Millimeter wave radar environment rasterization target attribute identification method, device and medium

Abstract

The application discloses a kind of millimeter wave radar environment gridding target attribute identification method, equipment and medium, belong to millimeter wave radar target detection technical field.The method includes: obtaining the driving information of vehicle and the point data of millimeter wave radar;According to driving information, convert multiple frames of point data in point data to current vehicle coordinate system;In current vehicle coordinate system, the measurement area is gridded and gridded feature extraction is carried out;The extracted features are input into the trained target attribute identification model, and the predicted value of the gridding target attribute is obtained.The application summarizes the overall characteristics of the point in the grid at the grid level, and constructs the grid environment features, fully extracts the features in the grid and around the grid, and uses deep learning method to mine the features and establish the relationship between grid features and target attributes, which can effectively identify the characteristics of the target, and provide prior information for subsequent target tracking, identification and feasible area calculation.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of millimeter wave radar target detection, and in particular to a millimeter wave radar environment gridding target attribute identification method, device and medium. BACKGROUND

[0002] In recent years, vehicle-mounted millimeter wave radars have developed rapidly, and their performance has gradually improved. The installation rate of millimeter wave radars in various types of vehicles has also been increasing. In addition to detecting moving targets, millimeter wave radars can also identify obstacles in the environment through multi-frame accumulation detection and be applied to the identification of drivable areas, the construction of electronic maps, and the identification of parking spaces. However, due to the interference of ground reflection noise, multipath reflection noise, frequency doubling noise, and other factors, and the influence of environmental noise, process noise, angle measurement capability, and resolution on the point feature of real targets, the identification of millimeter wave radar target point features and the elimination of false point features have become difficult problems that need to be solved in current applications.

[0003] Currently, in the field of millimeter wave radar target characteristic identification, information from other sensors or simple feature extraction, threshold judgment, and logical stacking are usually used to identify target characteristics. With the rapid development and wide application of deep learning algorithms, some methods using deep learning backbone networks for feature mining have also been proposed. However, the point cloud density of millimeter wave radars is not as good as that of laser radars, and the overall sample size is relatively small. The training effect of deep learning models with multiple layers on small samples is not ideal, which leads to insufficient feature mining by the backbone network. The existing technical solution proposes a method for feature extraction based on gridding, but it still uses the commonly used mean extraction and feature pyramid network method. This method does not extract point feature expansion and multi-factor interaction features, nor does it extract gridding environmental features. Therefore, this method cannot fully mine the information carried by the point itself and the local information of the gridding accumulated by multiple frames. SUMMARY

[0004] To at least partially solve one of the technical problems in the prior art, the purpose of the present application is to provide a millimeter wave radar environment gridding target attribute identification method, device and medium based on deep learning of a millimeter wave radar.

[0005] The first technical solution adopted by the present application is:

[0006] A millimeter wave radar environment gridding target attribute identification method, comprising the following steps:

[0007] Obtaining driving information of a vehicle and point data of a millimeter wave radar;

[0008] According to the driving information, converting multiple frames of point data in the point data into a current vehicle body coordinate system;

[0009] In the current vehicle body coordinate system, the measurement region is rasterized and raster features are extracted;

[0010] The extracted features are input into a trained target attribute recognition model to obtain predicted values of the raster target attributes.

[0011] Further, the driving information includes the vehicle speed V car and the yaw rate Yaw;

[0012] The point trail data includes 11-dimensional features of the azimuth angle θ, the distance Range, the pitch angle φ, the radial speed rSpeed, the horizontal ground speed abSpeed, the point trail coordinates x, y, z, the radar cross section RCS, the amplitude amp, and the signal-to-noise ratio SNR.

[0013] Further, the conversion of the multiple frames of point trail data in the point trail data into the current vehicle body coordinate system according to the driving information includes:

[0014] The motion displacement dx and dy of the vehicle between the previous frame and the current frame are calculated according to the driving information, and the in-vehicle displacement and angle transformation compensation is performed on the multiple frames of historical point trail data;

[0015] The calculation method of the motion displacement dx and dy is as follows:

[0016] dx = V car *T*cos(Yaw*T / 2)

[0017] dy = V car *T*sin(Yaw*T / 2)

[0018] In the formula, T is the frame period.

