A point cloud quality detection information generation method, device, equipment and medium

Abstract

The application discloses a point cloud quality detection information generation method, device, equipment and medium. The method comprises the following steps: acquiring at least one frame of point cloud frame corresponding to an acquisition area, wherein the point cloud frame comprises at least one point cloud data; for each point cloud frame, determining the map coordinates corresponding to each point cloud data according to the matching between each point cloud data corresponding to the point cloud frame and a region map; fusing each point cloud data corresponding to each point cloud frame according to the corresponding map coordinates to determine the fused point cloud corresponding to the acquisition area; generating a projection region map according to each point cloud data in the fused point cloud and the corresponding map coordinates of each point cloud data; and detecting the projection region map to determine the point cloud quality detection information. The application can improve the accuracy of point cloud quality detection information generation.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method, apparatus, equipment, and medium for generating point cloud quality inspection information. Background Technology

[0002] With the rapid development of technology, the types of radar have gradually increased, and it has been applied to all aspects of people's daily lives. Radar can emit lasers to detect the spatial position and reflection characteristics of targets, and convert the detection results of each laser beam into a three-dimensional point cloud data. The collected three-dimensional point cloud data can be used to construct a three-dimensional map of the environment.

[0003] Currently, a large amount of 3D point cloud data is first collected by radar, and then the collected 3D point cloud data is analyzed. That is, field data collection is carried out first, followed by indoor processing.

[0004] However, point cloud quality issues can only be discovered during later point cloud data processing, leading to high rework costs and extended project cycles. Summary of the Invention

[0005] This invention provides a method, apparatus, device, and medium for generating point cloud quality inspection information, so as to improve the accuracy of point cloud quality inspection information generation.

[0006] In a first aspect, embodiments of the present invention provide a method for generating point cloud quality detection information, the method comprising:

[0007] Acquire at least one point cloud frame corresponding to the acquisition area, and the point cloud frame includes at least one point cloud data.

[0008] For each point cloud frame, the map coordinates corresponding to each point cloud data are determined by matching the point cloud data corresponding to each point cloud frame with the regional map.

[0009] The point cloud data corresponding to each point cloud frame are fused according to the corresponding map coordinates to determine the fused point cloud corresponding to the collection area.

[0010] Based on the point cloud data and the corresponding map coordinates of each point cloud in the fused point cloud, a projection area map is generated;

[0011] The projected area map is inspected to determine the point cloud quality inspection information.

[0012] Secondly, embodiments of the present invention also provide a point cloud quality detection information generation device, the device comprising:

[0013] The point cloud frame acquisition module is used to acquire at least one point cloud frame corresponding to the acquisition area. The point cloud frame includes at least one point cloud data.

[0014] The data matching module is used to determine the map coordinates corresponding to each point cloud data by matching the point cloud data corresponding to each point cloud frame with the regional map.

[0015] The point cloud fusion module is used to fuse the point cloud data corresponding to each point cloud frame according to the corresponding map coordinates to determine the fused point cloud corresponding to the collection area.

[0016] The projection map acquisition module is used to generate a projection area map based on the point cloud data and the corresponding map coordinates of each point cloud in the fused point cloud.

[0017] The image detection module is used to detect the projected area map and determine the point cloud quality detection information.

[0018] Thirdly, embodiments of the present invention also provide a point cloud quality inspection information generation device, the point cloud quality inspection information generation device comprising:

[0019] At least one processor; and

[0020] A memory that is communicatively connected to at least one processor; wherein,

[0021] The memory stores a computer program that can be executed by at least one processor, such that the at least one processor is able to execute the point cloud quality detection information generation method of any embodiment of the present invention.

[0022] According to another aspect of the present invention, a computer-readable storage medium is provided, which stores computer instructions for causing a processor to execute and implement the point cloud quality detection information generation method of any embodiment of the present invention.

[0023] The technical solution of this invention involves acquiring at least one point cloud frame corresponding to the acquisition area, where each point cloud frame includes at least one point cloud data. For each point cloud frame, the map coordinates corresponding to each point cloud data are determined by matching the point cloud data corresponding to the frame with the regional map. The point cloud data corresponding to each point cloud frame are then fused according to their respective map coordinates to determine the fused point cloud corresponding to the acquisition area. A projection area map is generated based on the point cloud data in the fused point cloud and the corresponding map coordinates. The projection area map is then inspected to determine point cloud quality detection information. By generating a projection area map of the acquisition area using multi-dimensional point cloud data and performing image inspection on the projection area map to determine the point cloud quality, the accuracy of the point cloud data is improved, and invalid point cloud data is avoided, which could lead to a longer project cycle.

[0024] 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

[0025] 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.

[0026] Figure 1 This is a flowchart of a point cloud quality inspection information generation method provided by an embodiment of the present invention;

[0027] Figure 2 This is a flowchart of a point cloud quality inspection information generation method provided by an embodiment of the present invention;

[0028] Figure 3 This is a structural diagram of a point cloud quality detection information generation device according to an embodiment of the present invention;

[0029] Figure 4 This is a schematic diagram of the structure of a point cloud quality detection information generation device provided in an embodiment of the present invention. Detailed Implementation

[0030] 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.

[0031] 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 orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0032] The acquisition, storage, and application of point cloud data involved in the technical solutions of this invention comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0033] Example 1

[0034] Figure 1 This is a flowchart illustrating a method for generating point cloud quality inspection information according to Embodiment 1 of the present invention. This embodiment of the invention is applicable to the generation of point cloud quality inspection information. The method can be executed by a point cloud quality inspection information generation device, which can be implemented in hardware and / or software.

[0035] See Figure 1 The point cloud quality inspection information generation method shown includes:

[0036] S101. Obtain at least one point cloud frame corresponding to the acquisition area. The point cloud frame includes at least one point cloud data.

