Motion target detection method and device, electronic equipment and storage medium
By segmenting and feature-recognizing the temporal images of the radar imaging system, the problem of decreased performance in detecting weak moving targets was solved, and efficient and accurate target detection was achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- AGRICULTURAL BANK OF CHINA
- Filing Date
- 2022-10-25
- Publication Date
- 2026-06-26
Smart Images
Figure CN115601395B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of image processing technology, and more particularly to methods, apparatus, electronic devices, and storage media for moving target detection. Background Technology
[0002] Moving target detection technology based on radar imaging systems is widely used in military and civilian fields. However, the detection performance of conventional moving target detection methods drops significantly for moving targets with weak echo energy. As radar imaging technology develops towards higher frame rates, weak moving target detection methods based on radar time-series observations are receiving increasing attention because the accumulation of data from continuous observations can effectively improve the detection capability of weak moving targets, which is of great significance for fields such as military reconnaissance and traffic control.
[0003] Traditional moving target indication (MTI) techniques based on radar imaging systems typically utilize the difference in the Doppler spectrum between moving and stationary targets, or the defocusing or blurring characteristics of moving target images, to detect moving targets. However, when the energy of the moving target is very weak, causing it to be submerged in clutter or noise, these techniques often fail to detect the moving target effectively, leading to a decrease in detection accuracy. Summary of the Invention
[0004] This invention provides a method, apparatus, electronic device, and storage medium for detecting moving targets, thereby solving the problem of reduced detection efficiency in the prior art and improving the accuracy of moving target detection.
[0005] In a first aspect, embodiments of the present invention provide a moving target detection method, the method comprising:
[0006] Acquire a temporal image of a preset detection area, and generate a corresponding temporal signal based on the temporal image;
[0007] The time-series signal is segmented to obtain multiple segmented time-series signals, and local features corresponding to each segmented time-series signal are generated respectively.
[0008] Based on the local features, global features of the time-series signal are generated, and feature recognition is performed on the global features to determine the moving target in the preset detection area.
[0009] Optionally, acquiring the temporal image of the preset detection area includes:
[0010] The detection video of the preset detection area within a preset time interval is obtained, and the detection video is parsed to obtain at least one frame of time sequence image of the preset detection area.
[0011] Optionally, generating the corresponding time-series signal based on the time-series image includes:
[0012] The radar signal strength of each pixel in each of the time-series images is obtained sequentially;
[0013] A set of time-series signals corresponding to each of the aforementioned time-series images is generated based on the strength of each of the aforementioned radar signals.
[0014] Optionally, the timing signal is a one-dimensional timing signal;
[0015] The segmentation process of the timing signal yields multiple segmented timing signals, including:
[0016] Obtain a preset sliding window and determine the sliding step size of the sliding window;
[0017] Based on the sliding step size, the one-dimensional time-series signal is segmented to obtain multiple segmented time-series signals corresponding to the one-dimensional time-series signal.
[0018] Optionally, generating the local features corresponding to each of the segmented time-series signals includes:
[0019] For any segmented time-series signal, a preset feature statistics algorithm is obtained, and feature statistics are performed on the current segmented time-series signal based on the feature statistics algorithm to obtain the local features of the current segmented time-series signal; the local features include kernel function features, margin features, kurtosis features, impulse features, peak features and waveform features.
[0020] Optionally, generating global features of the time-series signal based on each of the local features includes:
[0021] Determine the maximum local feature, minimum local feature, and variance local feature of each of the aforementioned local features;
[0022] The global features of the time series signal are determined based on the maximum local features, the minimum local features, and the variance local features.
[0023] Optionally, the step of performing feature recognition on the global features to determine the moving target in the preset detection area includes:
[0024] A pre-trained detection model is obtained, and the global features are detected based on the detection model to determine the moving target in the preset detection area.
[0025] Secondly, embodiments of the present invention also provide a moving target detection device, the device comprising:
[0026] The timing signal generation module is used to acquire a timing image of a preset detection area and generate a corresponding timing signal based on the timing image;
[0027] The local feature determination module is used to segment the time-series signal to obtain multiple segmented time-series signals, and generate local features corresponding to each segmented time-series signal respectively.
[0028] The moving target determination module is used to generate global features of the time-series signal based on the local features, and to perform feature recognition on the global features to determine the moving target in the preset detection area.