[0019] Further, the rasterization of the measurement region and the extraction of the raster features include:

[0020] The measurement region is rasterized, and the multiple frames of historical point trails are traversed to establish the relationship between the point trails and the grids according to the coordinates x and y of the point trails;

[0021] All grids are traversed, and for all point trails in each grid, point trail feature extraction and grid overall feature extraction are performed respectively.

[0022] Further, the point trail features are extracted by the following method:

[0023] First, the 11 features of each point trail in the grid are subjected to univariate discretization processing, and the continuous variables are rounded down to obtain 11 new discretized features;

[0024] Secondly, according to the characteristics of the radar irradiation-reflection, multi-factor feature extraction is performed: M interaction features are screened out, assuming that the dimension of the interaction feature is n dimensions, where n<=M, a variable is selected from the M interaction features as the numerator, and n-1 variables are selected from the M-1 interaction features as the denominator, and a total of n is traversed from 2 to M, and a total of MultiVarNum multi-factor features are created:

[0025]

[0026] Further, the grid overall feature is extracted by the following method:

[0027] After extracting the features of the single track in the grid, the features of the multiple tracks in the grid are extracted by the mean value Mean, variance Var, standard deviation Std, maximum value Max, minimum value Min, median value Medium, and kurtosis;

[0028] The local environment features around the grid are obtained, and the environment features include the grid connectivity and the grid similarity; wherein the grid connectivity is used to measure the isolation degree of the grid, and the grid similarity is used to measure the similarity degree of the grid and the surrounding grid.

[0029] Further, the calculation formula of the grid connectivity is as follows:

[0030]

[0031] In the formula, N is the number of grids around the i-th grid; TotalNum is the total number of tracks in the grid; Bool(TotalNum j >0) represents whether the total number of tracks of the j-th grid around is greater than 0, and the result is 1 if it is true, otherwise 0;

[0032] The calculation formula of the grid similarity is as follows:

[0033]

[0034] In the formula, Value i,k is the value of the k-th feature of the i-th grid, and Value j,k is the value of the k-th feature of the j-th grid.

[0035] Further, the target attribute recognition model is a deep learning model, and the target attribute recognition model is trained by the following method:

[0036] The point track data collected by the millimeter wave radar is labeled, a grid is compared with an actual target, a corresponding grid attribute is marked according to a target attribute, and a training set is constructed;

[0037] The target attribute recognition model is trained by using the training set, wherein the input of the target attribute recognition model is the extracted point track features and the grid overall features of each grid, and the output is the association relationship between the grid attribute features and the grid target attributes.

[0038] The second technical scheme adopted by the present application is:

[0039] An electronic device, comprising a processor and a memory, the memory storing at least one instruction, at least one program, a code set or an instruction set, the at least one instruction, the at least one program, the code set or the instruction set being loaded and executed by the processor to implement the above-mentioned millimeter wave radar environment gridding target attribute recognition method.

[0040] The third technical scheme adopted by the present application is:

[0041] A computer-readable storage medium, the storage medium storing at least one instruction, at least one program, a code set or an instruction set, the at least one instruction, the at least one program, the code set or the instruction set being loaded and executed by a processor to implement the above-mentioned millimeter wave radar environment gridding target attribute recognition method.

[0042] The fourth technical scheme adopted by the present application is:

[0043] A computer program product or computer program, the computer program product or computer program comprising computer instructions stored in a computer-readable storage medium. The processor of the computer device can read the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the above-mentioned method.

[0044] The beneficial effects of the present application are: the present application aggregates the overall features of the point tracks in the grid layer, and constructs the grid environment features, fully extracts the features in the grid and around the grid, and further mines the features and establishes the association relationship between the grid features and the target attributes by using the deep learning method, which can effectively identify the characteristics of the target, and provides prior information for subsequent target tracking, identification and feasible area calculation. BRIEF DESCRIPTION OF DRAWINGS

[0045] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following introduces the drawings of the related technical solutions in the embodiments of the present application or the prior art. It should be understood that the drawings in the following introduction are only for the convenience of clearly describing part of the embodiments of the technical solutions of the present application, and those skilled in the art can obtain other drawings according to these drawings without paying creative labor.

[0046] Figure 1 is a flow diagram of a grid-based target attribute recognition method of a millimeter wave radar based on deep learning in an embodiment of the present application.