[0037] The acquisition area can be the geographical range of the target area scanned by the lidar, such as "Area A of the underground parking garage (including 3 pillars and 1 main passage)" or "Plot B of the outdoor park (100m×50m)," etc., which is a pre-defined or real-time scanning range that follows the movement of the equipment. A point cloud frame can be a collection of point cloud data output from a single complete lidar scan (e.g., 360° horizontal rotation), serving as the lidar's "data output unit." Each point cloud frame has a unique identifier (frame ID) and a timestamp, which can be understood as "a complete photograph of the point cloud." Point cloud data can be three-dimensional spatial data generated after a single laser beam detects a target, and is the smallest unit of a point cloud frame. Each point cloud data contains core information: three-dimensional coordinates (X / Y / Z in the sensor's local coordinate system), reflection intensity value, and a timestamp, which can be understood as "a pixel in a point cloud photograph."

[0038] Specifically, the collection area can be defined in one of two ways: The first way is to preset the range: before collection, manually select the area in the software interface (such as circling the area between the entrance and column 1 on the underground parking garage floor plan), or input the geographical coordinate range (such as X: 0-50m, Y: 0-30m); the second way is to follow in real time: when there is no preset range, follow the movement trajectory of the LiDAR device, and dynamically define an area with a radius of 20m as the collection area with the current position of the device as the center (adapted to mobile scanning scenarios). The lidar continuously scans the acquisition area at a fixed frequency (e.g., 10Hz), transmitting raw data packets in real time to the data receiving module via wired (TCP, IP, or USB) or wireless protocols. Each time the lidar completes a 360° horizontal rotation, it generates a "raw data packet," corresponding to the source data of a "point cloud frame." The raw data packet may contain: the ranging value of each laser beam, horizontal / vertical angle, reflection intensity, frame ID, and timestamp, among other low-level information. The data receiving module parses the raw data packet, converting the lidar's raw data (distance and angle) into "three-dimensional coordinates in the sensor's local coordinate system" using trigonometric functions. Combined with the reflection intensity, it generates individual "point cloud data." Invalid data (such as over-range points, no-echo points, and noise points) is filtered out, and the data is further filtered according to the acquisition area, retaining only the point cloud data whose coordinates fall within the acquisition area. These data are then categorized by "frame ID" to form point cloud frames corresponding to the acquisition area. Each frame contains all valid point cloud data within the acquisition area during that scan.

[0039] S102. For each point cloud frame, match the point cloud data corresponding to each point cloud frame with the regional map to determine the map coordinates corresponding to each point cloud data.

[0040] The regional map can be a pre-constructed 3D reference map covering the acquisition area, with a unified world coordinate system, including the spatial structure of the acquisition area (such as pillars and walls), and can be understood as a "3D base map of the acquisition area". The map coordinates can be the 3D coordinates of the point cloud data in the world coordinate system of the regional map, which is different from the sensor's local coordinate system. It is a unified spatial reference for all point cloud data and is used for multi-frame point cloud fusion.

[0041] Specifically, for all point cloud data in a given frame (e.g., frame ID=1), the curvature of each point is calculated, and "edge points" (with extremely high curvature, such as the edge of a pillar) and "planar points" (with extremely low curvature, such as the ground) are selected to form the feature set for the current frame. The corresponding regional map of the acquired area is retrieved, and "edge lines and planes" features (such as existing pillar edge lines and ground planes in the map) consistent with the feature type of the current frame are extracted to form a map feature set. Using the map coordinates of the previous frame's point cloud or the origin of the regional map as a reference, a pose matrix (describing the transformation relationship from the local coordinate system to the world coordinate system) is initialized. For each edge point in the current frame, the nearest edge line is searched in the regional map to establish a "point-line" correspondence; for each planar point in the current frame, the nearest plane is searched in the regional map to establish a "point-plane" correspondence; incorrect matches (such as matches that are too far apart or have excessively large angular deviations) are removed, retaining only valid matches with "highly compatible spatial location and geometric features". Based on effective matching pairs, the optimal pose matrix is ​​solved using a "nonlinear least squares optimization" algorithm. With the objective of minimizing the distance between the current frame's feature points and map features after pose transformation, an error function (such as the perpendicular distance from a point to a line or from a point to a surface) is constructed. Through iterative calculation using a numerical optimization algorithm, the rotation matrix (correcting device posture) and translation vector (correcting device position) that minimize the total error are found and combined to form the final pose matrix. If the error of the pose matrix is ​​less than a preset threshold (e.g., translation error < 1 cm), it is considered valid; otherwise, features are rematched, and the solution is repeated. Using the final pose matrix, all point cloud data of the current frame are transformed from the "sensor local coordinate system" to the "world coordinate system of the regional map": for each point cloud data point (local coordinates Xs, Ys, Zs) in the current frame, its three-dimensional coordinates (Xw, Yw, Zw) in the regional map are calculated according to the transformation rules of the pose matrix; the transformed map coordinates are then bound to the corresponding point cloud data, completing the determination of the map coordinates for all point cloud data in that frame.

[0042] S103. Merge the point cloud data corresponding to each point cloud frame according to the corresponding map coordinates to determine the fused point cloud corresponding to the collection area.

[0043] Among them, the fused point cloud can be a set of non-repeating 3D point clouds covering the entire acquisition area. It is a complete 3D point cloud map of the acquisition area, containing effective spatial information of all frame point clouds, with no duplicate points and no redundant data.

