Anti-interference multi-dimensional feature detection method and system

Abstract

The application discloses an anti-interference multi-dimensional feature detection method and system, and the method comprises the following steps: acquiring a current frame to-be-detected image of a to-be-detected target, a latest frame radar target information of a current frame, radio target information and a historical target track library; acquiring a plurality of target tracks based on an SVM classification model and a speed motion function according to the to-be-detected image; acquiring a total confidence corresponding to each target track according to the plurality of target tracks, the radio target information, the latest frame radar target information and the historical target track library; and selecting a target track with the minimum total confidence as a determined target. In the process of identifying a low, small and slow target, the hardware requirement is low, the operation speed is fast, the identification rate is high, and the anti-background interference capability is strong.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of image detection technology, in particular to an anti-interference multi-dimensional feature detection method and system. BACKGROUND

[0002] Low, small and slow targets have the characteristics of complex surrounding environment, slow target speed, and few image features. Under the framework of existing image detection algorithms, it is difficult to quickly and accurately identify them. The existing low, small and slow detection methods are as follows: 1. Using a neural network or other high hardware requirement machine learning algorithm to collect a large amount of data for target identification; 2. Using a primary image recognition algorithm to detect the target.

[0003] For the above two detection methods, the former has poor interpretability, high requirements for hardware and data collection, slow running speed, and high cost of core hardware localization. The latter has low detection rate. The biggest common shortcoming of the two is that they only detect from the image angle and are difficult to distinguish the target from the background. SUMMARY

[0004] The present application provides an anti-interference multi-dimensional feature detection method and system, which has low hardware requirements, fast operation speed, high recognition rate, and strong anti-background interference capability in the process of identifying low, small and slow targets.

[0005] In a first aspect, an anti-interference multi-dimensional feature detection method is provided, comprising the following steps:

[0006] Obtaining a current frame of a to-be-detected image of a to-be-detected target, a latest frame of radar target information of a current frame, radio target information, and a historical target track library;

[0007] Obtaining a plurality of target tracks based on an SVM classification model and a speed motion function according to the to-be-detected image;

[0008] Obtaining a total confidence corresponding to each target track according to a plurality of target tracks, the radio target information, the latest frame of radar target information, and the historical target track library;

[0009] Selecting a target track with the smallest total confidence as a determined target.

[0010] According to the first aspect, in a first possible implementation manner of the first aspect, the step of "obtaining a plurality of target tracks based on an SVM classification model and a speed motion function according to the to-be-detected image" specifically comprises the following steps:

[0011] Obtaining a plurality of target point tracks based on an SVM classification model according to the to-be-detected image;

[0012] Determining a plurality of target tracks in the plurality of target point tracks based on a speed motion function.

[0013] According to a second possible implementation of the first aspect, the step of obtaining the total confidence corresponding to each target track according to the plurality of target tracks, the radio target information, the latest frame of radar target information and the historical target track library, specifically comprises the following steps:

[0014] calculating an inter-frame correlation quality corresponding to each target track;

[0015] performing aggregation calculation on the plurality of target tracks to obtain an aggregation correlation quality corresponding to each target track;

[0016] obtaining a minimum radio correlation quality corresponding to each target track according to the radio target information and the plurality of target tracks;

[0017] obtaining a minimum radar correlation quality corresponding to each target track according to the latest frame of radar target information and the plurality of target tracks;

[0018] obtaining a minimum historical correlation quality corresponding to each target track according to the historical target track library and the plurality of target tracks;

[0019] obtaining the total confidence corresponding to each target track according to the inter-frame correlation quality, the aggregation correlation quality, the minimum radio correlation quality, the minimum radar correlation quality and the minimum historical correlation quality.

[0020] According to a third possible implementation of the first aspect, the step of calculating the inter-frame correlation quality corresponding to each target track, specifically comprises the following steps:

[0021] calculating an inter-frame correlation quality Q corresponding to each target track v as follows:

[0022]

[0023] wherein V n , V n-1 are respectively a track speed of a target track under the condition of two adjacent frames.

