interference modeling, and tracking stability analysis, a tracking performance evaluation mechanism is constructed, which not only improves the anti-interference ability and tracking stability of the system in a changing scene, but also provides data support for subsequent tracking optimization and strategy adjustment, and can work stably in complex actual application scenarios. Summary of the Invention

[0004] Aiming at the deficiencies of the prior art, this application proposes an intelligent target tracking data analysis system and method based on DSP.

[0005] To achieve the above object, this application provides the following technical solutions: An intelligent target tracking data analysis system and method based on DSP, which includes the following specific steps: Collect image data, extract visual features from the image and generate visual feature vectors; Extract motion features from the image to generate motion feature vectors, identify interfering objects and generate visual feature vectors and motion feature vectors of the interfering objects; Construct a target similarity interference model, and import the visual feature vectors and motion feature vectors of the target tracking object and the interfering object into the target similarity interference model to evaluate the similarity degree between the target tracking object and the interfering object; Construct an environmental interference model, and import the collected image environmental data into the environmental interference model to evaluate the environmental interference degree; Build a target tracking effect evaluation model, import target similarity interference and environmental interference into the target tracking effect evaluation model, and evaluate the target tracking effect in combination with the continuity and stability of target tracking.

[0006] Preferably, the steps of collecting image data, extracting visual features from the image and generating a visual feature vector are as follows: S11. Collect image data through a high-definition camera and other sensors, preprocess the collected original image data, including operations such as image denoising, normalization, and enhancement, perform time synchronization and spatial registration on multi-sensor data to ensure the consistency of different sensor data in time and space; S12. Convert the image to the HSV color space, construct a normalized color histogram for the target area, each channel is divided into a fixed number of intervals, count the number of pixels in each interval, and splice them into color features after normalization. Each color partition corresponds to one dimension; S13. Grayscale the image, analyze the gray-level co-occurrence relationship of the image in multiple directions based on the gray-level co-occurrence matrix (GLCM), and extract four texture features of the image: contrast, correlation, energy, and entropy; S14. Extract the edge information of the image through the Canny algorithm, calculate the geometric moment and central moment of the image based on the edge region, obtain the geometric invariant moment with rotation, translation, and scale invariance, form the edge shape feature, and identify and calculate the center of the target edge; S15. Generate a visual feature vector after normalizing the color feature, texture feature, and edge shape feature.

[0007] Preferably, the steps of extracting motion features from the image to generate a motion feature vector, identifying interference objects and generating visual and motion feature vectors of the interference objects are as follows: S21. Obtain the positions of the target tracking object in the current frame image and the previous frame image, grayscale the target area, the velocity vector is the ratio of the difference between the current frame position and the previous frame position to the time, the change in the motion direction angle is obtained through the arctangent function and the position of the target tracking object, and the acceleration is calculated through the rate of change of velocity between two frames. Generate a motion feature vector from the components of the velocity vector on the x-axis and y-axis, the change in the motion direction angle, and the acceleration. Among them, the velocity vector calculation formula is: , the motion direction angle change calculation formula is: , , the acceleration calculation formula is: , where, is the target center coordinate of the current frame, is the target center coordinate of the previous frame, is the time interval between two frames; S22. With the current tracking target position as the center, expand the detection window by a certain proportion. Perform image segmentation and connected component analysis within this area. The segmentation method is threshold segmentation based on color clustering. Region segmentation and connected component analysis based on color distribution are used to extract potential object regions. Filter out regions with too small an area or a large difference in size from the target. For each interference candidate region, extract and generate the visual feature vector and motion feature vector of the interference object, and keep them in the same feature space as the feature vector of the current tracking target.