[0029] Thirdly, embodiments of the present invention also provide an electronic device, comprising:
[0030] At least one processor; and
[0031] A memory communicatively connected to the at least one processor; wherein,
[0032] The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the moving target detection method according to any embodiment of the present invention.
[0033] Fourthly, embodiments of the present invention also provide a computer-readable storage medium storing computer instructions, which are used to cause a processor to execute and implement the moving target detection method described in any embodiment of the present invention.
[0034] The technical solution of this invention involves acquiring a time-series image of a preset detection area and generating a corresponding time-series signal based on the generated time-series image; segmenting the generated time-series signal to obtain multiple segmented time-series signals, and generating local features corresponding to each generated segmented time-series signal; generating global features of the time-series signal based on each generated local feature, and performing feature recognition on the generated global features to determine the moving target in the preset detection area. This technical solution transforms the moving target detection problem into a one-dimensional transient signal detection problem in the time domain. It eliminates the need to establish complex target motion models and system observation models. Furthermore, one-dimensional transient signal detection only requires attention to changes in the time domain, resulting in high applicability, strong model generalization ability, and a wider range of applicable scenarios. Moreover, by utilizing the obtained moving target detection model, each detection only requires forward extrapolation of the input signal, resulting in low computational load and fast detection speed. Therefore, this invention improves both the detection efficiency and detection quality of target detection.
[0035] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description
[0036] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0037] Figure 1 This is a flowchart of a moving target detection method provided in Embodiment 1 of the present invention;
[0038] Figure 2 This is a schematic diagram of a moving target detection device according to Embodiment 2 of the present invention;
[0039] Figure 3 This is a schematic diagram of the structure of an electronic device that implements the moving target detection method of this invention. Detailed Implementation
[0040] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.
[0041] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in sequences other than those illustrated or described herein.
[0042] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.
[0043] It is understood that before using the technical solutions disclosed in the various embodiments of this disclosure, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in this disclosure in an appropriate manner in accordance with relevant laws and regulations, and user authorization should be obtained.
[0044] For example, upon receiving a user's active request, a prompt message is sent to the user to explicitly inform them that the requested operation will require the acquisition and use of the user's personal information. This allows the user to independently choose whether to provide personal information to the software or hardware, such as the electronic device, application, server, or storage medium performing the operations of this disclosed technical solution, based on the prompt message.
[0045] As an optional but non-limiting implementation, in response to a user's active request, sending a prompt message to the user can be done via a pop-up window, where the prompt message can be presented in text format. Furthermore, the pop-up window can also include a selection control allowing the user to choose "agree" or "disagree" to provide personal information to the electronic device.
[0046] It is understood that the above notification and user authorization process are merely illustrative and do not constitute a limitation on the implementation of this disclosure. Other methods that comply with relevant laws and regulations may also be applied to the implementation of this disclosure.
[0047] It is understood that the data involved in this technical solution (including but not limited to the data itself, the acquisition or use of the data) shall comply with the requirements of relevant laws, regulations and related provisions.
[0048] Example 1
[0049] Figure 1 This is a flowchart illustrating a moving target detection method provided in Embodiment 1 of the present invention. This embodiment is applicable to situations involving the detection of moving targets. The method can be executed by a moving target detection device, which can be implemented in hardware and / or software. This moving target detection device can be configured in a smart terminal or a cloud server. Figure 1 As shown, the method includes:
[0050] S110. Obtain the time sequence image of the preset detection area, and generate the corresponding time sequence signal based on the generated time sequence image.
[0051] In this embodiment of the invention, the moving target detection method can be applied to various detection scenarios such as vehicle detection, marine detection, ship detection, and agricultural and forestry monitoring. Depending on the detection scenario, the content included in the preset detection area varies, and the detected moving targets also differ. Optionally, if detecting moving targets in the ocean, the preset detection area can be the ocean itself, and the detected moving targets may include ships, etc.; if monitoring agriculture and forestry, the preset detection area can be the area where the corresponding agricultural, forestry, or fruit trees are located, and the detected moving targets may include animals in the forest or fruit trees, such as birds. This embodiment does not limit the content of the preset detection area. The time-series image of the preset detection area can be a time-series image acquired within a preset time interval.
[0052] Optionally, the method for obtaining a temporal image of a preset detection region in this embodiment may include obtaining a detection video of the preset detection region within a preset time interval, performing video parsing on the generated detection video, and obtaining at least one frame of a temporal image of the preset detection region.