[0024] According to a fourth possible implementation of the first aspect, the step of performing aggregation calculation on the plurality of target tracks to obtain the aggregation correlation quality corresponding to each target track, specifically comprises the following steps:

[0025] obtaining an aggregation correlation quality Q corresponding to each target track gm as follows:

[0026]

[0027] wherein P is a target track spatial position.

[0028] According to a fifth possible implementation manner of the first aspect, in the step of obtaining a minimum radio correlation quality corresponding to each target track according to the radio target information and the plurality of target tracks, the step specifically comprises the following steps:

[0029] interpolating the azimuth in the radio target information and the azimuth in each target track to obtain a minimum radio azimuth correlation quality Q wm as follows:

[0030] Q wm = A m - A w Equation (Three);

[0031] interpolating the speed in the radio target information and the speed in each target track to obtain a minimum radio speed correlation quality Qs wm as follows:

[0032] Qs wm = s m - s w Equation (Four);

[0033] According to the minimum radio azimuth correlation quality Q wm and the minimum radio speed correlation quality Qs wm , a minimum radio correlation quality Q w corresponding to each target track is calculated as follows:

[0034] Q w = w1Q wm + w2Qs wm Equation (Five);

[0035] wherein A m is the azimuth of the target track m; A w is the azimuth of the radio target w; s m is the speed of the target track m; s w is the speed of the radio target w; w1 is a weight coefficient of Q wm ; and w2 is a weight coefficient of Qs wm .

[0036] According to a sixth possible implementation manner of the first aspect, the step of obtaining the minimum radar correlation quality corresponding to each target track according to the latest frame of radar target information and the plurality of target tracks specifically comprises the following steps:

[0037] performing interpolation calculation between the spatial position in the latest frame of radar target information and the spatial position in each target track to obtain a minimum radar spatial correlation quality Q rm as follows:

[0038] Q rm = P m -P r , formula (6);

[0039] performing interpolation calculation between the speed in the latest frame of radar target information and the speed in each target track to obtain a minimum radar speed correlation quality Qs rm as follows:

[0040] Qs rm = s m -s r , formula (7);

[0041] according to the minimum radar spatial correlation quality Q rm and the minimum radar speed correlation quality Qs rm , calculating a minimum radar correlation quality Q r corresponding to each target track as follows:

[0042] Q r = w3Q rm +w4Qs rm , formula (8);

[0043] wherein, P m is the spatial position of the target track m; P r is the spatial position of the radar target r; s m is the speed of the target track m; s r is the speed of the radar target r; w3 is a weight coefficient of Q rm ; and w4 is a weight coefficient of Qs rm .

[0044] According to a seventh possible implementation manner of the first aspect, the step of obtaining the minimum historical correlation quality corresponding to each target track according to the historical target track library and the plurality of target tracks specifically comprises the following steps:

[0045] interpolation between the speed of the historical target track library and the speed of each target track, to obtain a minimum historical correlation quality Qs corresponding to the minimum interpolation result of each target track hm As follows:

[0046] Qs hm = s m -s h Formula (IX);

[0047] In the formula, s m is the speed of the target track m; s h is the speed of the historical target track library target r.

[0048] According to a first aspect, in an eighth possible implementation manner of the first aspect, the step of obtaining a total confidence corresponding to each target track according to the inter-frame correlation quality, the aggregation correlation quality, the minimum radio correlation quality, the minimum radar correlation quality and the minimum historical correlation quality, specifically includes the following steps:

[0049] The total confidence Q corresponding to each target track is obtained as follows:

[0050] Q = w5Q v + w6Q gm + w7Q w + w8Q r + w9Qs hm Formula (X);

[0051] In the formula, w5 is a weight coefficient of the inter-frame correlation quality Q v ; w6 is a weight coefficient of the aggregation correlation quality Q gm ; w7 is a weight coefficient of the minimum radio correlation quality Q w ; w8 is a weight coefficient of the minimum radar correlation quality Q r ; and w9 is a weight coefficient of the minimum historical correlation quality Qs hm .