[0008] Preferably, for constructing the target similarity interference model, importing the visual feature vectors and motion feature vectors of the target tracking object and the interference object into the target similarity interference model to evaluate the similarity degree between the target tracking object and the interference object includes the following specific steps: S31. Substitute the visual feature vectors of the target tracking object and the interference object into the visual feature cosine similarity calculation formula to calculate the similarity degree of the visual features between the target tracking object and the interference object. Among them, the visual feature cosine similarity calculation formula is: , where is the visual feature vector of the target tracking object, is the visual feature vector of the interference object, is the modulus of the vector; S32. Substitute the motion feature vectors of the target tracking object and the interference object into the motion feature similarity calculation formula to calculate the similarity degree of the motion features between the target tracking object and the interference object. Among them, the motion feature similarity calculation formula is: , where is the motion feature vector of the target object in the current frame, is the motion feature vector of the interference object in the current frame, is the Euclidean distance between the motion feature vectors of the target tracking object and the interference object, used to measure the difference in motion behavior, is the maximum value of the modulus lengths of the two; S33. Substitute the visual feature cosine similarity and the motion feature similarity into the target similarity interference calculation formula to calculate the target similarity interference degree. Among them, the target similarity interference calculation formula is: , where and are weights.

[0009] Preferably, for constructing the environmental interference model, importing the collected image environmental data into the environmental interference model to evaluate the environmental interference degree includes the following specific steps: S41. Substitute the average brightness values of the current frame and the previous frame into the illumination interference calculation formula to evaluate the illumination interference. Among them, the illumination interference calculation formula is: , where is the average luminance value of the current frame, is the average luminance value of the previous frame; S42. Substitute the target effective area of the current frame and the previous frame into the partial occlusion interference calculation formula to evaluate the partial occlusion interference, where the partial occlusion interference calculation formula is: where, is the target effective area detected in the current frame, is the target effective area of the previous frame, standard area, the standard area is a constant to avoid the denominator being zero; S43. Substitute the number of image edge pixels into the scene complexity calculation formula to evaluate the scene complexity, where the scene complexity calculation formula is: where, is the total number of edge pixels detected in the image, extracted by the Canny edge detection algorithm, W is the image width, H is the image height, is the total number of pixels, is the number of significant moving or static objects in the scene, obtained by background subtraction statistics, is used to suppress the order of magnitude difference and smooth the complexity growth, quantifies the number of edges per unit area, reflecting the texture complexity; S44. Substitute the illumination interference, partial occlusion interference and scene complexity into the environmental interference calculation formula to evaluate the degree of environmental interference, where the environmental interference calculation formula is: where, r is the interference factor, including illumination interference, partial occlusion interference and scene complexity, is the interference factor weight, is the theoretical minimum value of the rth interference factor, the theoretical maximum value of the rth interference factor, linear normalization to eliminate the dimension difference.

[0010] Preferably, when constructing the target tracking effect evaluation model, import the target similarity interference and environmental interference into the target tracking effect evaluation model, and evaluate the target tracking effect in combination with the continuity and stability of target tracking, including the following specific steps: S51. Substitute the target similarity interference and environmental interference into the comprehensive interference calculation formula to evaluate the comprehensive interference situation, where in the comprehensive interference calculation formula: , k is the scaling coefficient used to control the curve steepness, Y is the total number of frames in the target tracking process, is the target similarity interference at the tth frame, The environmental interference at the t-th frame is used to generate a normalized comprehensive interference through fusing the target similarity interference value and the environmental interference value and mapping through the Sigmoid function. The integral is used to accumulate the interference intensity of each frame during the entire tracking process; S52. Substitute the actual center coordinates of the target tracking object and the center coordinates of the tracking frame into the target tracking stability calculation formula to evaluate the stability of the target tracking process. The target tracking stability calculation formula is: , where is the true target center coordinate, is the center coordinate of the tracking frame, is the image diagonal length; S53. Substitute the comprehensive interference, the target tracking stability, and the target tracking duration into the target tracking effect evaluation formula to evaluate the target tracking effect. The target tracking effect evaluation formula is: , where Z is the target tracking interruption duration, T is the total duration from the appearance to the disappearance of the target, is the continuity of the target tracking. Compare the target tracking effect evaluation with the threshold. If the target tracking effect value is higher than the threshold, the target tracking effect meets the requirements. If the target tracking effect value is lower than the threshold, it indicates that the target tracking effect does not meet the requirements.

[0011] The intelligent target tracking data analysis system based on DSP is implemented based on the above-mentioned intelligent target tracking data analysis method based on DSP, and specifically includes: An image acquisition module, which is used to acquire image data and perform preprocessing; A feature extraction module, which is used to extract visual features and motion features from the image and generate a visual feature vector and a motion feature vector; A target similarity interference evaluation module, which is used to evaluate the similarity degree between the target tracking object and the interference object through the visual feature vector and the motion feature vector of the target tracking object and the interference object; An environmental interference evaluation module, which is used to evaluate the environmental interference degree through the acquired image environmental data; A target tracking effect evaluation module, which is used to evaluate the target tracking effect by combining the target similarity interference and the environmental interference with the continuity and stability of the target tracking.

