Radar parameter adaptive point cloud information target detection method and device
By extracting and connecting point cloud features and resolution features of different scales, a point cloud information target detection method with radar parameters adaptive is realized, solving the problems of degradation of detection accuracy and weight applicability, and improving detection accuracy and adaptability.
Patent Information
- Application Number
- CN202411695888.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-25
- Publication Date
- 2025-05-06
AI Technical Summary
In the prior art, when migrating to other data sets using a single data set, the detection accuracy is severely reduced and it cannot adapt to complex road conditions in the real world. Joint training of multiple data sets will increase the computing power requirements and training costs, and the weight obtained is only applicable to multiple data sets of joint training and cannot be adapted to new data sets.
By extracting point cloud features and resolution features of different scales from the point cloud information and radar parameter information of the detection target, and connecting features of the same scale to generate point cloud detection results of the detection target, adaptive processing of input point cloud information and radar parameter information is realized, so that the weights obtained by the detector training on a single data set can be applied to other point cloud data sets.
The adaptability of the detector between different data sets is realized, detection accuracy is improved, computing power requirements and training costs are reduced, and weights are universally applicable to new data sets.
Smart Images
Figure CN119942518A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of point cloud detection technology, and in particular to a target detection method and device for radar parameter adaptive point cloud information. Background Art
[0002] CBDES (Computing Brain Development System) is the core of realizing autonomous driving technology and the commanding height in the development of intelligent connected vehicles. Since the module weights of CBDES functional software are learned in a data-driven manner, it is necessary to build a data closed-loop system to realize data collection, labeling, simulation and functional module updates.
[0003] In related technologies, with the continuous development of autonomous driving perception model algorithms, 3D target detectors can be used to train in the training set of a single point cloud dataset to obtain different weights to achieve target detection; multiple datasets can also be trained jointly, where each dataset corresponds to a set of models, the parameters of the models are interconnected during training, and a unified loss function is used for supervision to achieve target detection.
[0004] However, in related technologies, the method of using a single dataset to evaluate its own validation set works well, but when migrated to other datasets, the detection accuracy will be severely reduced and cannot adapt to the complex road conditions in the real world. The joint training of multiple datasets will increase the computing power requirements and training costs, and the obtained weights are only applicable to multiple datasets of joint training and cannot be adapted to new datasets, which urgently needs to be improved. Summary of the invention
[0005] The present application provides a target detection method and device based on radar parameter adaptive point cloud information to solve the problem that the detection accuracy of the method using a single data set is seriously reduced when migrating to other data sets in the related art, and it cannot adapt to the complex road conditions in the real world, and the joint training of multiple data sets will increase the computing power demand and training cost, and the obtained weights are only applicable to multiple data sets of joint training and cannot be adapted to new data sets.
[0006] The first aspect of the present application provides a target detection method for radar parameter adaptive point cloud information, comprising the following steps: based on the point cloud information of the detection target, obtaining point cloud features of different scales in the point cloud information; based on the radar parameter information of the radar that collects the point cloud information, extracting resolution features of different scales in the radar parameter information; connecting the point cloud features and resolution features of the same scale to obtain modal features of different scales of the detection target; and generating a point cloud detection result of the detection target using the point cloud information and the modal features of different scales.
[0007] Optionally, in one embodiment of the present application, before obtaining point cloud features of different scales in the point cloud information based on the detection target, it also includes: obtaining initial point cloud information of the detection target; performing range alignment and / or height alignment on the initial point cloud information to obtain point cloud information that meets preset conditions.
[0008] Optionally, in one embodiment of the present application, the step of acquiring point cloud features of different scales in the point cloud information based on the detected target includes: encoding and decoding the point cloud information to obtain the point cloud features of different scales.
[0009] Optionally, in one embodiment of the present application, the radar parameter information of the radar that collects the point cloud information is based on, and the resolution features of different scales in the radar parameter information are extracted, including: using the radar parameter information and the point cloud information to generate radar point cloud information of the detected target; upsampling the radar point cloud information to obtain the resolution features of the different scales.
[0010] Optionally, in one embodiment of the present application, before connecting point cloud features and resolution features of the same scale to obtain modal features of different scales of the detection target, it also includes: obtaining point cloud scale information of the point cloud features; obtaining radar scale information of the resolution features; and obtaining the same scale information between the point cloud features and the resolution features based on the point cloud scale information and the radar scale information.