[0044] Specifically, a "fused point cloud set" is created as a container to store the final, non-duplicate point cloud data. Because of the slight measurement error inherent in LiDAR, strict absolute equality is not required; instead, a precision threshold is used to determine "spatial sameness." Therefore, a "coordinate precision threshold" is set (e.g., ±0.02m, meaning the difference between the X / Y / Z coordinates of two points is ≤0.02m, thus considered "same map coordinates"). The point cloud data of all frames with determined map coordinates is traversed, and each point cloud data point is "placed" into its corresponding spatial position within the fused point cloud container according to its map coordinates. For example, the first frame (e.g., frame ID=1) is processed first: the 28,000 point cloud data points of this frame are placed into the fused point cloud container one by one according to their respective map coordinates (Xw, Yw, Zw). At this point, the container has no duplicate data, so it is stored directly. Subsequent frames (e.g., frame ID=2) are then processed: for each point cloud data point in this frame, its map coordinates are read first, and then the fused point cloud container is searched for "whether there is a stored point cloud with a coordinate difference ≤ precision threshold". If the difference between the map coordinates of the current point cloud data and the coordinates of the point cloud already stored in the container is less than or equal to the precision threshold, it is determined to be "same map coordinates". Point cloud data can be deduplicated according to the following rules: Rule 1 (prioritize reflection intensity): retain point cloud data with higher reflection intensity (e.g., if a point in the container already has an intensity of 180, and a new point has an intensity of 200, replace it with the new point); Rule 2 (prioritize latest): if the intensities are similar, retain the point cloud data collected in the latest frame (e.g., the point with frame ID=2 replaces the point with the same coordinates in frame ID=1); Rule 3 (default retention): without special rules, only the first stored point cloud data is retained, and subsequent points with the same coordinates are discarded; if there are no duplicates, store: if the coordinates of the current point cloud data have no matching entries in the container, store it directly in the container. After all frames of point cloud data have been processed, the set of unique point clouds stored in the fused point cloud container is the "fused point cloud corresponding to the collection area". This fused point cloud covers the entire collection area, and each map coordinate corresponds to only one point cloud data.

[0045] S104. Generate a projection area map based on the point cloud data and the corresponding map coordinates of each point cloud in the fused point cloud.

[0046] Among them, the projection area map can be a two-dimensional digital image generated by fusing point cloud projection, that is, a digital orthophoto map (DOM) of the acquisition area, which can intuitively show the coverage and reflection characteristics of the point cloud.

[0047] Specifically, a vertical orthophoto projection is used to project the 3D map coordinates (Xw, Yw, Zw) of each point in the fused point cloud onto the XY plane of the acquisition area. The extreme values ​​of the map coordinates of the fused point cloud are used as boundaries (e.g., Xw minimum 0m, maximum 20m, Yw minimum 0m, maximum 15m) to determine the range of the projection area map (completely consistent with the acquisition area). The pixel resolution of the projection area map is set (e.g., 0.02m / pixel, i.e., 1 pixel corresponds to a 2cm×2cm area in reality). The higher the resolution, the more detailed the image. Based on the projection range and resolution, a two-dimensional raster matrix (i.e., "image canvas") is constructed. If the projection range is Xw: 0-20m, Yw: 0-15m, and the resolution is 0.02m / pixel, then the number of matrix rows = 15 / 0.02 = 750 rows and the number of columns = 20 / 0.02 = 1000 columns. The matrix index is bound to the map coordinates: each raster (row / column index) in the matrix corresponds to a unique XY plane position in reality (e.g., the raster in the 100th row and the 200th column corresponds to the area with Xw = 4m and Yw = 2m). The initial values ​​of all rasters can be empty. The process iterates through each point cloud data point in the fused point cloud and projects it onto the corresponding position in the raster matrix. The map coordinates (Xw, Yw) of the point cloud data are converted into row / column indices of the raster matrix (e.g., Xw=4m, column index = 4 / 0.02 = 200; Yw=2m, row index = 2 / 0.02 = 100). If a raster corresponds to only one point cloud data point, the reflection intensity value of that point is directly assigned to the raster. If a raster corresponds to multiple point cloud data points (because multiple 3D points fall on the same 2D raster after projection), the maximum or average reflection intensity is assigned to the raster. For raster cells in the raster matrix that have "no point cloud data projection" (i.e., missed areas not covered by the point cloud within the acquisition area), preset pixel values ​​are assigned (e.g., 0 for black; or 255 for white), achieving visual distinction between "covered areas" and "missed areas." The filled raster matrix is ​​converted into a standard 2D image format (such as PNG or JPG) in "row / column order": the reflection intensity values ​​(0-255) of the raster are converted into grayscale pixel values ​​of the image (intensity 0 = black, intensity 255 = white, and intermediate values ​​are different grayscale values); the mapping relationship between image pixels and map coordinates is preserved (such as the pixels in the 200th column and 100th row of the image, corresponding to the real-world positions Xw=4m and Yw=2m); the final projection area map is output, which not only preserves the spatial location information of the collected area, but also reflects the reflection characteristics of the point cloud through grayscale, allowing for a direct view of the missed scan areas.

[0048] S105. Detect the projected area map and determine the point cloud quality detection information.

[0049] Among them, point cloud quality inspection information can be structured information about the quality of point cloud data in the collection area, which includes the type of point cloud problem, the location of the point cloud problem, and the rectification strategy of the point cloud scanning method.

[0050] Specifically, two core detection dimensions are pre-defined: Dimension 1: Point cloud coverage integrity, identifying "missed areas without point cloud coverage" in the projected area map. The judgment rule is that a continuous area in the projected area map with a pixel value of "preset data value (such as 0 / black)" and an area ≥ the preset minimum missed area (such as 0.1㎡) is judged as a "missed area"; Dimension 2: Point cloud density rationality, identifying "areas with insufficient point cloud quantity and density in the projected area map". The judgment rule is that the projected area map is divided into detection grids of a fixed size (such as 0.5m×0.5m). If the number of effective point cloud pixels in a grid is < the preset density threshold (such as at least 50 effective pixels in each 0.5m×0.5m grid), it is judged as a "density insufficient area". To achieve refined detection, the projected area map is divided into uniform "detection grids" according to preset sizes: Set the detection grid size: such as 0.5m × 0.5m; Bind the grid to map coordinates: Each detection grid corresponds to a unique geographical range within the collection area (such as the 3rd detection grid corresponding to Xw: 1-1.5m, Yw: 2-2.5m); Calculate the core data of each grid: number of effective pixels (the number of pixels whose reflection intensity ≠ preset data value); Data pixel ratio (number of black pixels / total number of grid pixels). Traverse all detection grids, analyze each grid according to preset rules, and mark the problem type: If the data pixel ratio of a grid is ≥80% and ≥2 consecutive grids form a closed area, mark it as a "missed area" and record the map coordinate range of the area; if the data pixel ratio is <80%, it is judged as "complete coverage" and no marking is required; if the number of effective pixels of a grid is <density threshold, it is marked as "insufficient density area" and the map coordinates, current number of effective pixels, and threshold difference of the area are recorded; if the number of effective pixels is ≥density threshold, it is judged as "density meets the standard" and no marking is required; if the same area meets both "missed area" and "insufficient density", it is marked as "missed area" first. For marked problem areas, the system automatically generates appropriate remedial suggestions: For missed scan areas: output the map coordinates and size of the "re-scan area"; for example, the operator moves to the Xw: 5~8m, Yw: 3~6m area and rescans; For areas with insufficient density: output the current density and threshold difference of the area; for example, "In the Xw: 2~4m, Yw: 4~6m area, increase the radar scanning frequency from 10Hz to 20Hz, or reduce the moving speed to 0.3m / s to rescan." The system integrates "problem type, problem location, and remedial suggestions" into structured point cloud quality inspection information, which can be marked with different colors on the projected area map for missed scans or areas with insufficient density.