[0012] An electronic device includes: a processor and a memory. Among them, a computer program that can be called by the processor is stored in the memory; The processor executes the above-mentioned intelligent target tracking data analysis method based on DSP by calling the computer program stored in the memory.

[0013] A computer-readable storage medium, characterized in that it stores instructions, which, when run on a computer, cause the computer to execute the above-mentioned DSP-based intelligent target tracking data analysis method.

[0014] Compared with the prior art, the beneficial effects of this application are as follows: This application collects image data, extracts visual features from the images and generates visual feature vectors, extracts motion features from the images to generate motion feature vectors, identifies interfering objects and generates visual feature vectors and motion feature vectors of the interfering objects, constructs a target similarity interference model, imports the visual feature vectors and motion feature vectors of the target tracking object and the interfering object into the target similarity interference model to evaluate the similarity degree between the target tracking object and the interfering object, constructs an environmental interference model, imports the collected image environmental data into the environmental interference model to evaluate the environmental interference degree, constructs a target tracking effect evaluation model, imports the target similarity interference and environmental interference into the target tracking effect evaluation model, and evaluates the target tracking effect in combination with the continuity and stability of target tracking, thereby improving the evaluation ability of the target tracking effect in an interference environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 is a schematic diagram of the overall process of the DSP-based intelligent target tracking data analysis method of this application; Figure 2 is a flow chart of the target similarity interference calculation of this application; Figure 3 is a flow chart of the target tracking effect evaluation of this application; Figure 4 is a schematic diagram of the overall framework of the DSP-based intelligent target tracking data analysis system of this application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0016] The following will clearly and completely describe the technical solutions in the embodiments of this application with reference to the accompanying drawings in the embodiments of this application. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all of the embodiments.

[0017] Embodiment 1 Please refer to Figures 1-3 , an embodiment provided by this application: a DSP-based intelligent target tracking data analysis method, which includes the following specific steps: Collect image data, extract visual features from the images and generate visual feature vectors; Extract motion features from the images to generate motion feature vectors, identify interfering objects and generate visual feature vectors and motion feature vectors of the interfering objects; Construct a target similarity interference model, and import the visual feature vectors and motion feature vectors of the target tracking object and the interference object into the target similarity interference model to evaluate the similarity between the target tracking object and the interference object; Construct an environmental interference model, and import the collected image environmental data into the environmental interference model to evaluate the degree of environmental interference; Construct a target tracking effect evaluation model, import the target similarity interference and environmental interference into the target tracking effect evaluation model, and evaluate the target tracking effect in combination with the continuity and stability of target tracking.