[0011] The second aspect of the present application provides a target detection device for radar parameter adaptive point cloud information, including: a first acquisition module, used to acquire point cloud features of different scales in the point cloud information based on the point cloud information of the detection target; an extraction module, used to extract resolution features of different scales in the radar parameter information based on radar parameter information of the radar that collects the point cloud information; a first generation module, used to connect point cloud features and resolution features of the same scale to obtain modal features of different scales of the detection target; a detection module, used to generate a point cloud detection result of the detection target using the point cloud information and the modal features of different scales.
[0012] Optionally, in one embodiment of the present application, it also includes: a second acquisition module, used to acquire the initial point cloud information of the detection target before acquiring point cloud features of different scales in the point cloud information based on the point cloud information of the detection target; an alignment module, used to perform range alignment and / or height alignment on the initial point cloud information to obtain point cloud information that meets preset conditions.
[0013] Optionally, in one embodiment of the present application, the first acquisition module includes: an encoding and decoding unit, configured to encode and decode the point cloud information to obtain point cloud features of different scales.
[0014] Optionally, in one embodiment of the present application, the extraction module includes: a generation unit, used to generate radar point cloud information of the detected target using the radar parameter information and the point cloud information; and an upsampling unit, used to upsample the radar point cloud information to obtain the resolution features of different scales.
[0015] Optionally, in one embodiment of the present application, it also includes: a third acquisition module, used to obtain point cloud scale information of the point cloud features before connecting point cloud features and resolution features of the same scale to obtain modal features of different scales of the detection target; a fourth acquisition module, used to obtain radar scale information of the resolution features; and a second generation module, used to obtain the same scale information between the point cloud features and the resolution features based on the point cloud scale information and the radar scale information.
[0016] The third aspect of the present application provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the target detection method based on radar parameter adaptive point cloud information as described in the above embodiment.
[0017] The fourth aspect of the present application provides a computer-readable storage medium, which stores a computer program. When the program is executed by a processor, it implements the above radar parameter adaptive point cloud information target detection method.
[0018] The fifth aspect of the present application provides a computer program product, including a computer program, which, when executed, implements the above radar parameter adaptive point cloud information target detection method.
[0019] The embodiment of the present application can extract point cloud features and resolution features of different scales from the acquired point cloud information and radar parameter information of the detection target, respectively, and connect the point cloud features and resolution features of the same scale, and then the modal features of different scales, so as to generate the point cloud detection result of the detection target, and realize the adaptive processing of the input point cloud information and radar parameter information, so that the weights obtained by the detector trained on a single data set can also be used in other point cloud data sets after updating the radar configuration information, and achieve good detection effect, providing a strong guarantee for realizing the CBDES data closure loop and applying it to the real world. Thus, the problem in the related technology that the method of using a single data set seriously reduces the detection accuracy when migrating to other data sets and cannot adapt to the complex road conditions in the real world is solved, and the joint training of multiple data sets will increase the computing power demand and training cost, and the obtained weights are only applicable to multiple data sets of joint training, and cannot be adapted to new data sets.
[0020] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS
[0021] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:
[0022] Figure 1 A flowchart of a target detection method for radar parameter adaptive point cloud information provided according to an embodiment of the present application;
[0023] FIG. 2( a ) is a flow chart showing the working principle of a target detection method based on radar parameter adaptive point cloud information provided in accordance with an embodiment of the present application;
[0024] FIG2( b ) is a block diagram of initial point cloud information provided according to an embodiment of the present application;
[0025] Figure 3 A block diagram of a target detection device for adaptively using point cloud information using radar parameters provided in an embodiment of the present application;
[0026] Figure 4 It is a schematic diagram of the structure of an electronic device provided according to an embodiment of the present application. DETAILED DESCRIPTION
[0027] Embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.