[0051] The technical solution of this invention involves acquiring at least one point cloud frame corresponding to the acquisition area, where each point cloud frame includes at least one point cloud data. For each point cloud frame, the map coordinates corresponding to each point cloud data are determined by matching the point cloud data corresponding to the frame with the regional map. The point cloud data corresponding to each point cloud frame are then fused according to their respective map coordinates to determine the fused point cloud corresponding to the acquisition area. A projection area map is generated based on the point cloud data in the fused point cloud and the corresponding map coordinates. The projection area map is then inspected to determine point cloud quality detection information. By generating a projection area map of the acquisition area using multi-dimensional point cloud data and performing image inspection on the projection area map to determine the point cloud quality, the accuracy of the point cloud data is improved, and invalid point cloud data is avoided, which could lead to a longer project cycle.

[0052] Example 2

[0053] Figure 2 This is a flowchart illustrating a method for generating point cloud quality inspection information according to Embodiment 2 of the present invention. Based on the above embodiments, this embodiment optimizes and improves the point cloud quality inspection information generation operation.

[0054] Furthermore, the process of "generating a projection area map based on the point cloud data and the corresponding map coordinates of each point cloud in the fused point cloud" is refined into "determining the map projection matrix based on the map location information corresponding to the regional map; updating the map matrix based on the corresponding map coordinates of each point cloud in the fused point cloud to obtain the updated map matrix; and generating a projection area map based on the map matrix and the reflection values ​​in each point cloud data," in order to improve the operation of generating point cloud quality inspection information.

[0055] It should be noted that for parts not described in detail in the embodiments of the present invention, please refer to the descriptions in other embodiments.

[0056] See Figure 2 The point cloud quality inspection information generation method shown includes:

[0057] S201. Obtain at least one point cloud frame corresponding to the acquisition area. The point cloud frame includes at least one point cloud data.

[0058] S202. Obtain at least one point cloud frame corresponding to the acquisition area. The point cloud frame includes at least one point cloud data.

[0059] S203. Merge the point cloud data corresponding to each point cloud frame according to the corresponding map coordinates to determine the fused point cloud corresponding to the collection area.

[0060] S204. Determine the map matrix based on the map location information corresponding to the regional map.

[0061] The map location information can be core geographic parameters in the regional map, the X, Y, and Z axis coordinates of various feature points (pillars, walls, and ground, etc.) in the regional map under the world coordinate system, and descriptive information about the location of each target point in the regional map, such as the origin of the coordinate system and the axis direction. The map matrix can be a two-dimensional matrix corresponding to the XY horizontal plane of the regional map.

[0062] Specifically, the X / Y / Z axis coordinates of all spatial feature points (such as column vertices, wall inflection points, and ground sampling points) in the regional map are read to form a feature point coordinate set. From this set, the following are selected: X-axis extreme values: Xmin (minimum X-coordinate) and Xmax (maximum X-coordinate), and the horizontal width in the X direction is calculated: WX = Xmax - Xmin; Y-axis extreme values: Ymin (minimum Y-coordinate) and Ymax (maximum Y-coordinate), and the horizontal length in the Y direction is calculated: LY = Ymax - Ymin. The world coordinate system origin coordinates of the regional map (e.g., (X0, Y0, Z0)) and the X / Y axis directions (e.g., X-axis along the main channel of the sampling area, Y-axis perpendicular to the main channel) are recorded to ensure that the matrix spatial orientation is consistent with reality. A standardized two-dimensional matrix is ​​created in memory: a two-dimensional array WX × LY is generated, and all matrix initial values ​​are set to "unfilled flag" (e.g., 0 or null).

[0063] S205. For each point cloud data in the fused point cloud, update the map matrix according to the corresponding map coordinates of the point cloud data to obtain the updated map matrix.

[0064] Specifically, based on the regional map coordinate system baseline (Xmin / Ymin, resolution res), a mapping formula is formulated to achieve accurate conversion from point cloud map coordinates to matrix indexes: Column index (X direction) calculation: col=floor((Xw-Xmin)÷res), (floor() is a floor function to ensure that the coordinates fall within the corresponding raster range, Xw is the map X coordinate of the point cloud data, and res is the set raster resolution), row index (Y direction) calculation: row=floor((Yw-Ymin)÷res), (Yw is the map Y coordinate of the point cloud data). If the calculated col / row exceeds the row and column range of the map matrix, it is determined to be invalid point cloud data and will not participate in matrix updates. The process iterates through all point cloud data in the merged point cloud, extracting the core information of each individual point cloud: map coordinates (Xw, Yw, Zw) and reflection intensity value (Intensity, ranging from 0 to 255). It then calculates the corresponding map matrix raster index (row, col). If the raster's current value is "unfilled" (0 / null), the reflection intensity value (Intensity) of the point cloud is directly assigned to that raster. If the raster already has a value (multiple point clouds mapped to the same raster), the maximum (preferred) or average reflection intensity of all mapped point clouds is taken and assigned to the raster to ensure the raster value reflects the core reflection characteristics of the area. If the point cloud coordinates exceed the area map range (the calculated index exceeds the matrix's row and column count), it is filtered out without updating the matrix. After traversing all point cloud data, for rasteres still marked "unfilled" in the map matrix (i.e., missed areas without point cloud coverage within the collection area), the initial label (e.g., 0) is kept unchanged for later differentiation between "data-rich areas" and "data-free areas." The completed two-dimensional matrix is ​​marked as "updated map matrix" and stored in system memory. The meaning of each grid value in the matrix is ​​as follows: 0 / null: the area corresponding to this grid has no point cloud coverage (missed scan); 1-255: the point cloud reflection intensity value of the area corresponding to this grid (the larger the value, the stronger the target reflectivity).