[0053] Specifically, a preset shooting device captures video of a preset detection area to obtain video data of the preset detection area within the shooting time period. Optionally, the shooting device may include a radar device. The captured video data is then analyzed. Optionally, analysis can be based on content; for example, generating a corresponding image frame when the video content changes, and sorting the generated multiple frames by time to generate a multi-frame time-series image of the preset detection area. Optionally, analysis can also be based on time; for example, generating an image frame for one second of video content, and sorting the generated multiple frames by time to generate a multi-frame time-series image of the preset detection area. The technical solution of this embodiment can also obtain time-series images in other ways, such as directly capturing images within a preset time interval using the shooting device, and sorting the captured images by time to generate a multi-frame time-series image of the preset detection area. It should be noted that the above methods for generating time-series images are all optional embodiments of this embodiment, and this embodiment does not limit the method of obtaining time-series images.
[0054] It should be noted that, based on time-series images generated by a radar device, when the motion of a moving target in the preset detection area is small, or when the size of the moving target in the preset detection area is small, the identification of the target in the image or the target exhibiting small motion becomes more difficult, thus reducing the accuracy of identification. To identify smaller moving targets and targets with small motion, the technical solution of this embodiment transforms the moving target detection problem in the image domain into a one-dimensional transient signal detection problem in the time domain. This fully utilizes the temporal information in the time series of the radar time-series image, further improving the ability to detect moving targets.
[0055] Optionally, the method for generating corresponding time-series signals based on the generated time-series images in this embodiment may include: sequentially acquiring the radar signal strength of each pixel in each generated time-series image; and generating a set of time-series signals corresponding to each generated time-series image based on each generated radar signal strength.
[0056] Specifically, for any given frame of time-series image, the radar signal data of the current time-series image is acquired, and the radar signal intensity of each pixel in the current frame's time-series image is determined based on the radar signal data. The pixels in the current frame's time-series image are sorted, their indices are determined, and a one-dimensional column vector is formed based on these indices. A one-dimensional time-series signal corresponding to the current frame's time-series image is generated based on the radar signal intensity corresponding to each pixel in the column vector. Further, based on the above method, the one-dimensional time-series signal corresponding to each frame's time-series image is determined, and the time-series signals of all time-series images are concatenated to obtain a set of one-dimensional time-series signals corresponding to each time-series image.
[0057] Based on this, moving targets are detected using time-series signals to obtain the final detection results.
[0058] It should be noted that, due to the different background environments and noise levels of different sampled data, the obtained one-dimensional time-series signal needs to be normalized before feature generation to ensure the robustness and compatibility of the subsequent detection results. For example, the normalization method used in this embodiment is as follows:
[0059]
[0060] Where x represents a one-dimensional time-series signal; u represents the mean of the one-dimensional time-series signal x; and σ represents the variance of x.
[0061] S120. The generated time sequence signal is segmented to obtain multiple segmented time sequence signals, and the local features corresponding to each generated segmented time sequence signal are generated respectively.
[0062] Because the target signal is short-lived and very weak, directly generating global statistical features from the signal would result in statistical features dominated by noise and background signals, with minimal impact from the target signal. Conversely, generating local statistical features only by segmenting the signal would lead to feature vectors with very high dimensionality, and the dimensionality of these local statistical feature vectors could not be guaranteed to be consistent across different scenarios. This would make subsequent application of machine learning algorithms extremely difficult, and the trained model would have poor generalization ability. Therefore, this invention employs a two-step feature generation method: first, the signal is segmented, and local statistical features are generated for each segment of the time-series signal; then, based on the generated local statistical features, global statistical features of the time-series signal are further constructed. This approach preserves the information of the target signal without inducing the curse of dimensionality.
[0063] Optionally, the method for segmenting the generated timing signal to obtain multiple segmented timing signals in this embodiment may include: obtaining a preset sliding window and determining the sliding step size of the generated sliding window; performing sliding segmentation on the generated one-dimensional timing signal based on the generated sliding step size to obtain multiple segmented timing signals corresponding to the generated one-dimensional timing signal.
[0064] Specifically, the sliding step size of the sliding window can be set based on the length of the segmented timing signal to be segmented, and the one-dimensional timing signal obtained above can be segmented based on the sliding step size to obtain at least one segmented timing signal.