[0028] The following describes the target detection method and device of radar parameter adaptive point cloud information of the embodiment of the present application with reference to the accompanying drawings. In view of the fact that the method using a single data set mentioned in the above background technology seriously reduces the detection accuracy when migrating to other data sets, and cannot adapt to the complex road conditions in the real world, and the joint training of multiple data sets will increase the computing power demand and training cost, and the obtained weights are only applicable to multiple data sets of joint training, and cannot adapt to the problem of being used for new data sets, the present application provides a target detection method of radar parameter adaptive point cloud information, in which point cloud features and resolution features of different scales can be extracted from the acquired point cloud information and radar parameter information of the detection target, and the point cloud features and resolution features of the same scale are connected, and then the modal features of different scales are generated to generate the point cloud detection result of the detection target, realize the adaptive processing of the input point cloud information and radar parameter information, so that the weight obtained by the detector on the single data set training can also be used in other point cloud data sets after updating the radar configuration information, and achieve good detection effect, providing a strong guarantee for realizing the closed loop of CBDES data and applying it to the real world. This solves the problem in the related art that when the method using a single dataset is migrated to other datasets, the detection accuracy is severely reduced and it cannot adapt to the complex road conditions in the real world. The joint training of multiple datasets will increase the computing power demand and training cost, and the obtained weights are only applicable to multiple datasets trained together and cannot be adapted to new datasets.
[0029] Specifically, Figure 1 The present invention provides a flowchart of a target detection method based on radar parameter adaptive point cloud information according to an embodiment of the present application.
[0030] like Figure 1 As shown, the target detection method of radar parameter adaptive point cloud information includes the following steps:
[0031] In step S101, based on the point cloud information of the detection target, point cloud features of different scales in the point cloud information are acquired.
[0032] It is understandable that the embodiments of the present application can divide the point cloud features into different scales, such as 1 / 2, 1 / 4, 1 / 8, 1 / 16 and 1 / 32, etc., and the present application does not impose any specific limitations.
[0033] Furthermore, the embodiments of the present application can perform feature extraction based on point cloud information of different scales, thereby obtaining point cloud features of different scales.
[0034] As a possible implementation method, the embodiment of the present application can obtain point cloud features of different scales based on the point cloud information of the detection target.
[0035] Optionally, in one embodiment of the present application, before obtaining point cloud features of different scales in the point cloud information based on the point cloud information of the detection target, it also includes: obtaining initial point cloud information of the detection target; performing range alignment and / or height alignment on the initial point cloud information to obtain point cloud information that meets preset conditions.
[0036] It can be understood that in the embodiment of the present application, due to the differences in the perception range of the initial point cloud information obtained by different radars, the embodiment of the present application can perform operations such as cropping, padding, and normalization according to the maximum ranging parameters of the radar so that the initial point cloud information has the same range. The specific settings can be made by technicians in this field according to actual conditions, and this application does not impose any specific restrictions.
[0037] That is to say, in some embodiments, the embodiments of the present application can perform range alignment on the initial point cloud information to obtain point cloud information that meets certain conditions. The specific settings can be made by technicians in this field according to actual conditions, and this application does not impose any specific restrictions.
[0038] In addition, it should be noted that due to the height differences of the initial point cloud information obtained by different radars in the embodiment of the present application, translation, rotation, scaling and other operations can be performed according to the radar height so that the initial point cloud information has the same height. The specific setting can be made by technicians in this field according to actual conditions, and this application does not impose any specific restrictions.
[0039] That is to say, in some embodiments, the embodiments of the present application can highly align the initial point cloud information to obtain point cloud information that meets certain conditions. The specific settings can be made by technicians in this field according to actual conditions, and this application does not impose any specific restrictions.
[0040] Optionally, in one embodiment of the present application, based on the point cloud information of the detection target, point cloud features of different scales in the point cloud information are obtained, including: encoding and decoding the point cloud information to obtain point cloud features of different scales.
[0041] As a possible implementation method, the embodiment of the present application can encode and decode point cloud information to obtain point cloud features with scales of 1 / 2, 1 / 4, 1 / 8, 1 / 16 and 1 / 32, etc. The specific scale can be set by technicians in this field according to actual conditions, and this application does not impose any specific restrictions.
[0042] In step S102, based on radar parameter information of the radar that collects the point cloud information, resolution features of different scales in the radar parameter information are extracted.
[0043] It can be understood that in the embodiments of the present application, the radar parameter information may include but is not limited to the maximum radar range, radar altitude, vertical angular resolution, horizontal angular resolution, etc., and the present application does not impose any specific restrictions.
[0044] In the actual implementation process, the embodiment of the present application can first determine the radar that collects point cloud information, and then obtain radar parameter information of different radars, and extract resolution features of different scales therefrom.
