A tracking target loss detection method and system based on color histogram distance

Abstract

The application discloses a tracking target loss detection method based on color histogram distance, comprising: obtaining a tracking image, wherein the tracking image comprises a target region and a background region; determining a tracking frame scaling ratio based on the target region of the tracking image to adjust tracking frame parameters to obtain a target center region, so that the ratio of the number of pixels of a target object in the target center region to the number of pixels of the background satisfies a preset condition; generating a color histogram according to the target center region; storing color histogram feature values in time sequence based on the color histogram to detect target loss according to the feature values; wherein the tracking frame is a rectangle, and the tracking frame parameters comprise a side length and a rotation angle. The application can accurately detect the occurrence of abnormal feature values to improve the long-time working reliability and intelligent level of a target tracking algorithm and an application system.

Description

Technical Field

[0001] This invention belongs to the field of target tracking, specifically relating to a target loss detection method and system based on color histogram distance. Background Technology

[0002] With the rapid development of computer vision technology, its application in military and civilian fields is becoming increasingly widespread. Target tracking technology, as an important branch of computer vision applications, has a wide range of applications in battlefield target reconnaissance, target attack locking, and specific target tracking in monitoring scenarios.

[0003] Currently, in target tracking systems, when a target disappears, goes out of the video view, or is obscured by obstacles, the tracking system and algorithm may search the vicinity of the last known target and incorrectly locate the object most similar to it. Therefore, target tracking systems typically require continuous visual monitoring by personnel to manually determine when a target is lost.

[0004] However, existing technologies have the following drawbacks: 1. They lack a confidence mechanism for tracking results or lack continuous analysis and application of confidence data, making it impossible to detect and utilize outliers in the confidence data, and failing to establish a confidence feature model for the entire lifecycle of target tracking from initial tracking, tracking process, to tracking loss; 2. When the target experiences partial occlusion or temporary disappearance, traditional target tracking algorithms will consider the tracking to have failed and stop tracking, resulting in poor system availability; 3. Traditional target tracking methods rely solely on the features used in the tracking algorithm for search and tracking, resulting in poor performance. Summary of the Invention

[0005] To address the aforementioned problems in the existing technology, this invention provides a tracking target loss detection method and system based on color histogram distance. The technical problem to be solved by this invention is achieved through the following technical solution:

[0006] A tracking target loss detection method based on color histogram distance includes:

[0007] Acquire an image to be tracked, wherein the image to be tracked includes a target region and a background region;

[0008] Based on the target region of the tracked image, the scaling ratio of the tracking box is determined to adjust the tracking box parameters to obtain the target center region, so that the ratio of the number of pixels of the target object to the number of pixels of the background in the target center region meets a preset condition.

[0009] A color histogram is generated based on the target center region;

[0010] The color histogram feature values ​​are stored sequentially based on the color histogram to perform target loss detection based on the feature values; wherein, the tracking box is rectangular, and the tracking box parameters include side length and rotation angle.

[0011] In one specific implementation, target loss detection based on the feature values ​​includes:

[0012] Set up a first-time sliding window, a second-time sliding window, and a third-time sliding window;

[0013] The mean of the feature values ​​within each time sliding window is used as the median Q2 in the box plot, and the lower quartile Q1, upper quartile Q3, and interquartile range IQR of the box plot are calculated.

[0014] The upper limit and lower limit of the sliding window are calculated based on the lower quartile Q1, the upper quartile Q3, and the interquartile range IQR, so as to determine the threshold range based on the upper limit and lower limit of the sliding window.

[0015] The target state is determined based on the first time sliding window, the second time sliding window, the third time sliding window, and the abnormal threshold, wherein the target state includes a normal state, a partially occluded state, a completely occluded state, and a lost state.

[0016] In one specific implementation, determining the target state based on a first time sliding window, a second time sliding window, a third time sliding window, and an anomaly threshold includes:

[0017] When the feature value in the first time sliding window exceeds the threshold range and the feature value in the second and third time sliding windows does not exceed the threshold range, the target state is determined to be a partial occlusion state.

[0018] When the feature values ​​detected in the first and second time sliding windows exceed the threshold range while the feature values ​​in the third time sliding window do not exceed the threshold range, the target state is determined to be a completely occluded state.