[0065] S206. Generate a projection area map based on the map matrix and the reflection values ​​in each point cloud data.

[0066] Specifically, all raster values ​​in the map matrix are categorized, and a unique pixel value mapping rule is set for different raster types, with pixel values ​​ranging from 0 to 255. Point cloud reflectance values ​​with raster values ​​between 1 and 255 are directly bound one-to-one with pixel values ​​(e.g., reflectance value 200, pixel value 200). Missed areas with raster values ​​of 0 or null are uniformly mapped to preset data pixel values ​​(usually set to 0, corresponding to black in a grayscale image, or 255, corresponding to white, creating a clear visual contrast with the effective reflectance pixel values). The current raster value is read to determine the raster type (effective reflectance value / unfilled marker); according to the mapping rule, the corresponding pixel value is assigned to the image pixel position corresponding to that raster; the row and column index order of the raster is preserved during the assignment process to ensure that the spatial arrangement of image pixels is completely consistent with the map matrix, thus accurately corresponding to the geographical location of the XY plane of the acquisition area. While generating the image, the bidirectional mapping relationship between the image pixels and the world coordinate system of the regional map is preserved and stored. The pixel array with completed pixel value assignment and preserved coordinate mapping relationship is converted into a visualized two-dimensional image format to generate the final projected area map. The width / height of the image is exactly the same as the number of columns / rows of the map matrix (as shown in the figure, the matrix is ​​750 rows × 1000 columns, and the image is 750 pixels high × 1000 pixels wide).

[0067] S207. Detect the projected area map and determine the point cloud quality detection information.

[0068] This invention determines the map projection matrix based on the map location information corresponding to the regional map; updates the map matrix based on the map coordinates of each point cloud data in the fused point cloud, resulting in an updated map matrix; and generates a projection area map based on the map matrix and the reflection values ​​in each point cloud data, transforming the point cloud data into a visualized projection area map, thereby improving the accuracy of point cloud quality detection.

[0069] Optionally, a projection area map is generated based on the map matrix and the reflection values ​​in each point cloud data, including: finding at least one point cloud location and at least one non-point cloud location in the map matrix; for each point cloud location, determining the reflection value in the point cloud data corresponding to the point cloud location as the image pixel value; for each non-point cloud location, determining a preset pixel value as the image pixel value corresponding to the non-point cloud location; and generating a projection area map based on the image pixel values ​​corresponding to each point cloud location and the image pixel values ​​corresponding to each non-point cloud location.

[0070] Specifically, iterate through the row and column indices (row, col) and corresponding raster values ​​V of each raster; if V ∈ [1, 255] (a point cloud reflection value), add the raster index (row, col) to the point cloud location set and associate it with the corresponding reflection value V; if V = 0 or V = null (an unfilled identifier), add the raster index (row, col) to the non-point cloud location set; based on the two types of location sets, assign image pixel values ​​one by one according to preset rules to ensure that visual features are clearly distinguishable: Point cloud location pixel value assignment: iterate through each raster in the point cloud location set The index (row, col) is used to directly determine the point cloud reflectance value V associated with this location as the corresponding image pixel value (i.e., pixel value = reflectance value). The "raster index - image pixel value" mapping relationship is recorded to ensure a one-to-one correspondence. For non-point cloud location pixel value assignment: each raster index (row, col) in the non-point cloud location set is traversed. A preset pixel value (e.g., 0, corresponding to black in a grayscale image) is determined as the image pixel value for that location. The preset pixel value is a fixed parameter and must form a clear visual contrast with the point cloud location pixel value (e.g., point cloud location pixel value ≥ 1, non-point cloud location = 0). The image pixel values ​​of all locations are integrated according to the spatial arrangement order of the map matrix and converted into a standard two-dimensional image.

[0071] By finding at least one point cloud location and at least one non-point cloud location in the map matrix, for each point cloud location, the reflection value in the point cloud data corresponding to the point cloud location is determined as the image pixel value; for each non-point cloud location, a preset pixel value is determined as the image pixel value corresponding to the non-point cloud location; based on the image pixel values ​​corresponding to each point cloud location and each non-point cloud location, a projection area map is generated. Different data processing methods are performed for different point cloud data, refining the data processing steps and improving the accuracy of data processing.

[0072] Optionally, the map matrix is ​​updated based on the corresponding map coordinates of the point cloud data to obtain the updated map matrix, including: finding the matrix position corresponding to the map coordinates in the map matrix based on the corresponding map coordinates of the point cloud data; obtaining a preset point cloud identifier; and updating the matrix value corresponding to the matrix position through the preset point cloud identifier to obtain the updated map matrix.

[0073] Specifically, based on the basic parameters of the map matrix (Xmin / Ymin, resolution res), a precise conversion formula between coordinates and matrix positions is formulated to ensure that each point cloud map coordinate uniquely matches the matrix position: Core mapping formula: Column index (X direction): col=floor((Xw-Xmin)÷res), Row index (Y direction): row=floor((Yw-Ymin)÷res), (Xw / Yw are the map coordinates of the point cloud data, Xmin / Ymin are the extreme values ​​of the XY axis of the regional map, res is the raster resolution of the map matrix, and floor() is the floor function); Extract the core information of a single point cloud data: map coordinates (Xw, Yw, Zw) and reflection intensity value (Intensity); Calculate the matrix row and column index (row, col) corresponding to the point cloud coordinates using the mapping formula; Verify the validity of the index: If the index is within the matrix row and column range, record the association between "point cloud data - matrix position"; if invalid, mark it as "point cloud outside the collection area" and filter it. Read the pre-configured point cloud labeling rules and obtain the corresponding label values. There are two preset rules (choose one, fixed throughout): Rule 1 (reflection value reuse), which directly takes the reflection intensity value (1-255) of the point cloud data, suitable for scenarios that need to retain the reflection features of the point cloud; and Rule 2 (dedicated label), which uses a fixed value (e.g., 1), suitable for extremely simple scenarios that only need to distinguish between "presence or absence of point clouds". Traverse the "point cloud data - matrix position" association list and locate each valid matrix position (row, col); obtain the label value V corresponding to the point cloud according to the preset rules; replace the original matrix value (initially 0 / null) at the (row, col) position in the matrix with V, or replace the original matrix value with the preset label 1 (indicating that there is point cloud data at the matrix position and that the position has not been missed), and output the updated map matrix.