[0045] Optionally, in one embodiment of the present application, based on radar parameter information of a radar that collects point cloud information, resolution features of different scales in the radar parameter information are extracted, including: using radar parameter information and point cloud information to generate radar point cloud information of a detected target; and upsampling the radar point cloud information to obtain resolution features of different scales.
[0046] It will be understood by those skilled in the art that, in the embodiments of the present application, radar parameter information may be used as a vector and concatenated with point cloud information to dynamically adjust the weights and parameters of the detection network according to the radar parameters.
[0047] Specifically, in an embodiment of the present application, radar parameter information is represented as a vector, which may include but is not limited to vertical angular resolution, horizontal angular resolution, etc. After being spliced with the point cloud information, the obtained feature layer is continuously upsampled four times by linear interpolation to generate resolution features of five different scales, corresponding to scales of 1 / 2, 1 / 4, 1 / 8, 1 / 16 and 1 / 32, etc., respectively. This application does not make specific restrictions.
[0048] Optionally, in one embodiment of the present application, before connecting point cloud features and resolution features of the same scale to obtain modal features of different scales of the detection target, it also includes: obtaining point cloud scale information of the point cloud features; obtaining radar scale information of the resolution features; and obtaining the same scale information between the point cloud features and the resolution features based on the point cloud scale information and the radar scale information.
[0049] In some embodiments, the embodiments of the present application can obtain point cloud scale information of point cloud features, such as 1 / 2, 1 / 4, 1 / 8, 1 / 16 and 1 / 32, etc., and the present application does not impose any specific limitations.
[0050] In some embodiments, the embodiments of the present application can obtain radar scale information of resolution characteristics, such as 1 / 2, 1 / 4, 1 / 8, 1 / 16 and 1 / 32, etc., which is not specifically limited in the present application.
[0051] In some embodiments, the embodiments of the present application can obtain the same scale information, 1 / 2, 1 / 4, 1 / 8, 1 / 16 and 1 / 32, etc. based on the point cloud scale information and the radar scale information.
[0052] In step S103, point cloud features and resolution features of the same scale are connected to obtain modal features of different scales of the detection target.
[0053] As a possible implementation method, the embodiment of the present application can connect the resolution feature with the point cloud feature of the same scale to obtain modal features of different scales.
[0054] In step S104, point cloud detection results of the detection target are generated using point cloud information and modal features of different scales.
[0055] During the actual implementation process, the embodiment of the present application can take point cloud information as input, and at the same time superimpose modal features of different scales with the point cloud information, perform feature detection, and output the point cloud detection results of the detection target.
[0056] The working principle of the radar parameter adaptive point cloud information target detection method proposed in the embodiment of the present application is described in detail below with reference to a specific embodiment.
[0057] Among them, Figure 2(a) is a flowchart of the working principle of the target detection method based on radar parameter adaptive point cloud information provided according to an embodiment of the present application.
[0058] Step S201: Acquire initial point cloud information and the radar corresponding to the acquisition of the initial point cloud information.
[0059] Step S202: performing range alignment and height alignment on the initial point cloud information to obtain aligned point cloud information.
[0060] Step S203: Encode and decode the point cloud information using an encoder and a decoder to obtain point cloud features of different scales.
[0061] Step S204: Based on the radar parameter information, upsampling is performed four times continuously by linear interpolation to generate resolution features of five different scales.
[0062] Step S205: Connect the resolution feature with the point cloud feature of the same scale to obtain modal features of different scales.
[0063] Step S206: taking the point cloud information and modal features of different scales as input, and performing feature detection through a point cloud detector.
[0064] Step S207: Output the point cloud detection result of the detection target.
[0065] According to the target detection method of radar parameter adaptive point cloud information proposed in the embodiment of the present application, point cloud features and resolution features of different scales can be extracted from the acquired point cloud information and radar parameter information of the detection target, and the point cloud features and resolution features of the same scale can be connected, and then the modal features of different scales can be generated to generate the point cloud detection result of the detection target, and the adaptive processing of the input point cloud information and radar parameter information can be realized, so that the weights obtained by the detector trained on a single data set can also be used in other point cloud data sets after updating the radar configuration information, and achieve good detection effects, providing a strong guarantee for realizing the CBDES data closed loop and applying it to the real world. Thus, the problem that the method of using a single data set in the related technology seriously reduces the detection accuracy when migrating to other data sets and cannot adapt to the complex road conditions in the real world, and the joint training of multiple data sets will increase the computing power demand and training cost, and the obtained weights are only applicable to multiple data sets of joint training, and cannot be adapted to new data sets.