[0019] When the feature values ​​in the first time sliding window, the second time sliding window, and the third time sliding window all exceed the threshold range, the target state is determined to be a lost state.

[0020] When the feature values ​​within the first, second, and third time sliding windows are all within the threshold range, the target state is determined to be a normal state.

[0021] In one specific implementation, the upper limit number of sliding windows = upper quartile Q3 + 1.5 × interquartile range IQR;

[0022] The lower limit of the sliding window is equal to the lower quartile Q1 - 1.5 × interquartile range IQR.

[0023] In one specific implementation, the color histogram includes a YUV or RGB color space.

[0024] This invention also provides a target loss detection system based on color histogram distance, comprising:

[0025] An image acquisition module is used to acquire an image to be tracked, wherein the image to be tracked includes a target region and a background region;

[0026] The center region generation module is used to determine the scaling ratio of the tracking box based on the target region of the tracking image to adjust the tracking box parameters to obtain the target center region, so that the ratio of the number of pixels of the target object to the number of pixels of the background in the target center region meets a preset condition.

[0027] A color histogram generation module is used to generate a color histogram based on the target center region;

[0028] The target detection module is used to store the feature values ​​of the color histogram in a time sequence based on the color histogram, so as to perform target loss detection based on the feature values; wherein, the tracking box is rectangular, and the tracking box parameters include side length and rotation angle.

[0029] In one specific embodiment, the target detection module includes:

[0030] A sliding window setting unit is used to set a first-time sliding window, a second-time sliding window, and a third-time sliding window;

[0031] The calculation unit module is used to take the mean of the feature values ​​within each time sliding window as the median Q2 in the box plot, and to calculate the lower quartile Q1, upper quartile Q3, and interquartile range IQR of the box plot;

[0032] The threshold range determination unit is used to calculate the upper limit number and lower limit number of sliding windows based on the lower quartile Q1, the upper quartile Q3 and the interquartile range IQR, so as to determine the threshold range based on the upper limit number and lower limit number of sliding windows;

[0033] The state detection unit is used to determine the target state based on a first time sliding window, a second time sliding window, a third time sliding window and an abnormal threshold, wherein the target state includes a normal state, a partially occluded state, a completely occluded state and a lost state.

[0034] In one specific implementation, the state detection unit is specifically used for:

[0035] When the feature value in the first time sliding window exceeds the threshold range and the feature value in the second and third time sliding windows does not exceed the threshold range, the target state is determined to be a partial occlusion state.

[0036] When the feature values ​​detected in the first and second time sliding windows exceed the threshold range while the feature values ​​in the third time sliding window do not exceed the threshold range, the target state is determined to be a completely occluded state.

[0037] When the feature values ​​in the first time sliding window, the second time sliding window, and the third time sliding window all exceed the threshold range, the target state is determined to be a lost state.

[0038] When the feature values ​​within the first, second, and third time sliding windows are all within the threshold range, the target state is determined to be a normal state.

[0039] In one specific implementation, the upper limit number of sliding windows = upper quartile Q3 + 1.5 × interquartile range IQR;

[0040] The lower limit of the sliding window is equal to the lower quartile Q1 - 1.5 × interquartile range IQR.

[0041] This invention also provides an electronic device, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus;

[0042] Memory, used to store computer programs;

[0043] When the processor executes the program stored in the memory, it implements the steps of the above-described target loss detection method based on color histogram distance.

[0044] The beneficial effects of this invention are:

[0045] 1. The target loss detection method based on color histogram distance of the present invention selects a local range of the target center, constructs color histogram feature values, establishes a feature value time series, and combines outlier data analysis methods, thereby accurately detecting the occurrence of abnormal feature values ​​and improving the long-term working reliability and intelligence level of the target tracking algorithm and application system;

[0046] 2. The target loss detection method based on color histogram distance of the present invention can dynamically adjust the parameter according to the specific target characteristics. The component of the target object in the reduced rectangle will be much larger than the component of the background pixels, ensuring the accuracy of the subsequent color histogram feature description.