[0018] It should be specifically noted in this embodiment that collecting image data, extracting visual features from the image and generating visual feature vectors include the following specific steps: S11. Collect image data through a high-definition camera and other sensors, and preprocess the collected original image data, including operations such as image denoising, normalization, and enhancement. Perform time synchronization and spatial registration on the multi-sensor data to ensure the consistency of different sensor data in time and space; Exemplarily, for the image data collected by the camera, use the Gaussian filtering algorithm to remove the noise in the image and improve the clarity of the image. For the radar and sensor data, make the data of different devices accurately correspond through coordinate transformation and time calibration; S12. Convert the image to the HSV color space, construct histograms for the hue, saturation, and brightness channels of the target area respectively. Each channel is segmented according to the set number of intervals, and the number of pixels in each segment is counted to construct a frequency histogram. Normalize the histogram of each channel. The value of each interval represents its relative proportion in the target area. Concatenate the normalized histograms of the three channels in sequence to form a one-dimensional color feature vector, and each dimension reflects the proportion of this color partition in the target area; S13. Grayscale the image, analyze the gray-level co-occurrence relationship of the image in multiple directions based on the gray-level co-occurrence matrix (GLCM), and extract four texture features of the image: contrast, correlation, energy, and entropy. Select a certain pixel distance (1 pixel) and direction (0 degrees, 45 degrees, 90 degrees, 135 degrees) to scan the entire grayscale image, and count the frequency of the gray-level value combinations of any two pixels in the image in each direction and distance. For example, if we check the gray-level values of every two adjacent pixels in the 0-degree direction, we get a matrix that represents the frequency of the pixel pairs with different gray-level values in the image. According to the gray-level co-occurrence matrix, extract four texture features of the image: contrast, correlation, energy, and entropy. Contrast measures the roughness of the image, correlation measures the relationship or consistency between gray-level values, energy measures the uniformity of the image texture, and entropy measures the complexity or uncertainty of the gray-level distribution of the image; S14. Extract the edge information of the image through the Canny algorithm, including steps of noise removal, gradient calculation, non-maximum suppression, and double thresholding, to obtain a clear edge map. Calculate the geometric moments and central moments of the image based on the edge region to obtain geometric invariant moments with rotation, translation, and scale invariance. Geometric moments are mathematical tools for describing the shape of an object and reflect the shape characteristics of the object in the image. Central moments are the normalization processing of geometric moments, constituting the edge shape characteristics. Calculate the weighted coordinates of each pixel point, where the weighted value is the gray value of the pixel. Sum the weighted coordinates of all pixels to obtain the centroid position of the target region, that is, the target center; S15. Normalize the color features, texture features, and edge shape features and generate a visual feature vector.

[0019] In this embodiment, it should be specifically noted that extracting the motion features of the image to generate a motion feature vector, identifying interference objects, and generating visual and motion feature vectors of the interference objects include the following specific steps: S21. Obtain the positions of the target tracking object in the current frame image and the previous frame image, perform grayscale processing on the target region. The velocity vector is the ratio of the difference between the current frame position and the previous frame position to the time. The change in the motion direction angle is obtained through the arctangent function and the position of the target tracking object, and the acceleration is calculated through the rate of change of velocity between two frames. Generate a motion feature vector from the components of the velocity vector on the x-axis and y-axis, the change in the motion direction angle, and the acceleration. Among them, the velocity vector calculation formula is: , the motion direction angle change calculation formula is: , , the acceleration calculation formula is: , where, is the target center coordinate of the current frame, is the target center coordinate of the previous frame, is the time interval between two frames; S22. With the current tracking target position as the center, expand the detection window by a certain ratio. Perform image segmentation and connected component analysis in this area. The segmentation method is threshold segmentation based on color clustering. Region segmentation based on color distribution and connected component analysis are used to extract potential object regions. Filter out regions with too small an area or a large difference in size from the target. For each interference candidate region, extract and generate visual and motion feature vectors of the interference object, and keep them in the same feature space as the feature vector of the current tracking target.

[0020] Exemplarily, in each frame of the image, interference objects that affect the tracking stability are detected. Taking the current tracking target position as the center, a detection window with a set ratio (1.5 times the width and height) is expanded outward. The image within the window is subjected to preliminary segmentation based on color distribution, and several connected region areas are extracted. For each connected region, screening is performed according to the area, aspect ratio, and similarity to the target size, and interference objects close to the target scale are retained. For each interference object, a visual feature vector and a motion feature vector are extracted according to the predefined visual feature and motion feature extraction methods.

[0021] In this embodiment, it should be specifically noted that constructing a target similarity interference model and importing the visual feature vectors and motion feature vectors of the target tracking object and the interference object into the target similarity interference model to evaluate the similarity degree between the target tracking object and the interference object includes the following specific steps: S31. Substitute the visual feature vectors of the target tracking object and the interference object into the visual feature cosine similarity calculation formula to calculate the similarity degree of the visual features between the target tracking object and the interference object. Among them, the visual feature cosine similarity calculation formula is: where, is the visual feature vector of the target tracking object, is the visual feature vector of the interference object, is the modulus of the vector; S32. Substitute the motion feature vectors of the target tracking object and the interference object into the motion feature similarity calculation formula to calculate the similarity degree of the motion features between the target tracking object and the interference object. Among them, the motion feature similarity calculation formula is: where, is the motion feature vector of the target object in the current frame, is the motion feature vector of the interference object in the current frame, is the Euclidean distance between the motion feature vectors of the target tracking object and the interference object, which is used to measure the difference in motion behavior, is the maximum value of the modulus lengths of the two, and normalization is performed to make the distance value related to the amount of motion. The closer the similarity is to 1, the closer the motion behavior is and the stronger the interference; S33. Substitute the visual feature cosine similarity and the motion feature similarity into the target similarity interference calculation formula to calculate the target similarity interference degree. Among them, the target similarity interference calculation formula is: where, and are weights; Exemplarily, in this embodiment, , indicating that the visual feature cosine similarity has a greater weight in the target similarity interference calculation, and the interference effect is stronger.