[0066] Next, a target detection device for adaptively detecting point cloud information using radar parameters proposed in an embodiment of the present application will be described with reference to the accompanying drawings.
[0067] Figure 3 A block diagram of a target detection device based on radar parameter adaptive point cloud information provided in an embodiment of the present application.
[0068] like Figure 3 As shown, the radar parameter adaptive point cloud information target detection device 10 includes: a first acquisition module 100, an extraction module 200, a first generation module 300 and a detection module 400.
[0069] The first acquisition module 100 is used to acquire point cloud features of different scales in the point cloud information based on the point cloud information of the detection target.
[0070] The extraction module 200 is used to extract resolution features of different scales in the radar parameter information based on the radar parameter information of the radar that collects the point cloud information.
[0071] The first generating module 300 is used to connect the point cloud features and resolution features of the same scale to obtain modal features of different scales of the detection target.
[0072] The detection module 400 is used to generate a point cloud detection result of a detection target by using point cloud information and modal features of different scales.
[0073] Optionally, in one embodiment of the present application, it also includes: a second acquisition module and an alignment module.
[0074] Among them, the second acquisition module is used to obtain initial point cloud information of the detection target before obtaining point cloud features of different scales in the point cloud information based on the point cloud information of the detection target.
[0075] The alignment module is used to perform range alignment and / or height alignment on the initial point cloud information to obtain point cloud information that meets preset conditions.
[0076] Optionally, in one embodiment of the present application, the first acquisition module 100 includes: an encoding and decoding unit.
[0077] The encoding and decoding units are used to encode and decode point cloud information to obtain point cloud features of different scales.
[0078] Optionally, in one embodiment of the present application, the extraction module 200 includes: a generation unit and an up-sampling unit.
[0079] Among them, the generating unit is used to generate radar point cloud information of the detected target by using radar parameter information and point cloud information.
[0080] The upsampling unit is used to upsample the radar point cloud information to obtain resolution features of different scales.
[0081] Optionally, in one embodiment of the present application, it further includes: a third acquisition module, a fourth acquisition module and a second generation module.
[0082] Among them, the third acquisition module is used to obtain point cloud scale information of the point cloud features before connecting the point cloud features and resolution features of the same scale to obtain modal features of different scales of the detection target.
[0083] The fourth acquisition module is used to acquire radar scale information of resolution characteristics.
[0084] The second generation module is used to obtain the same scale information between the point cloud features and the resolution features based on the point cloud scale information and the radar scale information.
[0085] It should be noted that the above explanation of the embodiment of the target detection method of radar parameter adaptive point cloud information is also applicable to the target detection device of radar parameter adaptive point cloud information of this embodiment, and will not be repeated here.
[0086] According to the target detection device of radar parameter adaptive point cloud information proposed in the embodiment of the present application, point cloud features and resolution features of different scales can be extracted from the acquired point cloud information and radar parameter information of the detection target, and the point cloud features and resolution features of the same scale can be connected, and then the modal features of different scales can be generated to generate the point cloud detection result of the detection target, and the adaptive processing of the input point cloud information and radar parameter information can be realized, so that the weights obtained by the detector trained on a single data set can also be used in other point cloud data sets after updating the radar configuration information, and achieve good detection effect, providing a strong guarantee for realizing the CBDES data closed loop and applying it to the real world. Thus, the problem that the method of using a single data set in the related technology seriously reduces the detection accuracy when migrating to other data sets and cannot adapt to the complex road conditions in the real world, and the joint training of multiple data sets will increase the computing power demand and training cost, and the obtained weights are only applicable to multiple data sets of joint training, and cannot be adapted to new data sets, etc.
[0087] Figure 4 This is a schematic diagram of the structure of an electronic device provided according to an embodiment of the present application. The electronic device may include:
[0088] Memory 401 , processor 402 , and a computer program stored in the memory 401 and executable on the processor 402 .
[0089] When the processor 402 executes the program, the target detection method based on radar parameter adaptive point cloud information provided in the above embodiment is implemented.
[0090] Furthermore, the electronic device further comprises:
[0091] The communication interface 403 is used for communication between the memory 401 and the processor 402 .
[0092] The memory 401 is used to store computer programs that can be executed on the processor 402 .
[0093] The memory 401 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.