[0052] The second aspect provides an anti-interference multi-dimensional feature detection system, including the following steps:

[0053] A data acquisition module is configured to acquire a current frame of a to-be-detected image of a to-be-detected target, a most recent frame of radar target information of the current frame, radio target information and a historical target track library;

[0054] A track module is in communication connection with the data acquisition module and is configured to acquire a plurality of target tracks based on an SVM classification model and a speed motion function according to the to-be-detected image;

[0055] A confidence module, in communication connection with the data acquisition module and the track module, is configured to acquire a total confidence corresponding to each target track according to the multiple target tracks, the radio target information, the latest frame of radar target information and the historical target track library;

[0056] A target determination module, in communication connection with the confidence module, is configured to select a target track with the minimum total confidence as a determined target.

[0057] Compared with the prior art, the advantages of the present application are as follows: in the process of identifying a low, small and slow target, not only the image features are considered, but also the kinematics, radar radio target features and historical track features are introduced, the complex neural network algorithm is not used, the calculation amount is not large, the hardware demand is low, the operation speed is fast, the identification rate is high, the anti-background interference capability is strong, and the interpretability is strong. BRIEF DESCRIPTION OF DRAWINGS

[0058] Figure 1 is a flowchart of an embodiment of an anti-interference multi-dimensional feature detection method of the present application;

[0059] Figure 2 is a flowchart of another embodiment of an anti-interference multi-dimensional feature detection method of the present application;

[0060] Figure 3 is a structural schematic diagram of an anti-interference multi-dimensional feature detection system of the present application. DETAILED DESCRIPTION

[0061] Reference will now be made in detail to the embodiments of the present application, examples of which are illustrated in the accompanying drawings. While the present application will be described in conjunction with the embodiments, it will be understood that they are not intended to limit the present application to these embodiments. On the contrary, the present application is intended to cover alternatives, modifications, and equivalents, which can be included within the spirit and scope of the present application as defined by the appended claims. It should be noted that the method steps described herein can be implemented by any functional block or functional arrangement, and any functional block or functional arrangement can be implemented as a physical entity or a logical entity, or a combination of both.

[0062] In order for those skilled in the art to better understand the present application, the present application will be further described in detail below in conjunction with the drawings and specific embodiments.

[0063] Note: the examples to be introduced next are only specific examples, and are not intended to limit the embodiments of the present application to the specific steps, values, conditions, data, sequences, etc. Those skilled in the art can use the concept of the present application to construct more embodiments not mentioned in the present specification by reading the present specification.

[0064] Reference Figure 1As shown, the embodiment of the present application provides an anti-interference multi-dimensional feature detection method, comprising the following steps:

[0065] S100, acquiring a current frame to-be-detected image of a to-be-detected target, a latest frame radar target information of the current frame, radio target information and a historical target track library;

[0066] S200, acquiring a plurality of target tracks based on an SVM classification model and a speed motion function according to the to-be-detected image;

[0067] S300, acquiring a total confidence corresponding to each target track according to the plurality of target tracks, the radio target information, the latest frame radar target information and the historical target track library;

[0068] S400, selecting a target track with the minimum total confidence as a determined target.

[0069] Specifically, in the embodiment, the present application considers not only image features but also kinematics, radar radio target features, historical track features and the like in the process of identifying low, small and slow targets, does not use a complex neural network algorithm, has a small amount of calculation, low hardware requirements, fast operation speed, high recognition rate, strong anti-background interference capability and strong interpretability.

[0070] Preferably, in another embodiment of the present application, the step of "S200, acquiring a plurality of target tracks based on an SVM classification model and a speed motion function according to the to-be-detected image" specifically comprises the following steps:

[0071] S210, acquiring a plurality of target point tracks based on an SVM classification model according to the to-be-detected image;

[0072] S220, determining a plurality of target tracks in the plurality of target point tracks based on a speed motion function.