[0047] 3. The target loss detection method based on color histogram distance of the present invention adopts a multi-box graph detection method and uses three different time sliding windows of different lengths, so as to meet the needs of different working scenarios, making it more practical and improving the stability of the target tracking algorithm.

[0048] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0049] Figure 1 This is a schematic diagram of a target loss detection method based on color histogram distance provided in an embodiment of the present invention;

[0050] Figure 2 This is a schematic diagram of the state detection and judgment logic provided in an embodiment of the present invention;

[0051] Figure 3 This is a block diagram of a target loss detection system module based on color histogram distance provided in an embodiment of the present invention;

[0052] Figure 4 This is a schematic diagram of the electronic device provided by the present invention. Detailed Implementation

[0053] The present invention will be further described in detail below with reference to specific embodiments, but the implementation of the present invention is not limited thereto.

[0054] Example 1

[0055] Please see Figure 1 , Figure 1 This is a schematic flowchart of a target loss detection method based on color histogram distance provided in an embodiment of the present invention, including:

[0056] S1. Acquire the image to be tracked, wherein the image to be tracked includes a target region and a background region; in this embodiment, the image to be tracked generally refers to the target video data or image data that needs to be tracked. When the target tracking system is working, it usually generates a target bounding box that contains the target object and part of the background scene where the target is located. The target bounding box is usually rectangular, and the aspect ratio of the rectangle is consistent with the external dimensions of the target. The target region refers to the region inside the rectangle. The target region generally includes a complete target pixel, and the region outside the rectangle is called the background region, which does not contain the target pixel.

[0057] S2. Based on the target region of the tracking image, determine the scaling ratio of the tracking box to adjust the tracking box parameters to obtain the target center region, so that the ratio of the number of pixels of the target object to the number of pixels of the background in the target center region meets the preset condition.

[0058] It should be noted that the bounding box can be divided into two types: rotatable and non-rotatable, depending on the algorithm. Under stable tracking conditions, the tracking system locks the target object at the center of the bounding box. Due to the irregular shape of the target and the inconsistency between the rotation angle and the length and width of the video frame, a large number of background pixels will exist in the edge area of ​​the bounding box, which will significantly hinder tracking. This embodiment sets a tracking box scaling parameter λ (e.g., 0.5 < λ < 0.9). The system can dynamically adjust this parameter according to the specific target characteristics. After scaling down, the target object component within the bounding box will be much larger than the background pixel component (e.g., setting the preset condition for this ratio to 1:0.05-1:0.01), ensuring the accuracy of subsequent color histogram feature description.

[0059] The target loss detection method based on color histogram distance of the present invention can dynamically adjust the parameter according to the specific target characteristics. The component of the target object in the reduced rectangle will be much larger than the component of the background pixels, ensuring the accuracy of the subsequent color histogram feature description.

[0060] S3. Generate a color histogram based on the target center region;

[0061] The color histogram in this embodiment can use both YUV and RGB color spaces, which can be selected according to the image data format of the video source. For YUV format, preferably, when the image quality is poor, the weight of the Y channel can be increased, and the luminance channel should be given more consideration.

[0062] S4. Based on the color histogram, store the color histogram feature values ​​in time sequence to perform target loss detection according to the feature values; wherein, the tracking box is rectangular, and the tracking box parameters include side length and rotation angle.

[0063] In one specific implementation, target loss detection based on the feature values ​​includes:

[0064] S41. Set a first time sliding window, a second time sliding window, and a third time sliding window;

[0065] The mean of the feature values ​​within each time sliding window is used as the median Q2 in the box plot, and the lower quartile Q1, upper quartile Q3, and interquartile range IQR of the box plot are calculated.

[0066] S42. Calculate the upper limit number and lower limit number of sliding windows based on the lower quartile Q1, the upper quartile Q3 and the interquartile range IQR, so as to determine the threshold range based on the upper limit number and lower limit number of sliding windows;

[0067] S43. Determine the target state based on the first time sliding window, the second time sliding window, the third time sliding window and the abnormal threshold, wherein the target state includes a normal state, a partially occluded state, a completely occluded state and a lost state.

[0068] In one specific implementation, step S43 specifically includes:

[0069] When the feature value in the first time sliding window exceeds the threshold range and the feature value in the second and third time sliding windows does not exceed the threshold range, the target state is determined to be a partial occlusion state.