[0047] Figure 2 is a step flow chart of a grid-based target attribute recognition method of a millimeter wave radar environment in an embodiment of the present application. DETAILED DESCRIPTION

[0048] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the drawings, wherein the same or similar reference signs represent the same or similar elements or elements with the same or similar functions throughout. The embodiments described below by referring to the drawings are exemplary and are only used to explain the present application, and cannot be understood as a limitation of the present application. For the step numbers in the following embodiments, they are only set for the convenience of description and explanation, and the order between the steps is not limited in any way, and the execution order of each step in the embodiments can be adaptively adjusted according to the understanding of those skilled in the art.

[0049] In the description of the present application, it should be understood that the orientation description, such as the orientation or position relationship indicated by up, down, front, back, left, right, etc. is based on the orientation or position relationship shown in the drawings, and is only for the convenience of describing the present application and simplifying the description, and therefore cannot be understood as indicating or implying that the device or element indicated must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation of the present application.

[0050] In the description of the present application, the meaning of several is one or more, and the meaning of multiple is more than two. Greater than, less than, more than, etc. are understood as not including the number, and above, below, etc. are understood as including the number. If it is described as first, second, it is only for the purpose of distinguishing technical features, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features or the order of indicated technical features.

[0051] In the description of the present application, unless otherwise explicitly limited, the words such as setting, installing, connecting, etc. should be broadly understood, and those skilled in the art can reasonably determine the specific meaning of the above words in the present application in combination with the specific content of the technical solutions.

[0052] In view of the existing technical problems, the application provides a grid target attribute recognition method for a millimeter wave radar based on deep learning. First, the method processes received millimeter wave radar point track data, and compensates for position deviation of a plurality of historical point tracks of a previous frame and a current frame based on vehicle driving information using a dead reckoning method. Secondly, at the point track feature level, a plurality of features are created by combining a discretization method and a multi-feature interaction method, thereby fully mining the internal law of the point track and enhancing the robustness of the point track feature. Then, at the grid level, the overall features of the point tracks in the grid are summarized based on a plurality of statistical methods, and a plurality of grid environment features are constructed, thereby fully extracting the features in the grid and around the grid. Finally, the deep learning method is used to further mine the features and establish the mapping relationship between the grid features and the target attributes, so that the trained deep learning model can effectively recognize the characteristics of the target, thereby providing prior information for subsequent target tracking, recognition and feasible area calculation.

[0053] Embodiment 1

[0054] As shown in Figure 1 and Figure 2 , the embodiment provides a millimeter wave radar environment grid target attribute recognition method, which comprises the following steps:

[0055] S1, acquiring driving information of a vehicle and point track data of a millimeter wave radar.

[0056] In some embodiments, the driving information of the vehicle can be acquired by using an existing sensor, wherein the driving information comprises a vehicle speed V car and a yaw rate Yaw.

[0057] In some embodiments, the vehicle-mounted millimeter wave radar scans the environment and acquires millimeter wave radar echo data. First, target point track information is obtained through millimeter wave radar signal processing. Then, the target point track is projected to the heading angle direction of the vehicle, and the current frame point track is converted from a polar coordinate system to a Cartesian coordinate system, thereby obtaining the final point track data, including 11-dimensional features of a bearing angle θ, a distance Range, a pitch angle φ, a radial speed rSpeed, a horizontal ground speed abSpeed, point track coordinates x, y, z, a reflection cross section RCS, an amplitude amp, and a signal-to-noise ratio SNR.

[0058] S2, converting a plurality of point tracks in the point track data to a current vehicle body coordinate system according to the driving information.

[0059] After obtaining the driving information of the vehicle and the point cloud data of the millimeter wave radar, the multi-frame point cloud information is converted to the current vehicle body coordinate system through coordinate conversion and DR trajectory prediction. Specifically, the motion displacement dx and dy of the vehicle between the previous frame and the current frame is calculated, and the displacement and angle transformation compensation of the multi-frame historical point cloud data is performed. dx and dy are calculated as follows:

[0060] dx = V car *T*cos(Yaw*T / 2)

[0061] dy = V car *T*sin(Yaw*T / 2)

[0062] where T is the frame period.