[0073] Specifically, in the embodiment, a 32*32 sliding box (the size of the sliding box can be changed according to actual conditions) is used to divide the current frame to-be-detected image of the to-be-detected target into a plurality of to-be-detected images, the image is taken as a one-dimensional vector as a lightweight feature vector of the SVM for first layer screening, and a large number of primary undetermined targets are obtained in the picture; the SVM detection constraint function is as follows:

[0074]

[0075] w is a learning parameter of a training model (the training model is obtained by training historical target images in advance), the physical meaning of which is a hyperplane normal vector, b is a hyperplane intercept, "s.t." is a short for Subject to, indicating a constraint item during optimization, i.e. a condition that must be met; i.e. the target point track that meets the constraint function condition.

[0076] When the target point traces in the continuous N frames of images are obtained, the same target point traces are inter-frame correlated, and the speed motion function is as follows:

[0077]

[0078] Therefore, the target point trace satisfying the above speed motion function condition is considered as a target track, V is the target speed, Vmin is the minimum speed, Vmax is the maximum speed, and Vdelta is the maximum speed difference between two frames.

[0079] Preferably, in another embodiment of the present application, the step of "S300, obtaining the total confidence corresponding to each target track according to the plurality of target tracks, the radio target information, the latest frame of radar target information, and the historical target track library" specifically comprises the following steps:

[0080] S310, calculating the inter-frame correlation quality corresponding to each target track;

[0081] S320, performing aggregation degree calculation on the plurality of target tracks to obtain the aggregation correlation quality corresponding to each target track;

[0082] S330, obtaining the minimum radio correlation quality corresponding to each target track according to the radio target information and the plurality of target tracks;

[0083] S340, obtaining the minimum radar correlation quality corresponding to each target track according to the latest frame of radar target information and the plurality of target tracks;

[0084] S350, obtaining the minimum historical correlation quality corresponding to each target track according to the historical target track library and the plurality of target tracks;

[0085] S360, obtaining the total confidence corresponding to each target track according to the inter-frame correlation quality, the aggregation correlation quality, the minimum radio correlation quality, the minimum radar correlation quality, and the minimum historical correlation quality.

[0086] Specifically, in the embodiment, since the existing low small and slow detection method has the following two kinds: 1, using neural network and other high hardware requirement machine learning algorithm to collect a large amount of data for target identification; 2, using a primary image recognition algorithm to detect the target; the former has poor interpretability, high requirements for hardware and data acquisition, slow running speed, difficulty in core hardware localization, high cost, and the latter has low detection rate; the biggest common shortcoming of the two is that they only detect from the image angle, and it is difficult to distinguish the target from the background; therefore, in view of the above problems, the application not only considers the image features, but also introduces kinematics, track clustering relationship, radar radio aiming feature, historical track feature, the algorithm calculation of each layer is not large, the hardware requirement is low, the operation speed is fast, the identification rate is high, the anti-background interference ability is strong, and the interpretability is strong.

[0087] Preferably, in another embodiment of the application, the step of "S310, calculating the inter-frame correlation quality corresponding to each target track" specifically comprises the following steps:

[0088] Calculate the inter-frame correlation quality Q corresponding to each target track v As follows:

[0089]

[0090] In the formula, V n , V n-1 Respectively, the track speed of a target track under the condition of adjacent two frames; the smaller the inter-frame correlation quality, the more stable the target track.

[0091] Preferably, in another embodiment of the application, the step of "S320, calculating the aggregation degree of a plurality of target tracks to obtain the aggregation correlation quality corresponding to each target track" specifically comprises the following steps:

[0092] Obtain the aggregation correlation quality Q corresponding to each target track gm As follows:

[0093]

[0094] In the formula, P is the spatial position of the target track.

[0095] Specifically, in the embodiment, it is generally considered that a great difference between the target point and the background is that the background often extracts multiple tracks, while the target is relatively isolated, so the distance between the track and the track is used as an index to establish the aggregation correlation quality, and the formula is as shown in formula (two), and the aggregation correlation quality is the reciprocal sum of the distance of a track and all other tracks.