[0074] By finding the corresponding matrix position in the map matrix based on the map coordinates of the point cloud data, a preset point cloud identifier is obtained; the matrix value corresponding to the matrix position is updated with the preset point cloud identifier to obtain the updated map matrix. Based on the matrix value in the map matrix, it is possible to distinguish which positions in the map matrix were missed.

[0075] Optionally, the projection area map is inspected to determine point cloud quality inspection information, including: dividing the projection area map according to the acquired grid division information to determine at least one image grid; for each image grid, counting the number of point clouds corresponding to the image grid; detecting the number of point clouds corresponding to the image grid to generate first inspection information; detecting the pixel values ​​of each image in the projection area map to determine second inspection information; and determining point cloud quality inspection information based on the first inspection information and the second inspection information.

[0076] Specifically, analyze the grid division information, extract the core parameter of the physical size of the detection grid (such as W = 0.5 m and H = 0.5 m) and the division method (uniform grid); combine the pixel resolution res of the projection area map (such as 0.02 m / pixel) to calculate the pixel size of a single image grid: Pixel width: Colgrid = W÷res (such as 0.5÷0.02 = 25 pixels); Pixel height: Rowgrid = H÷res (such as 0.5÷0.02 = 25 pixels); Divide the projection area map into M×N image grids in the order of "from left to right, from top to bottom"; assign a unique identifier to each image grid (such as grid ID = 1, 2, 3...), and record the coordinate range of the acquisition area map corresponding to the grid (such as grid 1 corresponding to Xw: 0 - 0.5 m, Yw: 0 - 0.5 m). Traverse a single image grid: Locate all pixel positions covered by the image grid; count the number of pixels whose pixel value ≠ the preset data pixel value (such as 0), and this number is the number of point clouds corresponding to the image grid, and store it in the format of "grid ID - number of point clouds - coordinate range", such as grid ID = 5, number of point clouds = 48, coordinates Xw: 1 - 1.5 m, Yw: 2 - 2.5 m. Preset the minimum effective number of point clouds threshold T for a single image grid (such as 50); judge grid by grid: If the number of point clouds ≥ T: Determine it as "density up to standard", and record "grid ID - up to standard - number of point clouds"; If the number of point clouds < T: Determine it as "insufficient density", and record "grid ID - insufficient density - number of point clouds - threshold difference (T - number of point clouds) - coordinate range"; Integrate the judgment results of all grids to generate the first detection information. Preset the detection rule: If the proportion of data pixel values (such as 0) in the image grid is ≥ 80% and at least 2 consecutive grids form a closed area, determine it as a "missed scan area", and record the grid ID, coordinate range and non - data proportion; Generate the second detection information: If the same grid is determined as "missed scan" and "insufficient density" at the same time, it is preferentially marked as "missed scan"; Match the preset rectification suggestions for each problem type (such as missed scan corresponding to rescan, insufficient density corresponding to reducing the radar moving speed, etc. Further, it can also be visually marked, and different - colored marks are used to mark various problem grids on the projection area map (such as missed scan = red and insufficient density = yellow).

[0077] By dividing the projection area map according to the obtained grid division information, at least one image grid is determined; for each image grid, the number of point clouds corresponding to the image grid is counted; the number of point clouds corresponding to the image grid is detected to generate the first detection information; the pixel values of each image in the projection area map are detected to determine the second detection information; according to the first detection information and the second detection information, the point cloud quality detection information is determined, and the point cloud quality is determined through multi - dimensional detection information, which improves the accuracy of point cloud quality detection.

[0078] Optionally, the number of point clouds corresponding to the image raster is detected, and first detection information is generated, including: comparing the number of point clouds corresponding to the image raster with a number threshold to obtain a number comparison result; when the number comparison result is that the number of point clouds is less than or equal to the number threshold, determining the update speed according to the current scanning speed, and determining the first detection information according to the update speed; when the number comparison result is that the number of point clouds is greater than the number threshold, determining the first detection information as empty, and continuing to detect the pixel values ​​of each image in the projection area map to determine the second detection information.

[0079] Specifically, the threshold number T of a single image raster (e.g., 50) and the current scanning speed Vcurr of the LiDAR (e.g., 1.0 m / s, automatically extracted from the acquisition log) are read; the number of point clouds N of the current image raster (e.g., 48) is obtained; if N≤T: the quantity comparison result = "insufficient density", proceed to step 2; if N>T: the quantity comparison result = "density meets the standard", proceed to step 3. Step 2: Insufficient density scenario (N≤T): When it is determined that the density is insufficient, based on the preset "scanning speed-point cloud density" association rule, the core reason for the insufficient density is "scanning speed is too fast" (the faster the speed, the fewer the number of point clouds per unit area); the scanning speed is adjusted in reverse according to the "density gap ratio", the core formula is: Vupdate=Vcurr×TN, (example: Vcurr=1.0m / s, N=48, T=50, Vupdate=1.0×(48 / 50)=0.96m / s; if N=25, T=50, Vupdate=0.5m / s); the update speed must not be lower than At the device's minimum scanning speed (e.g., 0.2 m / s), if the calculated result is lower than the minimum value, the minimum value is taken as the update speed. Integrated parameters generate structured results, with core fields including: Detection dimension: point cloud density; Problem type: insufficient density; Image raster ID / corresponding map coordinates: e.g., raster ID=10, coordinates Xw:1-1.5m / Yw:2-2.5m; Core parameters: current point cloud quantity N, quantity threshold T, difference (T−N); Speed ​​parameters: current scanning speed Vcurr, suggested update speed Vupdate; Rectification suggestion: "Adjust the scanning speed to Vupdate and rescan the area." Density compliance scenario: When the density is determined to be compliant, the first detection information of the image raster is marked as "null / empty," with no problem judgment or rectification suggestion; the step of "detecting the pixel values ​​of each image in the projected area map and determining the second detection information" is automatically executed without interrupting the overall detection chain.