[0022] In this embodiment, it should be specifically noted that constructing an environmental interference model and importing the collected image environmental data into the environmental interference model to evaluate the degree of environmental interference includes the following specific steps: S41. Substitute the average brightness values of the current frame and the previous frame into the illumination interference calculation formula to evaluate the illumination interference. The illumination interference calculation formula is: , where is the average brightness value of the current frame, is the average brightness value of the previous frame; S42. Substitute the target effective area sizes of the current frame and the previous frame into the partial occlusion interference calculation formula to evaluate the partial occlusion interference. The partial occlusion interference calculation formula is: , where is the detected target effective area size of the current frame, is the target effective area size of the previous frame, Standard area, the standard area is a constant to avoid the denominator being zero; S43. Substitute the number of image edge pixels into the scene complexity calculation formula to evaluate the scene complexity. The scene complexity calculation formula is: , where is the total number of detected edge pixels in the image, extracted by the Canny edge detection algorithm, W is the image width, H is the image height, is the total number of pixels, and the influence of resolution differences is eliminated by normalizing the edge quantity, is the number of significant moving or static objects in the scene, obtained by background subtraction statistics, is used to suppress the order of magnitude differences and smooth the complexity growth, quantifies the number of edges per unit area and reflects the texture complexity; S44. Substitute the illumination interference, partial occlusion interference, and scene complexity into the environmental interference calculation formula to evaluate the degree of environmental interference. The environmental interference calculation formula is: , where r is the interference factor, including illumination interference, partial occlusion interference, and scene complexity, is the interference factor weight, is the theoretical minimum value of the rth interference factor, the theoretical maximum value of the rth interference factor, and linear normalization eliminates the dimension difference; Exemplarily, the illumination interference of the current frame is 60, the theoretical range is [30, 90], the partial occlusion interference is 0.4, the theoretical range is [0, 1], the scene complexity is 0.75, the theoretical range is [0.2, 1.0], and the interference factor weights are 0.4, 0.35, and 0.25 respectively. Substitute them into the environmental interference calculation formula to calculate the degree of environmental interference.

[0023] In this embodiment, it should be specifically noted that to construct a target tracking effect evaluation model, import the target similarity interference and environmental interference into the target tracking effect evaluation model, and evaluate the target tracking effect in combination with the continuity and stability of target tracking, including the following specific steps: S51. Substitute the target similarity interference and environmental interference into the comprehensive interference calculation formula to evaluate the comprehensive interference situation. Among them, in the comprehensive interference calculation formula: , where k is a scaling coefficient used to control the steepness of the curve, Y is the total number of frames in the target tracking process, is the target similarity interference at the t-th frame, is the environmental interference at the t-th frame. By fusing the target similarity interference value and the environmental interference value, a normalized comprehensive interference is generated through the Sigmoid function mapping. The integral is used to accumulate the interference intensity of each frame in the entire tracking process to comprehensively evaluate the comprehensive interference level endured by the system in the full-frame sequence; Exemplarily, in this embodiment, the scaling coefficient is set to k = 5 to improve the sensitivity to medium and high interference levels; S52. Substitute the actual center coordinates of the target tracking object and the center coordinates of the tracking frame into the target tracking stability calculation formula to evaluate the stability of the target tracking process. Among them, the target tracking stability calculation formula is: , where is the real target center coordinate, is the center coordinate of the tracking frame, is the length of the image diagonal. By normalizing the tracking error of each frame and taking the reciprocal of the average, the overall stability during the tracking process is measured. The smaller the error, the higher the stability score; S53. Substitute the comprehensive interference, target tracking stability, and target tracking duration into the target tracking effect evaluation formula to evaluate the target tracking effect. Among them, the target tracking effect evaluation formula is: , where Z is the target tracking interruption duration, T is the total duration from the appearance to the disappearance of the target, is the continuity of target tracking. Compare the target tracking effect evaluation with the threshold. If the target tracking effect value is higher than the threshold, the target tracking effect meets the requirements. If the target tracking effect value is lower than the threshold, it indicates that the target tracking effect does not meet the requirements; Exemplarily, if the target tracking effect value is lower than the threshold and it is detected that the increase in environmental interference leads to a decline in tracking performance, according to the interference type, dynamically adjust the weights in the feature fusion strategy. When the light changes significantly, reduce the weight of the color feature and increase the weights of the texture and edge features. When occlusion occurs, enhance the importance of the motion feature in the matching process and weaken the dependence on static visual features. When the continuity score drops, indicating that the target is briefly lost, adaptively expand the search area for the next frame to increase the probability of re-capturing the target.