[0096] Preferably, in another embodiment of the present application, the step of "S330, obtaining the minimum radio correlation quality corresponding to each target track according to the radio target information and the plurality of target tracks" specifically comprises the following steps:

[0097] interpolating the azimuth in the radio target information and the azimuth in each target track to obtain the minimum radio azimuth correlation quality Q wm as follows:

[0098] Q wm = A m -A w Equation (Three);

[0099] interpolating the speed in the radio target information and the speed in each target track to obtain the minimum radio speed correlation quality Qs wm as follows:

[0100] Qs wm =s m -s w Equation (Four);

[0101] according to the minimum radio azimuth correlation quality Q wm and the minimum radio speed correlation quality Qs wm , calculating the minimum radio correlation quality Q w corresponding to each target track as follows:

[0102] Q w =w1Q wm +w2Qs wm Equation (Five);

[0103] wherein A m is the azimuth of the target track m; A w is the azimuth of the radio target w; s m is the speed of the target track m; s w is the speed of the radio target w; w1 is the weight coefficient of Q wm ; and w2 is the weight coefficient of Qs wm .

[0104] Therefore, the smaller the minimum radio correlation quality Q w , the more consistent the target track is with the radio target information.

[0105] Preferably, in another embodiment of the present application, the step of "S340, obtaining the minimum radar correlation quality corresponding to each target track according to the latest frame of radar target information and the plurality of target tracks" specifically comprises the following steps:

[0106] interpolating the spatial position in the latest frame of radar target information and the spatial position in each target track to obtain the minimum radar spatial correlation quality Q rm as follows:

[0107] Q rm = P m -P r Equation (Six);

[0108] interpolating the speed in the latest frame of radar target information and the speed in each target track to obtain the minimum radar speed correlation quality Qs rm as follows:

[0109] Qs rm = s m -s r Equation (Seven);

[0110] according to the minimum radar spatial correlation quality Q rm and the minimum radar speed correlation quality Qs rm , calculating the minimum radar correlation quality Q r corresponding to each target track as follows:

[0111] Q r = w3Q rm +w4Qs rm Equation (Eight);

[0112] wherein P m is the spatial position of the target track m; P r is the spatial position of the radar target r; s m is the speed of the target track m; s r is the speed of the radar target r; w3 is the weight coefficient of Q rm ; and w4 is the weight coefficient of Qs rm .

[0113] Therefore, the smaller the minimum radar correlation quality Q r , the more consistent the target track is with the radar target information.

[0114] Preferably, in another embodiment of the present application, the "S350, obtaining the minimum historical correlation quality corresponding to each target track according to the historical target track library and the plurality of target tracks" step specifically comprises the following steps:

[0115] interpolating the speed in the historical target track library and the speed in each target track to obtain the minimum historical correlation quality Qs corresponding to each target track with the minimum interpolation result hm As follows:

[0116] Qs hm = s m -s h Equation (IX);

[0117] In the equation, s m is the speed of target track m; s h is the speed of historical target track library target r.

[0118] Specifically, in the embodiment, the information of the historical target track library is used to generate historical track correlation parameters by comparing with the target track. For the track speed that has been identified and confirmed by manual confirmation, its speed is recorded in the historical target track library as a track historical data discrimination feature. The target track is compared with the historical target track library to obtain the minimum historical correlation quality Qs corresponding to each target track with the minimum interpolation result hm .

[0119] Preferably, in another embodiment of the present application, the "S360, obtaining the total confidence corresponding to each target track according to the inter-frame correlation quality, the aggregation correlation quality, the minimum radio correlation quality, the minimum radar correlation quality and the minimum historical correlation quality" step specifically comprises the following steps:

[0120] The total confidence Q corresponding to each target track is obtained as follows:

[0121] Q = w5Q v + w6Q gm + w7Q w + w8Q r + w9Qs hm Equation (X);

[0122] In the equation, w5 is the weight coefficient of the inter-frame correlation quality Q v ; w6 is the weight coefficient of the aggregation correlation quality Q gm ; w7 is the weight coefficient of the minimum radio correlation quality Q w ; w8 is the weight coefficient of the minimum radar correlation quality Q rw9 is a weight coefficient of the minimum historical correlation quality Qs hm w9 is a weight coefficient of the minimum historical correlation quality Qs

[0123] Finally, the total confidence corresponding to all target tracks is sorted, and the target track with the minimum total confidence is selected as the determined target (suspected target).