[0070] When the feature values ​​detected in the first and second time sliding windows exceed the threshold range while the feature values ​​in the third time sliding window do not exceed the threshold range, the target state is determined to be a completely occluded state.

[0071] When the feature values ​​in the first time sliding window, the second time sliding window, and the third time sliding window all exceed the threshold range, the target state is determined to be a lost state.

[0072] When the feature values ​​within the first, second, and third time sliding windows are all within the threshold range, the target state is determined to be a normal state.

[0073] For details on making the above judgments, please refer to [link / reference]. Figure 2 The judgment logic is as follows: The feature values ​​of the first, second, and third time-sliding windows are judged sequentially. That is, if no abnormal feature value appears in the first time-sliding window, the second and third time-sliding windows are considered normal, and the state is directly judged as normal. If the first time-sliding window shows an abnormality, the second time-sliding window needs to be further judged. If the second time-sliding window does not show an abnormality, the third time-sliding window is considered normal, and the state is judged as partially occluded. If the second time-sliding window shows an abnormality, the third time-sliding window needs to be judged. When the feature values ​​in the first, second, and third time-sliding windows all exceed the threshold range, the target state is determined to be lost; otherwise, the target is considered completely occluded, and the state is determined to be completely occluded. It should be noted that the times of the first, second, and third time-sliding windows can be set according to the situation. The target loss detection method based on color histogram distance of the present invention adopts a multi-box graph detection method and uses three time sliding windows of different lengths, so as to meet the needs of different working scenarios, making it more practical and improving the stability of the target tracking algorithm.

[0074] In one specific implementation, the upper limit number of sliding windows = upper quartile Q3 + 1.5 × interquartile range IQR;

[0075] The lower limit of the sliding window is equal to the lower quartile Q1 - 1.5 × interquartile range IQR.

[0076] The target loss detection method based on color histogram distance of the present invention selects a local area at the center of the target, constructs color histogram feature values, establishes a time series of feature values, and combines outlier data analysis methods to accurately detect the occurrence of abnormal feature values, thereby improving the long-term reliability and intelligence level of target tracking algorithms and application systems.

[0077] Please see Figure 3 The present invention also provides a target loss detection system based on color histogram distance, comprising:

[0078] Image acquisition module 31 is used to acquire an image to be tracked, wherein the image to be tracked includes a target region and a background region;

[0079] The center region generation module 32 is used to determine the scaling ratio of the tracking box based on the target region of the tracking image to adjust the tracking box parameters to obtain the target center region, so that the ratio of the number of pixels of the target object to the number of pixels of the background in the target center region meets a preset condition.

[0080] The color histogram generation module 33 is used to generate a color histogram based on the target center region;

[0081] The target detection module 34 is used to store the feature values ​​of the color histogram in a time sequence based on the color histogram, so as to perform target loss detection based on the feature values; wherein, the tracking box is a rectangle, and the tracking box parameters include side length and rotation angle.

[0082] In one specific embodiment, the target detection module includes:

[0083] A sliding window setting unit is used to set a first-time sliding window, a second-time sliding window, and a third-time sliding window;

[0084] The calculation unit module is used to take the mean of the feature values ​​within each time sliding window as the median Q2 in the box plot, and to calculate the lower quartile Q1, upper quartile Q3, and interquartile range IQR of the box plot;

[0085] The threshold range determination unit is used to calculate the upper limit number and lower limit number of sliding windows based on the lower quartile Q1, the upper quartile Q3 and the interquartile range IQR, so as to determine the threshold range based on the upper limit number and lower limit number of sliding windows;

[0086] The state detection unit is used to determine the target state based on a first time sliding window, a second time sliding window, a third time sliding window and an abnormal threshold, wherein the target state includes a normal state, a partially occluded state, a completely occluded state and a lost state.

[0087] In one specific implementation, the state detection unit is specifically used for:

[0088] When the feature value in the first time sliding window exceeds the threshold range and the feature value in the second and third time sliding windows does not exceed the threshold range, the target state is determined to be a partial occlusion state.