[0065] Based on this, signal features are generated for each segment of the time sequence signal to obtain the local signals corresponding to each segment of the time sequence signal.
[0066] Optionally, for any segmented time-series signal, the method for generating each local feature of the current segmented time-series signal may include obtaining a preset feature statistics algorithm, performing feature statistics on the current segmented time-series signal based on the feature statistics algorithm, and obtaining the local features of the current segmented time-series signal.
[0067] In this embodiment, local features include kernel function features, margin features, kurtosis features, impulse features, peak features, and waveform features. Correspondingly, different feature statistical algorithms are used depending on the generated local features. Optionally, for kernel function features, the method for statistically analyzing kernel function features may include:
[0068]
[0069] Where x represents the signal in the segmented time sequence signal, and x, l, y represent different positions of the time sequence in the segmented time sequence signal.
[0070] Optionally, for margin features, methods for statistical margin features may include:
[0071]
[0072] Where x represents the signal in the segmented time sequence signal, and x, l, y represent different positions of the time sequence in the segmented time sequence signal.
[0073] Optionally, methods for statistically analyzing kurtosis features may include:
[0074]
[0075] Where x represents the signal in the segmented time sequence signal, and x, l, y represent different positions of the time sequence in the segmented time sequence signal.
[0076] Optionally, for pulse characteristics, methods for statistical pulse characteristics may include:
[0077]
[0078] Where x represents the signal in the segmented time sequence signal, and x, l, y represent different positions of the time sequence in the segmented time sequence signal.
[0079] Optionally, for peak features, methods for statistically analyzing peak features may include:
[0080]
[0081] Where x represents the signal in the segmented time sequence signal, and x, l, y represent different positions of the time sequence in the segmented time sequence signal.
[0082] Optionally, for waveform features, the methods for waveform feature identification may include:
[0083]
[0084] Where x represents the signal in the segmented time sequence signal, and x, l, y represent different positions of the time sequence in the segmented time sequence signal.
[0085] S130. Generate global features of the time-series signal based on each generated local feature, and perform feature recognition on the generated global features to determine the moving target in the preset detection area.
[0086] It should be noted that if the aforementioned local features are directly used to construct global features, the feature vector extracted by each detection unit will have a very large dimension. Furthermore, the dimension of local statistical feature vectors cannot be guaranteed to be consistent across different scenarios. This would make the subsequent object detection model's detection learning during training extremely difficult, and the trained detection model would also have weak generalization ability. To address this issue, this invention utilizes the maximum, minimum, and variance of local statistical features to construct global features.
[0087] Optionally, the method for generating global features of a time series signal based on each generated local feature in this embodiment may include: determining the maximum local feature, minimum local feature, and variance local feature of each generated local feature; and determining the global features of the generated time series signal based on the maximum local feature, minimum local feature, and variance local feature.
[0088] In this embodiment, the maximum local features include the maximum kernel function feature, the maximum margin feature, the maximum kurtosis feature, the maximum impulse feature, the maximum peak value feature, and the maximum waveform feature. The minimum local features include the minimum kernel function feature, the minimum margin feature, the minimum kurtosis feature, the minimum impulse feature, the minimum peak value feature, and the minimum waveform feature.
[0089] Specifically, the kernel function features, margin features, kurtosis features, impulse features, peak features, and waveform features of each segment of the time-series signal are obtained separately. Based on the above local features, the maximum kernel function feature, maximum margin feature, maximum kurtosis feature, maximum impulse feature, maximum peak feature, and maximum waveform feature are determined, as well as the minimum kernel function feature, minimum margin feature, minimum kurtosis feature, minimum impulse feature, minimum peak feature, and minimum waveform feature. Furthermore, variance kernel function features, variance margin features, variance degree features, variance impulse features, variance peak features, and variance waveform features can also be determined based on the above local features. Further, the maximum and minimum values and variances of the local features are used to construct global features. The final global features are shown in the following formula:
[0090]
[0091] Where var, max, and min are used to calculate the variance, maximum value, and minimum value, respectively. It should be noted that the global features constructed in this embodiment are also one-dimensional feature vectors.
[0092] Based on this, the method for determining moving targets in a preset detection region based on global features in this embodiment may include: acquiring a pre-trained detection model, detecting generated global features based on the generated detection model, and determining moving targets in the generated preset detection region.