[0024] It should be noted here that the value-taking method of various set parameters in this embodiment is as follows: obtain representative target tracking interference data, and at the same time obtain the target tracking effect. Hire experts to manually judge whether the target tracking meets the requirements. At the same time, substitute the obtained historical data into the calculation results and judgment results of each step in this embodiment and substitute them into the fitting software to output the value-taking of various set parameters that meet the highest judgment accuracy rate.

[0025] The advantages of this embodiment compared with the prior art are as follows: this application collects image data, extracts visual features from the image and generates visual feature vectors, extracts motion features from the image to generate motion feature vectors, identifies interference objects and generates visual feature vectors and motion feature vectors of the interference objects, constructs a target similarity interference model, imports the visual feature vectors and motion feature vectors of the target tracking object and the interference object into the target similarity interference model to evaluate the similarity degree between the target tracking object and the interference object, constructs an environmental interference model, imports the collected image environmental data into the environmental interference model to evaluate the environmental interference degree, constructs a target tracking effect evaluation model, imports the target similarity interference and environmental interference into the target tracking effect evaluation model, and evaluates the target tracking effect in combination with the continuity and stability of target tracking, improving the evaluation ability of the target tracking effect in an interference environment.

[0026] Embodiment 2 As Figure 4 shown, the intelligent target tracking data analysis system based on DSP is implemented based on the above-mentioned intelligent target tracking data analysis method based on DSP. It specifically includes an image acquisition module, a feature extraction module, a target similarity interference evaluation module, an environmental interference evaluation module, and a target tracking effect evaluation module. The image acquisition module is used to collect image data and perform preprocessing; the feature extraction module is used to extract visual features and motion features from the image and generate visual feature vectors and motion feature vectors; the target similarity interference evaluation module is used to evaluate the similarity degree between the target tracking object and the interference object through the visual feature vectors and motion feature vectors of the target tracking object and the interference object; the environmental interference evaluation module is used to evaluate the environmental interference degree through the collected image environmental data; the target tracking effect evaluation module is used to evaluate the target tracking effect by combining the target similarity interference and environmental interference with the continuity and stability of target tracking.

[0027] It should be specifically noted here that the image acquisition module has an image preprocessing unit built-in to complete preliminary operations such as gray-scale equalization, noise suppression, and edge enhancement. The image acquisition module supports the multi-resolution image acquisition mode, and the image data will be directly transmitted to the internal storage unit of the DSP through the DMA channel or cache, reducing data transmission delay and improving the overall response speed of the system.

[0028] Embodiment 3 This embodiment provides an electronic device, including: a processor and a memory, where a computer program that can be called by the processor is stored in the memory; The processor executes the above-mentioned intelligent target tracking data analysis method based on DSP by calling the computer program stored in the memory.

[0029] This electronic device may have relatively large differences due to different configurations or performances, and can include one or more processors (Central Processing Units, CPUs) and one or more memories. Among them, at least one computer program is stored in the memory, and this computer program is loaded and executed by the processor to implement the intelligent target tracking data analysis method based on DSP provided by the above method embodiment. This electronic device can also include other components for implementing device functions. For example, this electronic device can also have components such as wired or wireless network interfaces and input / output interfaces for data input and output. This embodiment will not be elaborated here.