[0063] S3, in the current vehicle body coordinate system, the measurement region is rasterized and the grid features are extracted.

[0064] In some embodiments, step S3 specifically includes the following steps:

[0065] S31: Rasterization. The entire measurement region is rasterized, and the point cloud and grid ownership relationship is established according to the point cloud coordinates x and y.

[0066] S32: Grid feature extraction. All grids are traversed, and for all point clouds in each grid, the grid overall feature is extracted by the grid feature extraction method.

[0067] As an optional implementation, the grid features include point cloud features and grid overall features, and the specific extraction method is as follows:

[0068] A1, point cloud feature extraction

[0069] First, the 11 features of each point cloud in the grid are processed by single variable discretization, and the continuous variables are rounded down to obtain 11 new discrete features, which can enhance the robustness of the original features.

[0070] Secondly, according to the characteristics of the radar's illumination-reflection, multi-factor feature extraction is performed. First, M interactive features are selected, and the following 4 features (M = 4) are selected: reflection cross section RCS, amplitude amp, signal-to-noise ratio SNR and distance Range. The above features can effectively mine the internal feature logic of the millimeter wave radar detection point cloud. The extraction method of multi-factor interactive feature is as follows: assuming that the dimension of interactive feature is n-dimensional (n ≤ M), first select a variable as the numerator from M interactive features, and then select n-1 variables as the denominator from M-1 interactive features, then there are An arrangement combination. The value of n traverses 2 to M, then a total of MultiVarNum multi-factor characteristics will be created:

[0071]

[0072] A2, grid overall feature extraction

[0073] On the basis of single point trace feature extraction within the grid, the grid overall feature and the grid surrounding environment feature are extracted.

[0074] Firstly, the mean, variance, standard deviation, maximum, minimum, median, kurtosis of all features of multiple point traces within the grid are extracted, and in addition, the total number of point traces within the grid is calculated to obtain the overall feature of a single grid point trace.

[0075] Secondly, on the basis of the above, the local environment features around the grid are obtained, and the following several environment features are constructed:

[0076] 1) Grid connectivity, used to measure the isolation degree of the grid; small connectivity indicates that the grid is more likely to be an object edge, isolated false target and small target, and the expression is as follows:

[0077]

[0078] Wherein; N is the number of grids around the ith grid; Bool(TotalNum j >0) indicates whether the total number of point traces of the jth grid around is greater than 0, and the expression result is 1 if it is true, otherwise 0.

[0079] 2) Grid similarity, used to measure the similarity of the grid and the surrounding grid; the higher the similarity, the greater the probability of being the same target, and the expression is as follows:

[0080]

[0081] Wherein, Value i,k is the value of the kth feature of the ith grid.

[0082] S4, input the extracted features into the trained target attribute recognition model to obtain the predicted value of the grid target attribute.

[0083] In the embodiment, the point trace data collected by the millimeter wave radar in real time is rasterized, and the grid feature is extracted. The extracted features are input into the trained deep learning model, so as to obtain the predicted value of the grid target attribute and output, which provides prior information for subsequent target tracking and recognition.

[0084] As an optional implementation, the target attribute recognition model is a deep learning model, and is trained in the following manner: first, the data collected by the millimeter wave radar is labeled, and the grid is compared with the actual target, and the corresponding grid attribute is marked according to the attribute of the target. For example, if the target of this grid is an object that affects vehicle driving (such as a vehicle, a stone pier, a no-parking sign, a railing, etc.), it is marked as a "high" attribute, and if the target does not affect vehicle driving (such as a dropped ground lock, a manhole cover, etc.), it is marked as a "low" attribute, and if the position corresponding to the dot trace has no target, it is marked as a "false" attribute. Second, a deep learning model is used for training, the input of which is the extracted grid features and environmental features of each grid, and the output is the attribute of the grid ("high", "low" and "false" attributes), thereby establishing the association between the grid features and the grid target attributes.

[0085] Embodiment 2

[0086] The embodiments of the present application also provide an electronic device, which includes a processor and a memory, and the memory stores at least one instruction, at least one program, a code set or an instruction set, which are loaded and executed by the processor to implement a millimeter wave radar environment gridding target attribute recognition method as shown in Figure 1 and Figure 2

[0087] It can be understood that the memory can include a random access memory (RAM) and a read-only memory (ROM). Optionally, the memory includes a non-transitory computer-readable storage medium. The memory can be used to store instructions, programs, codes, code sets or instruction sets. The memory can include a program storage area and a data storage area, wherein the program storage area can store instructions for implementing an operating system, instructions for at least one function, instructions for implementing each of the above methods, etc.; and the data storage area can store data created according to the use of the server, etc.