[0093] Specifically, for the detection model, training samples are acquired, and one-dimensional global features corresponding to the sample images are obtained based on these training samples. The pixels corresponding to the determined global features are then identified, with pixels where the target has passed designated as positive instances and labeled +1, while other pixels are designated as negative instances and labeled -1. A training dataset is then constructed using these pixel labels and the global features, and a binary classifier is trained using Decision Tree and AdaBoost algorithms, ultimately resulting in the trained detection model.
[0094] Specifically, the trained detection model is obtained, and the global features obtained based on the above implementation method are input into the detection model to obtain the detection result output by the model. The detection result may include whether a moving target exists in a preset detection region.
[0095] The technical solution of this invention involves acquiring a time-series image of a preset detection area and generating a corresponding time-series signal based on the generated time-series image; segmenting the generated time-series signal to obtain multiple segmented time-series signals, and generating local features corresponding to each generated segmented time-series signal; generating global features of the time-series signal based on each generated local feature, and performing feature recognition on the generated global features to determine the moving target in the preset detection area. This technical solution transforms the moving target detection problem into a one-dimensional transient signal detection problem in the time domain. It eliminates the need to establish complex target motion models and system observation models. Furthermore, one-dimensional transient signal detection only requires attention to changes in the time domain, resulting in high applicability, strong model generalization ability, and a wider range of applicable scenarios. Moreover, by utilizing the obtained moving target detection model, each detection only requires forward extrapolation of the input signal, resulting in low computational load and fast detection speed. Therefore, this invention improves both the detection efficiency and detection quality of target detection.
[0096] Example 2
[0097] Figure 2 This is a schematic diagram of a moving target detection device provided in Embodiment 2 of the present invention. Figure 2 As shown, the device includes: a timing signal generation module 210, a local feature determination module 220, and a moving target determination module 230;
[0098] The timing signal generation module 210 is used to acquire a timing image of a preset detection area and generate a corresponding timing signal based on the timing image;
[0099] The local feature determination module 220 is used to segment the time sequence signal to obtain multiple segmented time sequence signals, and generate local features corresponding to each segmented time sequence signal respectively.
[0100] The moving target determination module 230 is used to generate global features of the time-series signal based on each of the local features, and to perform feature recognition on the global features to determine the moving target in the preset detection area.
[0101] Based on the above embodiments, optionally, the timing signal generation module 210 includes:
[0102] The temporal image acquisition unit is used to acquire the detection video of the preset detection area within a preset time interval, perform video parsing on the detection video, and obtain at least one frame of temporal image of the preset detection area.
[0103] Based on the above embodiments, optionally, the timing signal generation module 210 includes:
[0104] A radar signal strength acquisition unit is used to sequentially acquire the radar signal strength of each pixel in each of the time-series images;
[0105] The timing signal generation unit is used to generate a set of timing signals corresponding to each timing image based on the strength of each radar signal.
[0106] Based on the above embodiments, optionally, the timing signal is a one-dimensional timing signal;
[0107] Local feature determination module 220 includes:
[0108] A sliding step size determination unit is used to acquire a preset sliding window and determine the sliding step size of the sliding window;
[0109] The segmented timing signal acquisition unit is used to perform sliding segmentation on the one-dimensional timing signal based on the sliding step size to obtain multiple segmented timing signals corresponding to the one-dimensional timing signal.
[0110] Based on the above embodiments, optionally, the local feature determination module 220 includes:
[0111] The local feature acquisition unit is used to acquire a preset feature statistics algorithm for any segmented time-series signal, and perform feature statistics on the current segmented time-series signal based on the feature statistics algorithm to obtain the local features of the current segmented time-series signal; the local features include kernel function features, margin features, kurtosis features, impulse features, peak features and waveform features.
[0112] Based on the above embodiments, optionally, the moving target determination module 230 includes:
[0113] The feature selection unit is used to determine the maximum local feature, minimum local feature, and variance local feature of each of the local features;
[0114] A global feature determination unit is used to determine the global features of the time series signal based on the maximum local feature, the minimum local feature, and the variance local feature.
[0115] Based on the above embodiments, optionally, the moving target determination module 230 includes:
[0116] The moving target determination unit is used to acquire a pre-trained detection model, detect the global features based on the detection model, and determine the moving target in the preset detection area.