[0080] The number of point clouds corresponding to the image raster is compared with a number threshold to obtain a number comparison result. When the number comparison result is that the number of point clouds is less than or equal to the number threshold, the update speed is determined according to the current scanning speed, and the first detection information is determined according to the update speed. When the number comparison result is that the number of point clouds is greater than the number threshold, the first detection information is determined to be empty, and the detection of each image pixel value in the projection area map is continued to determine the second detection information. When the point cloud data density in the image raster is low, the radar scanning speed is automatically reduced to collect point cloud data and improve the quality of point cloud data.

[0081] Optionally, the second detection information is determined by detecting the pixel values ​​of each image in the projection area map, including: identifying the pixel values ​​of each image in the projection area map to obtain the image pixel position corresponding to at least one preset pixel value; determining the image area based on the pixel positions of each image; determining the rescanning area based on the image area and the area map; and determining the second detection information based on the rescanning area.

[0082] Specifically, read the preset pixel value (e.g., Vnull=0), the pixel resolution res of the projected area map (e.g., 0.02m / pixel), and the coordinate extreme values ​​(Xmin / Ymin) of the area map; traverse each pixel of the projected area map and read its pixel value V; if V=Vnull, record the row and column index (row,col) of the pixel to form a "preset pixel value position list"; if V is different from Vnull, it is determined to be a valid point cloud pixel and skipped. To avoid ineffective remediation caused by scattered pixels, discrete preset pixel values ​​are aggregated into continuous image regions. The region aggregation rule uses "8-neighborhood connectivity" (i.e., adjacent pixels on all sides and diagonally are considered continuous). Pixel positions meeting the following conditions are aggregated into an image region: all pixel values ​​are preset pixel values; pixel positions are continuous (without interruptions); the number of pixels in a single image region is greater than or equal to the preset minimum number of pixels (e.g., 25 pixels, corresponding to a 0.5m × 0.5m physical area, avoiding single-point noise); a unique ID is assigned to each aggregated image region (e.g., region ID = 1, 2...); the boundary pixel positions of each image region are recorded (minimum / maximum row, minimum / maximum col), converting the image region into a physical rescanned region in the world coordinate system of the regional map, achieving accurate mapping of "image features - real-world location": Image Region Boundary Transformation. Map coordinates: Minimum X-coordinate of the image region: Xstart = Xmin + colmin × res; Maximum X-coordinate of the image region: Xend = Xmin + colmax × res; Minimum Y-coordinate of the image region: Ystart = Ymin + rowmin × res; Maximum Y-coordinate of the image region: Yend = Ymin + rowmax × res; Determine the parameters for the rescanned region: Geographic coordinate range: (Xstart, Ystart) − (Xend, Yend); Region area: S = (Xend − Xstart) × (Yend − Ystart); Region center point coordinates: Xcenter = (Xstart + Xend) / 2, Ycenter = (Ystart + Yend) / 2; Compare the rescanned region with the feature points (such as pillars or walls) on the region map to improve on-site identifiability. Integrate all geographic information of the rescanned area to generate structured second detection information: The core fields of the second detection information are: problem type: missed scan (no point cloud coverage), area identifier: a unique code for the rescanned area, geographic information: coordinate range, center point coordinates and area, area description: a natural language description of the map features of the associated area (e.g., "0.5-1.5m west of pillar No. 1 in Garage A area"), and rectification suggestion: "Rescan the rescanned area at a speed of ≤0.5m / s to ensure complete point cloud coverage."

[0083] By identifying the pixel values ​​of each image in the projection area map, the image pixel position corresponding to at least one preset pixel value is obtained; based on the pixel position of each image, the image area is determined; based on the image area and the area map, the rescan area is determined; based on the rescan area, the second detection information is determined, and the missed scan area is rescanned in a timely manner to improve the quality of point cloud data.

[0084] Example 3

[0085] Figure 3 This is a schematic diagram of a point cloud quality inspection information generation device provided in Embodiment 3 of the present invention. This embodiment of the present invention is applicable to the generation of point cloud quality inspection information. The device can execute a point cloud quality inspection information generation method and can be implemented in hardware and / or software.

[0086] See Figure 3 The point cloud quality detection information generation device shown includes: a point cloud frame acquisition module 301, a data matching module 302, a point cloud fusion module 303, a projection map acquisition module 304, and an image detection module 305, wherein...

[0087] The point cloud frame acquisition module 301 is used to acquire at least one point cloud frame corresponding to the acquisition area, and the point cloud frame includes at least one point cloud data.

[0088] The data matching module 302 is used to determine the map coordinates corresponding to each point cloud data by matching the point cloud data corresponding to each point cloud frame with the regional map.

[0089] The point cloud fusion module 303 is used to fuse the point cloud data corresponding to each point cloud frame according to the corresponding map coordinates to determine the fused point cloud corresponding to the collection area.

[0090] The projection map acquisition module 304 is used to generate a projection area map based on the point cloud data in the fused point cloud and the corresponding map coordinates of each point cloud data.

[0091] The image detection module 305 is used to detect the projected area map and determine the point cloud quality detection information.

[0092] The technical solution of this invention involves acquiring at least one point cloud frame corresponding to the acquisition area, where each point cloud frame includes at least one point cloud data. For each point cloud frame, the map coordinates corresponding to each point cloud data are determined by matching the point cloud data corresponding to the frame with the regional map. The point cloud data corresponding to each point cloud frame are then fused according to their respective map coordinates to determine the fused point cloud corresponding to the acquisition area. A projection area map is generated based on the point cloud data in the fused point cloud and the corresponding map coordinates. The projection area map is then inspected to determine point cloud quality detection information. By generating a projection area map of the acquisition area using multi-dimensional point cloud data and performing image inspection on the projection area map to determine the point cloud quality, the accuracy of the point cloud data is improved, and invalid point cloud data is avoided, which could lead to a longer project cycle.