[0124] Referring to FIG. 1 Figure 2 As shown in FIG. 1, the anti-interference multi-dimensional feature detection method provided by the embodiment of the present application comprises the following steps:

[0125] 1. acquiring a current frame of a to-be-detected image of a to-be-detected target;

[0126] 2. acquiring a plurality of target point tracks based on an SVM classification model according to the to-be-detected image;

[0127] 3. determining a plurality of target tracks in the plurality of target point tracks based on a speed motion function;

[0128] 4. calculating an inter-frame correlation quality corresponding to each target track;

[0129] 5. calculating a clustering degree of the plurality of target tracks to acquire a clustering correlation quality corresponding to each target track;

[0130] 6. acquiring a minimum radio correlation quality corresponding to each target track according to the radio target information and the plurality of target tracks;

[0131] 7. acquiring a minimum radar correlation quality corresponding to each target track according to the latest frame of radar target information and the plurality of target tracks;

[0132] 8. acquiring a minimum historical correlation quality corresponding to each target track according to the historical target track library and the plurality of target tracks;

[0133] 9. acquiring a total confidence corresponding to each target track according to the inter-frame correlation quality, the clustering correlation quality, the minimum radio correlation quality, the minimum radar correlation quality and the minimum historical correlation quality;

[0134] 10. selecting a target track with the minimum total confidence as a determined target.

[0135] Referring to FIG. 1 Figure 3 As shown in FIG. 1, the anti-interference multi-dimensional feature detection system provided by the embodiment of the present application comprises the following steps:

[0136] a data acquisition module, configured to acquire a current frame of a to-be-detected image of a to-be-detected target, the latest frame of radar target information of a current frame, radio target information and a historical target track library;

[0137] The track module is connected in communication with the data acquisition module, and is configured to acquire a plurality of target tracks based on an SVM classification model and a speed motion function according to the image to be detected;

[0138] The confidence module is connected in communication with the data acquisition module and the track module, and is configured to acquire a total confidence corresponding to each target track according to the plurality of target tracks, the radio target information, the latest frame of radar target information and the historical target track library.

[0139] The target determination module is connected in communication with the confidence module, and is configured to select a target track with the minimum total confidence as a determined target.

[0140] Therefore, the present application not only considers the features of images, but also introduces kinematics, track clustering relationship, radar radio target feature and historical track feature, the algorithm calculation amount of each layer is not large, the hardware demand is low, the operation speed is fast, the recognition rate is high, the background interference resistance is strong, and the interpretability is strong.

[0141] Specifically, the present embodiment corresponds to the above-mentioned method embodiment one by one, and the functions of each module have been described in detail in the corresponding method embodiment, and therefore will not be described one by one.

[0142] Based on the same inventive concept, the present application also provides a computer readable storage medium having a computer program stored thereon, and the computer program is executed by a processor to implement all method steps or part of method steps of the above method.

[0143] The present application implements all or part of the above method, and can also be completed by a computer program to instruct related hardware, and the computer program can be stored in a computer readable storage medium, and the computer program can implement the steps of each method embodiment when executed by a processor. The computer program includes computer program code, which can be in the form of source code, object code, executable file or some intermediate form. The computer readable medium can include any entity or device capable of carrying computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the content of the computer readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction, for example, in some jurisdictions, according to legislation and patent practice, the computer readable medium does not include electrical carrier signals and telecommunication signals.

[0144] Based on the same inventive concept, the embodiments of the present application further provide an electronic device, comprising a memory and a processor, the memory storing a computer program running on the processor, and the processor implements all the method steps or part of the method steps in the above method when executing the computer program.

[0145] The processor can be a central processing unit (CPU), and can also be other general-purpose processors, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor and the like, and the processor is a control center of the computer device, and connects all parts of the computer device through various interfaces and lines.