[0117] The moving target detection device provided in the embodiments of the present invention can execute the moving target detection method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the method execution.
[0118] Example 3
[0119] Figure 3 A schematic diagram of an electronic device 10 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.
[0120] like Figure 3 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 may also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.
[0121] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.
[0122] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as moving target detection methods.
[0123] In some embodiments, the moving target detection method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the moving target detection method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the moving target detection method by any other suitable means (e.g., by means of firmware).
[0124] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.
[0125] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.
[0126] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.
[0127] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).
[0128] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.
[0129] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.
[0130] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.
[0131] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.
Claims
1. A method for detecting moving targets, characterized in that, include: Acquire a temporal image of a preset detection area, and generate a corresponding temporal signal based on the temporal image; The timing signal is a one-dimensional timing signal; The time-series signal is segmented to obtain multiple segmented time-series signals, and local features corresponding to each segmented time-series signal are generated respectively. Based on the local features, global features of the time-series signal are generated, and feature recognition is performed on the global features to determine the moving target in the preset detection area; The step of generating corresponding time-series signals based on the time-series images includes: sequentially acquiring the radar signal strength of each pixel in each time-series image; and generating a set of time-series signals corresponding to each time-series image based on the radar signal strength. The step of segmenting the timing signal to obtain multiple segmented timing signals includes: obtaining a preset sliding window and determining the sliding step size of the sliding window; and performing sliding segmentation on the one-dimensional timing signal based on the sliding step size to obtain multiple segmented timing signals corresponding to the one-dimensional timing signal. The step of generating local features corresponding to each segmented time-series signal includes: for any segmented time-series signal, obtaining a preset feature statistics algorithm, performing feature statistics on the current segmented time-series signal based on the feature statistics algorithm, and obtaining local features of the current segmented time-series signal; the local features include kernel function features, margin features, kurtosis features, impulse features, peak features, and waveform features; The step of generating global features of the time series signal based on each of the local features includes: determining the maximum local feature, minimum local feature, and variance local feature of each of the local features; and determining the global features of the time series signal based on the maximum local feature, the minimum local feature, and the variance local feature.
2. The method according to claim 1, characterized in that, The acquisition of the time-series image of the preset detection region includes: The detection video of the preset detection area within a preset time interval is obtained, and the detection video is parsed to obtain at least one frame of time sequence image of the preset detection area.
3. The method according to claim 1, characterized in that, The step of performing feature recognition on the global features to determine the moving target in the preset detection area includes: A pre-trained detection model is obtained, and the global features are detected based on the detection model to determine the moving target in the preset detection area.
4. A moving target detection device, characterized in that, include: The timing signal generation module is used to acquire a timing image of a preset detection area and generate a corresponding timing signal based on the timing image; The timing signal is a one-dimensional timing signal; The local feature determination module is used to segment the time-series signal to obtain multiple segmented time-series signals, and generate local features corresponding to each segmented time-series signal respectively. The moving target determination module is used to generate global features of the time-series signal based on each of the local features, and to perform feature recognition on the global features to determine the moving target in the preset detection area; The timing signal generation module includes: a radar signal strength acquisition unit, used to sequentially acquire the radar signal strength of each pixel in each timing image; and a timing signal generation unit, used to generate a set of timing signals corresponding to each timing image based on the radar signal strength. The local feature determination module includes: a sliding step size determination unit, used to acquire a preset sliding window and determine the sliding step size of the sliding window; and a segmented time series signal acquisition unit, used to perform sliding segmentation on the one-dimensional time series signal based on the sliding step size to obtain multiple segmented time series signals corresponding to the one-dimensional time series signal. The local feature determination module includes: a local feature acquisition unit, used to acquire a preset feature statistics algorithm for any segmented time-series signal, and perform feature statistics on the current segmented time-series signal based on the feature statistics algorithm to obtain the local features of the current segmented time-series signal; the local features include kernel function features, margin features, kurtosis features, impulse features, peak features, and waveform features; The moving target determination module includes: a feature selection unit, used to determine the maximum local feature, minimum local feature, and variance local feature of each of the local features; and a global feature determination unit, used to determine the global features of the time-series signal based on the maximum local feature, the minimum local feature, and the variance local feature.
5. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the moving target detection method according to any one of claims 1-3.
6. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the moving target detection method according to any one of claims 1-3.