[0088] ​The processor can include one or more processing cores. The processor connects various parts within the entire server by various interfaces and lines, executes various functions of the server and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory, and calling data stored in the memory. Alternatively, the processor can be implemented in at least one of a hardware form of a digital signal processing (DSP), a field-programmable gate array (FPGA), and a programmable logic array (PLA). The processor can be integrated with one or several combinations of a central processing unit (CPU) and a modem. Among them, the CPU mainly processes operating systems and application programs; the modem is used to process wireless communication. It can be understood that the above-mentioned modem can also not be integrated into the processor, but can be realized by a single chip.

[0089] Since the electronic device is an electronic device corresponding to the millimeter wave radar environment gridding target attribute identification method of the embodiments of the present application, and the principle of solving problems of the electronic device is similar to that of the method, the implementation of the electronic device can be referred to the implementation process of the above-mentioned method embodiments, and the repeated parts will not be repeated.

[0090] Embodiment 3

[0091] The embodiments of the present application also provide a computer readable storage medium, wherein at least one instruction, at least one program, a code set or an instruction set are stored in the storage medium, and the at least one instruction, the at least one program, the code set or the instruction set are loaded and executed by a processor to implement a millimeter wave radar environment gridding target attribute identification method as shown in Figure 1 and Figure 2 .

[0092] Those skilled in the art can understand that all or part of the steps of various methods in the above embodiments can be instructed by programs to relevant hardware, and the programs can be stored in a computer readable storage medium, including Read-Only Memory (ROM), Random Access Memory (RAM), Programmable Read-only Memory (PROM), Erasable Programmable Read Only Memory (EPROM), One-time Programmable Read-Only Memory (OTPROM), Electrically-Erasable Programmable Read-Only Memory (EEPROM), Compact Disc Read-Only Memory (CD-ROM) or other optical disk storage, magnetic disk storage, magnetic tape storage, or any other medium that can be used to carry or store data which can be read by a computer.

[0093] Since the storage medium is a storage medium corresponding to the method of identifying the target attribute of the millimeter wave radar environment grid in the embodiments of the present application, and the principle of solving the problem of the storage medium is similar to that of the method, the implementation of the storage medium can refer to the implementation process of the above method embodiments, and the repeated parts will not be described again.

[0094] Embodiment 4

[0095] In some possible implementation manners, various aspects of the method of the embodiments of the present application can also be implemented in the form of a program product, which includes program codes for causing a computer device to execute the steps of the method of identifying the target attribute of the millimeter wave radar environment grid according to various exemplary embodiments of the present application described above in the specification when the program product is run on the computer device. The executable computer program codes or "codes" for executing various embodiments can be written in a high-level programming language such as C, C++, C#, Smalltalk, Java, JavaScript, Visual Basic, Structured Query Language (for example, Transact-SQL), Perl, or in various other programming languages.

[0096] It should be understood that various aspects of the application can be implemented in hardware, software, firmware or a combination of them. In the above embodiments, various steps or methods can be implemented in software or firmware which is stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, and in another embodiment, any of the following technologies, known in the art, or their combination, can be used: discrete logic circuitry having logic gates for implementing logic functions upon an application of data signals, application-specific integrated circuits having appropriate combinational logic gates, programmable gate arrays (PGA), field-programmable gate arrays (FPGA), and so on.

[0097] In the description of the specification, the description of the terms "one embodiment", "some embodiments", "an example", "a specific example", or "some examples" and the like means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In the specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Also, the specific features, structures, materials or characteristics described can be combined in any appropriate manner in any one or more embodiments or examples. Furthermore, the person skilled in the art can combine and combine the different embodiments or examples described in the specification and the features of the different embodiments or examples, without contradiction.

[0098] The above embodiments are only for the purpose of illustrating the technical concept and characteristics of the present application, and the purpose is to enable those skilled in the art to understand the content of the present application and to implement it, and cannot limit the protection scope of the present application. Any equivalent changes or modifications made according to the essence of the present application should be covered within the protection scope of the present application.