[0030] Embodiment 4 This embodiment proposes a computer-readable storage medium, on which a rewritable computer program is stored; When the computer program runs on a computer device, it enables the computer device to execute the above-mentioned intelligent target tracking data analysis method based on DSP.

[0031] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other arbitrary combination. 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 according to the embodiments of the present application 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 computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center in a wired network or / and wireless network manner. The computer-readable storage medium can be any available medium that the computer can access, or a data storage device such as a server or data center that contains one or more collections of available media. The available medium can be a magnetic medium (such as a floppy disk, hard disk, or magnetic tape), an optical medium (such as a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

Claims

1. An intelligent target tracking data analysis method based on DSP, characterized in that, It includes the following specific steps: Collect image data, extract visual features from the image and generate visual feature vectors; Extract motion features from the image to generate motion feature vectors, identify interfering objects and generate visual and motion feature vectors of the interfering objects; Construct a target similarity interference model, and import the visual and motion feature vectors of the target tracking object and the interfering object into the target similarity interference model to evaluate the similarity between the target tracking object and the interfering object; Construct an environmental interference model, and import the collected image environmental data into the environmental interference model to evaluate the degree of environmental interference; Construct a target tracking effect evaluation model, import the target similarity interference and environmental interference into the target tracking effect evaluation model, and evaluate the target tracking effect in combination with the continuity and stability of target tracking.

2. The DSP-based intelligent target tracking data analysis method according to claim 1, characterized in that The step of collecting image data, extracting visual features from the image and generating visual feature vectors includes the following specific steps: S11. Collect image data through a high-definition camera and other sensors, preprocess the collected original image data, and perform time synchronization and spatial registration on the multi-sensor data; S12. Convert the image to the HSV color space, construct a normalized color histogram for the target area, each channel is divided into a fixed number of intervals, count the number of pixels in each interval, and splice them into color features after normalization. Each color partition corresponds to one dimension; S13. Grayscale the image, analyze the gray-level co-occurrence relationship of the image in multiple directions based on the gray-level co-occurrence matrix, and extract four texture features of the image: contrast, correlation, energy, and entropy; S14. Extract the edge information of the image through the Canny algorithm, calculate the geometric moment and central moment of the image based on the edge area, obtain the geometric invariant moment with rotation, translation, and scale invariance, form the edge shape feature, and identify and calculate the center of the target edge; S15. Normalize the color features, texture features, and edge shape features and generate visual feature vectors.

3. The DSP-based intelligent target tracking data analysis method according to claim 2, characterized in that The step of extracting motion features from the image to generate motion feature vectors, identifying interfering objects and generating visual and motion feature vectors of the interfering objects includes the following specific steps: S21. Obtain the positions of the target tracking object in the current frame image and the previous frame image, grayscale the target area, the velocity vector is the ratio of the difference between the current frame position and the previous frame position to the time, the change in the motion direction angle is obtained through the arctangent function and the position of the target tracking object, and the acceleration is calculated through the rate of change of velocity between two frames. Generate motion feature vectors from the components of the velocity vector on the x-axis and y-axis, the change in the motion direction angle, and the acceleration; S22. Take the current tracking target position as the center, expand the detection window by a certain proportion, perform image segmentation and connected component analysis in this area, extract the potential interfering object area, and extract and generate the visual and motion feature vectors of the interfering object.

4. The DSP-based intelligent target tracking data analysis method according to claim 3, wherein The step of constructing a target similarity interference model, importing the visual and motion feature vectors of the target tracking object and the interfering object into the target similarity interference model to evaluate the similarity between the target tracking object and the interfering object includes the following specific steps: S31. Substitute the visual feature vectors of the target tracking object and the interference object into the visual feature cosine similarity calculation formula to calculate the similarity degree of the visual features between the target tracking object and the interference object. The visual feature cosine similarity calculation formula is as follows: , where is the visual feature vector of the target tracking object, is the visual feature vector of the interference object, is the modulus of the vector; S32. Substitute the motion feature vectors of the target tracking object and the interference object into the motion feature similarity calculation formula to calculate the similarity degree of the motion features between the target tracking object and the interference object. The motion feature similarity calculation formula is as follows: , where is the motion feature vector of the target object in the current frame, is the motion feature vector of the interference object in the current frame, is the Euclidean distance between the motion feature vectors of the target tracking object and the interference object; S33. Substitute the visual feature cosine similarity and the motion feature similarity into the target similarity interference calculation formula to calculate the target similarity interference degree, where the target similarity interference calculation formula is: , where and are weights.