[0146] The memory can be used to store computer programs and / or modules, and the processor realizes various functions of the computer device by running or executing the computer programs and / or modules stored in the memory, and calling data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application program required by a function (such as a sound playing function, an image playing function, etc.); and the data storage area can store data created according to the use of the mobile phone (such as audio data, video data, etc.). In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one magnetic disk storage device, a flash memory device, or other volatile solid-state memory devices.

[0147] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, a server or a computer program product. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Moreover, the present 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 and optical storage, etc.) containing computer-usable program code.

[0148] The present application is described in reference to the appended drawings figures and / or block diagrams of methods, apparatus (systems), servers, and computer program products according to embodiments of the application. It will be understood that each block of the flowchart and / or block diagrams, and combinations of blocks in the flowchart 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, create means for implementing the functions specified in the flowchart and / or block diagram block or blocks. Figure 1 one or more flowcharts and / or blocks Figure 1 one or more flowcharts and / or blocks

[0149] These computer program instructions can also be stored in a computer- readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instructions which implement the function specified in the flowchart and / or block diagram block or blocks. Figure 1 one or more flowcharts and / or blocks Figure 1 one or more flowcharts and / or blocks

[0150] The computer program instructions can also be loaded into a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart and / or block diagram block or blocks. Figure 1 one or more flowcharts and / or blocks Figure 1 one or more flowcharts and / or blocks

[0151] Obviously, numerous modifications and variations of the present application are possible in light of the above teachings. It is therefore to be understood that within the scope of the appended claims and their equivalents, the application can be practiced otherwise than as specifically described.

Claims

1. A method of interference-resistant multi-dimensional feature detection, characterized in that, The method comprises the following steps: Obtaining a current frame to-be-detected image of a to-be-detected target, a latest frame radar target information of the current frame, radio target information and a historical target track library; Obtaining a plurality of target tracks based on an SVM classification model and a speed motion function according to the to-be-detected image; Obtaining a total confidence corresponding to each target track according to the plurality of target tracks, the radio target information, the latest frame radar target information and the historical target track library; Selecting a target track with minimum total confidence as a determined target; The step of obtaining the total confidence corresponding to each target track according to the plurality of target tracks, the radio target information, the latest frame radar target information and the historical target track library comprises the following steps: Calculating an interframe correlation quality corresponding to each target track; Calculating a clustering correlation quality corresponding to each target track by performing clustering degree calculation on the plurality of target tracks; Obtaining a minimum radio correlation quality corresponding to each target track according to the radio target information and the plurality of target tracks; Obtaining a minimum radar correlation quality corresponding to each target track according to the latest frame radar target information and the plurality of target tracks; Obtaining a minimum historical correlation quality corresponding to each target track according to the historical target track library and the plurality of target tracks; Obtaining the total confidence corresponding to each target track according to the interframe correlation quality, the clustering correlation quality, the minimum radio correlation quality, the minimum radar correlation quality and the minimum historical correlation quality.

2. The method of claim 1, wherein, The step of obtaining a plurality of target tracks based on an SVM classification model and a speed motion function according to the to-be-detected image comprises the following steps: Obtaining a plurality of target point tracks based on an SVM classification model according to the to-be-detected image; Determining a plurality of target tracks in the plurality of target point tracks based on a speed motion function.

3. The anti-interference multidimensional feature detection method as described in claim 1, characterized in that, The step of calculating an interframe correlation quality corresponding to each target track comprises the following steps: calculating an inter-frame correlation quality corresponding to each target track as follows: Formula (One); In the formula, , are the track speeds of a target track in two adjacent frames, respectively.

4. The method of claim 1, wherein, The step of calculating a clustering correlation quality corresponding to each target track by performing clustering degree calculation on the plurality of target tracks comprises the following steps: acquiring the corresponding aggregated relevance quality of each target track as follows: Formula (II); In the formula, P is a spatial position of a target track.