[0089] When the feature values ​​detected in the first and second time sliding windows exceed the threshold range while the feature values ​​in the third time sliding window do not exceed the threshold range, the target state is determined to be a completely occluded state.

[0090] When the feature values ​​in the first time sliding window, the second time sliding window, and the third time sliding window all exceed the threshold range, the target state is determined to be a lost state.

[0091] When the feature values ​​within the first, second, and third time sliding windows are all within the threshold range, the target state is determined to be a normal state.

[0092] In one specific implementation, the upper limit number of sliding windows = upper quartile Q3 + 1.5 × interquartile range IQR;

[0093] The lower limit of the sliding window is equal to the lower quartile Q1 - 1.5 × interquartile range IQR.

[0094] This invention also provides an electronic device, such as... Figure 4 As shown, it includes a processor 41, a communication interface 42, a memory 43, and a communication bus 44, wherein the processor 41, the communication interface 42, and the memory 43 communicate with each other through the communication bus 44.

[0095] Memory 43 is used to store computer programs;

[0096] When processor 41 executes the program stored in memory 43, it performs the following steps:

[0097] S1. Obtain the image to be tracked;

[0098] S2. Based on the target region of the tracking image, determine the scaling ratio of the tracking box to adjust the tracking box parameters to obtain the target center region, so that the ratio of the number of pixels of the target object to the number of pixels of the background in the target center region meets the preset condition.

[0099] S3. Generate a color histogram based on the target center region;

[0100] S4. Based on the color histogram, store the color histogram feature values ​​in time sequence to perform target loss detection according to the feature values.

[0101] The communication bus mentioned in the above electronic devices can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc. For ease of illustration, only one thick line is used to represent it in the diagram, but this does not mean that there is only one bus or one type of bus.

[0102] The communication interface is used for communication between the aforementioned electronic devices and other devices.

[0103] The memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0104] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0105] The method provided in this invention can be applied to electronic devices. Specifically, the electronic device can be a desktop computer, a portable computer, a smart mobile terminal, a server, etc. No limitation is made herein; any electronic device that can implement this invention falls within the protection scope of this invention.

[0106] For the embodiments of the device / electronic device / storage medium, since they are basically similar to the method embodiments, the description is relatively simple, and relevant parts can be referred to in the description of the method embodiments.

[0107] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0108] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. In addition, those skilled in the art can combine and integrate the different embodiments or examples described in this specification.

[0109] Although this application has been described herein in conjunction with various embodiments, those skilled in the art, by reviewing the accompanying drawings, disclosure, and appended claims, will understand and implement other variations of the disclosed embodiments in carrying out the claimed application. In the claims, the word "comprising" does not exclude other components or steps, and "a" or "an" does not exclude a plurality. A single processor or other unit can implement several functions listed in the claims. While different dependent claims may recite certain measures, this does not mean that these measures cannot be combined to produce good results.

[0110] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems (devices), or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects, all of which are collectively referred to herein as "modules" or "systems." Furthermore, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The computer program may be stored / distributed in a suitable medium, provided with or as part of other hardware, or may take other distribution forms, such as via the Internet or other wired or wireless telecommunications systems.

[0111] This application is described with reference to flowchart illustrations and / or block diagrams of methods, systems (devices), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0112] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0113] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0114] The above description, in conjunction with specific preferred embodiments, provides a further detailed explanation of the present invention. It should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, various simple deductions or substitutions can be made without departing from the concept of the present invention, and all such modifications and substitutions should be considered within the scope of protection of the present invention.