[0094] If the memory 401, the processor 402 and the communication interface 403 are implemented independently, the communication interface 403, the memory 401 and the processor 402 can be connected to each other through a bus and communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 4 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.
[0095] Optionally, in a specific implementation, if the memory 401, the processor 402 and the communication interface 403 are integrated on a chip, the memory 401, the processor 402 and the communication interface 403 can communicate with each other through an internal interface.
[0096] The processor 402 may be a central processing unit (CPU), or an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application.
[0097] An embodiment of the present application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above radar parameter adaptive point cloud information target detection method.
[0098] An embodiment of the present application also provides a computer program product, including a computer program, which, when executed, implements the above radar parameter adaptive point cloud information target detection method.
[0099] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or N embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.
[0100] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the features. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise clearly and specifically defined.
[0101] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code comprising one or N executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present application belong.
[0102] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purpose of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or N wirings (electronic devices), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically by optically scanning the paper or other medium and then editing, interpreting or processing in other suitable ways as necessary and then storing it in a computer memory.
[0103] It should be understood that the various parts of the present application can be implemented by hardware, software, firmware or a combination thereof. In the above embodiment, the N steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. If implemented by hardware, as in another embodiment, it can be implemented by any one or a combination of multiple of the following technologies known in the art: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.
[0104] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.
[0105] In addition, each functional unit in each embodiment of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.
[0106] The storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, etc. Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present application. A person of ordinary skill in the art may change, modify, replace and modify the above embodiments within the scope of the present application.
Claims
1. A target detection method based on radar parameter adaptive point cloud information, characterized in that: The following steps are involved: Based on the point cloud information of the detection target, obtaining point cloud features of different scales in the point cloud information; Extracting resolution features of different scales in the radar parameter information based on radar parameter information of the radar that collects the point cloud information; Connecting point cloud features and resolution features of the same scale to obtain modal features of different scales of the detection target; The point cloud information and the modal features of different scales are used to generate a point cloud detection result of the detection target.
2. The method according to claim 1, characterized in that Before acquiring point cloud features of different scales in the point cloud information based on the detected target, the method further includes: Acquire initial point cloud information of the detection target; The initial point cloud information is range aligned and / or height aligned to obtain point cloud information that meets preset conditions.
3. The method according to claim 1, characterized in that: The step of obtaining point cloud features of different scales in the point cloud information based on the point cloud information of the detection target includes: The point cloud information is encoded and decoded to obtain point cloud features of different scales.
4. The method according to claim 1, characterized in that: The extracting resolution features of different scales in the radar parameter information based on the radar parameter information of the radar that collects the point cloud information includes: Generate radar point cloud information of the detected target using the radar parameter information and the point cloud information; The radar point cloud information is upsampled to obtain the resolution features of different scales.
5. The method according to claim 1, characterized in that Before connecting the point cloud features and the resolution features of the same scale to obtain the modal features of the detection target at different scales, the method further includes: Obtaining point cloud scale information of the point cloud feature; Obtaining radar scale information of the resolution feature; The same scale information between the point cloud feature and the resolution feature is obtained based on the point cloud scale information and the radar scale information.
6. A target detection device based on radar parameter adaptive point cloud information, characterized in that: include: A first acquisition module is used to acquire point cloud features of different scales in the point cloud information based on the point cloud information of the detection target; An extraction module, configured to extract resolution features of different scales in the radar parameter information based on radar parameter information of the radar that collects the point cloud information; A generation module, used to connect point cloud features and resolution features of the same scale to obtain modal features of different scales of the detection target; A detection module is used to generate a point cloud detection result of the detection target by using the point cloud information and the modal features of different scales.
7. The device according to claim 6, characterized in that Also includes: A second acquisition module is used to acquire initial point cloud information of the detection target before acquiring point cloud features of different scales in the point cloud information based on the point cloud information of the detection target; The alignment module is used to perform range alignment and / or height alignment on the initial point cloud information to obtain point cloud information that meets preset conditions.
8. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the target detection method based on radar parameter adaptive point cloud information as described in any one of claims 1 to 5.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that: The program is executed by a processor to implement the target detection method of radar parameter adaptive point cloud information as described in any one of claims 1 to 5.
10. A computer program product, characterized in that It comprises a computer program, which, when executed, is used to implement the target detection method of radar parameter adaptive point cloud information as described in any one of claims 1 to 5.