5. The method of claim 1, wherein, The step of obtaining a minimum radio correlation quality corresponding to each target track according to the radio target information and the plurality of target tracks comprises the following steps: interpolating between the bearing in the radio target information and the bearing in each of the target tracks to obtain a minimum radio bearing correlation quality corresponding to the minimum interpolation result of each of the target tracks as follows: Formula (3); interpolating between the speed in the radio target information and the speed in each of the target tracks to obtain a minimum radio speed correlation quality for each of the target tracks corresponding to a minimum interpolated result as follows: Formula (IV); According to a minimum radio bearing correlation quality With a minimum radio speed correlation quality , the minimum radio correlation quality corresponding to each of the target track is calculated As follows: Formula (5); wherein is the bearing of the target track m; is the bearing of the radio target w; is the speed of the target track m; is the speed of the radio target w; is the is the weight factor of the is the is the weight factor of the 6. The method of claim 1, wherein, The step of obtaining a minimum radar correlation quality corresponding to each target track according to the latest frame radar target information and the plurality of target tracks comprises the following steps: interpolating between the spatial position in the latest frame of radar target information and the spatial position in each of the target tracks to obtain a minimum radar spatial correlation quality corresponding to an interpolation result minimum of each of the target tracks as follows: Formula (VI); interpolating between the speed in the latest frame of radar target information and the speed in each target track to obtain a minimum radar speed correlation quality corresponding to the minimum interpolation result of each target track as follows: Formula (VII); According to a minimum radar spatial correlation quality With a minimum radar speed correlation quality , the minimum radar correlation quality corresponding to each of the target tracks is calculated as follows: Formula (8); wherein is the spatial position of the target track m; is the spatial position of the radar target r; is the velocity of the target track m; is the velocity of the radar target r; is the weight coefficient of is the weight coefficient of is the weight coefficient of is the weight coefficient of 7. The method of claim 1, wherein, The step of obtaining a minimum historical correlation quality corresponding to each target track according to the historical target track library and the plurality of target tracks comprises the following steps: interpolating between the speeds in the historical target track library and the speed in each of the target tracks to obtain a minimum historical correlation quality for each of the target tracks corresponding to a minimum interpolation result as follows: Formula (IX); wherein is the velocity of the target track m; is the velocity of the historical target track library target r.

8. The method of claim 1, wherein, The step of obtaining the total confidence corresponding to each target track according to the interframe correlation quality, the clustering correlation quality, the minimum radio correlation quality, the minimum radar correlation quality and the minimum historical correlation quality comprises the following steps: acquiring total confidence corresponding to each of the target tracks as follows: Formula (10); wherein is a weight coefficient for the inter-frame correlation quality ; is a weight coefficient for the aggregated correlation quality ; is a weight coefficient for the minimum radio correlation quality ; is a weight coefficient for the minimum radar correlation quality ; is a weight coefficient for the minimum history correlation quality .

9. An interference-resistant multi-dimensional feature detection system characterized by, The method comprises the following steps: The data acquisition module is configured to acquire a current frame to-be-detected image of a to-be-detected target, a most recent frame radar target information of the current frame, radio target information and a historical target track library; The track module is in communication connection with the data acquisition module and is configured to acquire a plurality of target tracks based on an SVM classification model and a speed motion function according to the to-be-detected image; The confidence module is in communication connection with the data acquisition module and the track module and is configured to acquire a total confidence corresponding to each target track according to the plurality of target tracks, the radio target information, the most recent frame radar target information and the historical target track library; The target determination module is in communication connection with the confidence module and is configured to select a target track with minimum total confidence as a determined target; The step of "acquiring a total confidence corresponding to each target track according to the plurality of target tracks, the radio target information, the most recent frame radar target information and the historical target track library" specifically includes the following steps: calculating an inter-frame correlation quality corresponding to each target track; calculating a clustering correlation quality corresponding to each target track by performing clustering degree calculation on the plurality of target tracks; acquiring a minimum radio correlation quality corresponding to each target track according to the radio target information and the plurality of target tracks; acquiring a minimum radar correlation quality corresponding to each target track according to the most recent frame radar target information and the plurality of target tracks; acquiring a minimum historical correlation quality corresponding to each target track according to the historical target track library and the plurality of target tracks; acquiring a total confidence corresponding to each target track according to the inter-frame correlation quality, the clustering correlation quality, the minimum radio correlation quality, the minimum radar correlation quality and the minimum historical correlation quality.

Images