5. The DSP-based intelligent target tracking data analysis method according to claim 4, characterized in that, The construction of the environmental interference model, importing the collected image environmental data into the environmental interference model to evaluate the degree of environmental interference includes the following specific steps: S41. Substitute the average luminance values of the current frame and the previous frame into the illumination interference calculation formula to evaluate the illumination interference, where the illumination interference calculation formula is: , where is the average luminance value of the current frame, is the average luminance value of the previous frame; S42. Substitute the target effective region areas of the current frame and the previous frame into the partial occlusion interference calculation formula to evaluate the partial occlusion interference, where the partial occlusion interference calculation formula is: , where is the target effective region area detected in the current frame, is the target effective region area of the previous frame, is the standard area; S43. Substitute the number of image edge pixels into the scene complexity calculation formula to evaluate the scene complexity. The scene complexity calculation formula is as follows: , where is the total number of edge pixels detected in the image, W is the image width, H is the image height, is the total number of pixels, is the number of significant moving and static objects in the scene; S44. Substitute the light interference, partial occlusion interference, and scene complexity into the environmental interference calculation formula to evaluate the degree of environmental interference. The environmental interference calculation formula is as follows: , where r is the interference factor, which is the light interference, partial occlusion interference, and scene complexity, is the interference factor weight, is the theoretical minimum value of the r-th interference factor, The theoretical maximum value of the r-th interference factor.

6. The DSP-based intelligent target tracking data analysis method according to claim 5, characterized in that, The construction of the target tracking effect evaluation model, importing the target similarity interference and environmental interference into the target tracking effect evaluation model, and evaluating the target tracking effect by combining the continuity and stability of target tracking includes the following specific steps: S51. Substitute the target similarity interference and environmental interference into the comprehensive interference calculation formula to evaluate the comprehensive interference situation. In the comprehensive interference calculation formula: , where k is the scaling factor, Y is the total number of frames in the target tracking process, is the target similarity interference at the t-th frame, is the environmental interference at the t-th frame; S52. Substitute the actual center coordinates of the target tracking object and the center coordinates of the tracking box into the target tracking stability calculation formula to evaluate the stability of the target tracking process. The target tracking stability calculation formula is as follows: , where is the true target center coordinate, is the center coordinate of the tracking box, is the length of the image diagonal; S53. Substitute the comprehensive interference, target tracking stability, and target tracking duration into the target tracking effect evaluation formula to evaluate the target tracking effect. The target tracking effect evaluation formula is as follows: , where Z is the target tracking interruption duration, and T is the total duration from the appearance to the disappearance of the target. is the continuity of target tracking. Compare the target tracking effect evaluation with the threshold. If the target tracking effect value is higher than the threshold, the target tracking effect meets the requirements. If the target tracking effect value is lower than the threshold, it indicates that the target tracking effect does not meet the requirements.

7. The DSP-based intelligent target tracking data analysis system is implemented based on the DSP-based intelligent target tracking data analysis method according to any one of claims 1-6, and is characterized in that Specifically, it includes: An image acquisition module for collecting image data and performing preprocessing; A feature extraction module for extracting visual features and motion features from the image and generating visual feature vectors and motion feature vectors; A target similarity interference evaluation module for evaluating the similarity degree between the target tracking object and the interference object through the visual feature vectors and motion feature vectors of the target tracking object and the interference object; An environmental interference evaluation module for evaluating the degree of environmental interference through the collected image environmental data; A target tracking effect evaluation module for evaluating the target tracking effect by combining the target similarity interference and environmental interference with the continuity and stability of target tracking.

8. An electronic device, comprising: A processor and a memory, wherein the memory stores a computer program that can be called by the processor; It is characterized in that the processor executes the DSP-based intelligent target tracking data analysis method according to any one of claims 1-6 by calling the computer program stored in the memory.

9. A computer-readable storage medium, characterized in that, Instructions are stored, and when the instructions run on a computer, the computer executes the DSP-based intelligent target tracking data analysis method according to any one of claims 1-6.

Images