Claims

1. A method for identifying target properties of a millimeter wave radar environment gridding, characterized in that, The method comprises the following steps: Obtain driving information of a vehicle and point cloud data of a millimeter wave radar; According to the driving information, convert multiple frames of point cloud data in the point cloud data to a current vehicle body coordinate system; In the current vehicle body coordinate system, perform grid processing on a measurement area and extract grid features; Input the extracted features into a trained target attribute recognition model to obtain predicted values of the grid target attributes; The grid processing and grid feature extraction comprise: Perform grid processing on the measurement area, traverse multiple frames of historical point cloud data, and establish a relationship between the point cloud and the grid according to the coordinates x and y of the point cloud; Traverse all grids, and for all point clouds in each grid, perform point cloud feature extraction and grid overall feature extraction respectively; The point cloud feature is extracted by the following method: Perform univariate discretization processing on the feature of each point cloud in the grid to obtain new discretized features; According to the characteristics of the radar irradiation-reflection, multi-factor feature extraction is performed: M interaction features are screened out, it is assumed that the dimension of the interaction features is n dimensions, where n≤M, a variable is selected from the M interaction features as a numerator, and n-1 variables are selected from the M-1 interaction features as a denominator, and a total of An arrangement combination; the value of n is traversed from 2 to M, and a total of MultiVarNum multi-factor features are created: The grid overall feature is extracted by the following method: After extracting the feature of a single point cloud in the grid, perform feature extraction on the mean, variance, standard deviation, maximum, minimum, median, and kurtosis of all features of multiple point clouds in the grid; Obtain local environmental features around the grid, and the environmental features include grid connectivity and grid similarity; the grid connectivity is used to measure the degree of isolation of the grid, and the grid similarity is used to measure the similarity between the grid and surrounding grids.

2. The method of claim 1, wherein, The travel information includes a vehicle speed V of the vehicle car and a yaw rate Yaw. The point cloud data includes 11-dimensional features of the azimuth angle θ, the distance Range, the elevation angle φ, the radial velocity rSpeed, the horizontal ground speed abSpeed, the point cloud coordinates x, y, z, the reflection cross section RCS, the amplitude amp, and the signal-to-noise ratio SNR.

3. The method of claim 2, wherein, The conversion of multiple frames of point cloud data in the point cloud data to the current vehicle body coordinate system according to the driving information comprises: According to the driving information, calculate the motion displacement dx and dy of the vehicle between the previous frame and the current frame, and compensate for the displacement and angle transformation of the multiple frames of historical point cloud data; The calculation method of the motion displacement dx and dy is as follows: dx = V car *T*cos(Yaw*T / 2) dy = V car *T*sin(Yaw*T / 2) In the formula, T is the frame period.

4. The method of claim 1, wherein, The calculation formula of the grid connectivity is as follows: In the formula, N is the number of grids around the i-th grid; TotalNum is the total number of point traces in the grid; Bool(TotalNum j >0) indicates whether the total number of point traces in the j-th grid around is greater than 0, and the result of the expression is 1 if it is true, and 0 otherwise. The calculation formula of the grid similarity is as follows: In the formula, Value i,k is the value of feature k for the i-th grid, Value j,k is the value of feature k for the j-th grid.

5. The method of claim 1, wherein, The target attribute recognition model is a deep learning model, and the target attribute recognition model is trained by the following method: label the point cloud data collected by the millimeter wave radar, compare the grid with the actual target, mark the grid attribute according to the attribute of the target, and construct a training set; Train the target attribute recognition model using the training set; The input of the target attribute recognition model is the extracted point cloud feature and grid overall feature of each grid, and the output is the association relationship between the attribute feature of the grid and the grid target attribute.

6. An electronic device, comprising: The electronic device comprises a processor and a memory, and the memory stores at least one instruction, at least one program, a code set, or an instruction set, which are loaded and executed by the processor to implement the method according to any one of claims 1 to 5.

7. A computer-readable storage medium, characterized in that, The storage medium stores at least one instruction, at least one program, a code set or an instruction set, which is loaded and executed by the processor to implement the method as claimed in any one of claims 1 to 5.

Images