Claims

1. A tracking target loss detection method based on color histogram distance, characterized in that, include: Acquire an image to be tracked, wherein the image to be tracked includes a target region and a background region; Based on the target region of the image to be tracked, the scaling ratio of the tracking box is determined to adjust the tracking box parameters to obtain the target center region, so that the ratio of the number of pixels of the target object to the number of pixels of the background in the target center region meets a preset condition, the preset condition being 1:0.05-1:0.01; A color histogram is generated based on the target center region; The color histogram feature values ​​are stored sequentially based on the color histogram to perform target loss detection based on the feature values; wherein, the tracking box is rectangular, and the tracking box parameters include side length and rotation angle; Target loss detection based on the aforementioned feature values ​​includes: Set up a first-time sliding window, a second-time sliding window, and a third-time sliding window; The mean of the feature values ​​within each time sliding window is used as the median Q2 in the box plot, and the lower quartile Q1, upper quartile Q3, and interquartile range IQR of the box plot are calculated. The upper limit and lower limit of the sliding window are calculated based on the lower quartile Q1, the upper quartile Q3, and the interquartile range IQR, so as to determine the threshold range based on the upper limit and lower limit of the sliding window. The target state is determined based on a first time sliding window, a second time sliding window, a third time sliding window, and an anomaly threshold. The target state includes a normal state, a partially occluded state, a completely occluded state, and a lost state. Specifically, it includes: When the feature value in the first time sliding window exceeds the threshold range and the feature value in the second and third time sliding windows does not exceed the threshold range, the target state is determined to be a partial occlusion state. When the feature values ​​detected in the first and second time sliding windows exceed the threshold range while the feature values ​​in the third time sliding window do not exceed the threshold range, the target state is determined to be a completely occluded state. When the feature values ​​in the first time sliding window, the second time sliding window, and the third time sliding window all exceed the threshold range, the target state is determined to be a lost state. When the feature values ​​within the first, second, and third time sliding windows are all within the threshold range, the target state is determined to be a normal state.

2. The target loss detection method based on color histogram distance according to claim 1, characterized in that, The upper limit of the sliding window = upper quartile Q3 + 1.5 × interquartile range IQR; The lower limit of the sliding window is equal to the lower quartile Q1 - 1.5 × interquartile range IQR.

3. The target loss detection method based on color histogram distance according to claim 1, characterized in that, The color histogram includes either the YUV or RGB color space.

4. A target loss detection system based on color histogram distance, characterized in that, include: An image acquisition module is used to acquire an image to be tracked, wherein the image to be tracked includes a target region and a background region; The center region generation module is used to determine the scaling ratio of the tracking box based on the target region of the image to be tracked, so as to adjust the tracking box parameters to obtain the target center region, and make the ratio of the number of pixels of the target object to the number of pixels of the background in the target center region meet the preset conditions. A color histogram generation module is used to generate a color histogram based on the target center region; The target detection module is used to store the feature values ​​of the color histogram in a time sequence based on the color histogram, so as to perform target loss detection based on the feature values; wherein, the tracking box is rectangular, and the tracking box parameters include side length and rotation angle; The target detection module includes: A sliding window setting unit is used to set a first-time sliding window, a second-time sliding window, and a third-time sliding window; The calculation unit module is used to take the mean of the feature values ​​within each time sliding window as the median Q2 in the box plot, and to calculate the lower quartile Q1, upper quartile Q3, and interquartile range IQR of the box plot; The threshold range determination unit is used to calculate the upper limit number and lower limit number of sliding windows based on the lower quartile Q1, the upper quartile Q3 and the interquartile range IQR, so as to determine the threshold range based on the upper limit number and lower limit number of sliding windows; A state detection unit is used to determine the target state based on a first time sliding window, a second time sliding window, a third time sliding window, and an anomaly threshold, wherein the target state includes a normal state, a partially occluded state, a completely occluded state, and a lost state; the state detection unit is specifically used for: When the feature value in the first time sliding window exceeds the threshold range and the feature value in the second and third time sliding windows does not exceed the threshold range, the target state is determined to be a partial occlusion state. When the feature values ​​detected in the first and second time sliding windows exceed the threshold range while the feature values ​​in the third time sliding window do not exceed the threshold range, the target state is determined to be a completely occluded state. When the feature values ​​in the first time sliding window, the second time sliding window, and the third time sliding window all exceed the threshold range, the target state is determined to be a lost state. When the feature values ​​within the first, second, and third time sliding windows are all within the threshold range, the target state is determined to be a normal state.

5. The target loss detection system based on color histogram distance according to claim 4, characterized in that, The upper limit of the sliding window = upper quartile Q3 + 1.5 × interquartile range IQR; The lower limit of the sliding window is equal to the lower quartile Q1 - 1.5 × interquartile range IQR.

6. An electronic device, characterized in that, It includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; A processor, when executing a program stored in memory, implements the method of any one of claims 1-3.

Images