[0093] Optionally, the projection image acquisition module 304 includes:

[0094] The map matrix determination unit is used to determine the map matrix based on the map location information corresponding to the regional map.

[0095] The map matrix update unit is used to update the map matrix based on the map coordinates of each point cloud data in the fused point cloud, so as to obtain the updated map matrix.

[0096] The projection area determination unit is used to generate a projection area map based on the map matrix and the reflection values ​​in each point cloud data.

[0097] Optionally, the projection area determination unit is specifically used for:

[0098] Find at least one point cloud location and at least one non-point cloud location in the map matrix;

[0099] For each point cloud location, the reflection value in the point cloud data corresponding to that location is determined as the image pixel value;

[0100] For each non-point cloud location, a preset pixel value is determined as the image pixel value corresponding to the non-point cloud location;

[0101] A projection area map is generated based on the image pixel values ​​corresponding to each point cloud location and the image pixel values ​​corresponding to each non-point cloud location.

[0102] Optional, map matrix update unit, specifically used for:

[0103] Based on the corresponding map coordinates of the point cloud data, the matrix position corresponding to the map coordinates is found in the map matrix;

[0104] Obtain the preset point cloud identifiers;

[0105] The updated map matrix is ​​obtained by updating the matrix values ​​corresponding to the preset point cloud markers.

[0106] Optionally, the image detection module 305 includes:

[0107] A grid division unit is used to divide the projected area map according to the acquired grid division information to determine at least one image grid.

[0108] The quantity counting unit is used to count the number of point clouds corresponding to each image raster.

[0109] The quantity detection unit is used to detect the number of point clouds corresponding to the image raster and generate the first detection information;

[0110] The pixel value detection unit is used to detect the pixel values ​​of each image in the projection area map and determine the second detection information;

[0111] The information determination unit is used to determine point cloud quality detection information based on the first detection information and the second detection information.

[0112] Optional, quantity detection unit, specifically used for:

[0113] The number of point clouds corresponding to the image raster is compared with the number threshold to obtain the number comparison result;

[0114] When the number comparison result is that the number of point clouds is less than or equal to the number threshold, the update speed is determined according to the current scanning speed, and the first detection information is determined according to the update speed.

[0115] When the number comparison result shows that the number of point clouds is greater than the number threshold, the first detection information is determined to be empty, and the detection of each image pixel value in the projection area map is continued to determine the second detection information.

[0116] Optional, the information determination unit is specifically used for:

[0117] Identify the pixel values ​​of each image in the projection area map to obtain the image pixel position corresponding to at least one preset pixel value;

[0118] Determine the image region based on the position of each pixel in the image;

[0119] Based on the image area and the area map, determine the rescan area;

[0120] The second detection information is determined based on the rescanned area.

[0121] The point cloud quality detection information generation device provided in this embodiment of the invention can execute the point cloud quality detection information generation method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of executing the point cloud quality detection information generation method.

[0122] Example 4

[0123] Figure 4 A schematic diagram of the structure of a point cloud quality inspection information generation device 400 that can be used to implement embodiments of the present invention is shown.

[0124] like Figure 4 As shown, the point cloud quality inspection information generation device 400 includes at least one processor 401 and a memory, such as a read-only memory (ROM) 402 and a random access memory (RAM) 403, communicatively connected to the at least one processor 401. The memory stores computer programs executable by the at least one processor. The processor 401 can perform various appropriate actions and processes based on the computer program stored in the ROM 402 or loaded into the RAM 403 from the storage unit 408. The RAM 403 can also store various programs and data required for the operation of the point cloud quality inspection information generation device 400. The processor 401, ROM 402, and RAM 403 are interconnected via a bus 404. An input / output (I / O) interface 405 is also connected to the bus 404.

[0125] Multiple components in the point cloud quality inspection information generation device 400 are connected to the I / O interface 405, including: an input unit 406, such as a keyboard, mouse, etc.; an output unit 407, such as various types of displays, speakers, etc.; a storage unit 408, such as a disk, optical disk, etc.; and a communication unit 409, such as a network card, modem, wireless communication transceiver, etc. The communication unit 409 allows the point cloud quality inspection information generation device 400 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0126] Processor 401 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 401 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, digital signal processors (DSPs), and any suitable processor, controller, microcontroller, etc. Processor 401 performs the various methods and processes described above, such as the point cloud quality inspection information generation method.

[0127] In some embodiments, the point cloud quality inspection information generation method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 408. In some embodiments, part or all of the computer program may be loaded and / or installed on the point cloud quality inspection information generation device 400 via ROM 402 and / or communication unit 409. When the computer program is loaded into RAM 403 and executed by processor 401, one or more steps of the point cloud quality inspection information generation method described above may be performed. Alternatively, in other embodiments, processor 401 may be configured to perform the point cloud quality inspection information generation method by any other suitable means (e.g., by means of firmware).

[0128] Various implementations 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), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations 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.

[0129] 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.

[0130] 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.

[0131] To provide user interaction, the systems and techniques described herein can be implemented on a point cloud quality inspection information generation device, which includes: 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 point cloud quality inspection information generation device. Other types of devices can also be used to provide user interaction; 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).

[0132] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or 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.

[0133] A computing system can include clients and servers. Clients and servers are generally geographically separated and typically interact via 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. It addresses the shortcomings of traditional physical hosts and VPS (Virtual Private Server) services, such as high management difficulty and weak business scalability.

[0134] 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.

[0135] 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 generating point cloud quality inspection information, characterized in that, The method includes: Acquire at least one point cloud frame corresponding to the acquisition area, wherein the point cloud frame includes at least one point cloud data; For each point cloud frame, the map coordinates corresponding to each point cloud data are determined by matching the point cloud data corresponding to the point cloud frame with the regional map. The point cloud data corresponding to each point cloud frame are fused according to the corresponding map coordinates to determine the fused point cloud corresponding to the collection area. Based on the point cloud data and the corresponding map coordinates of each point cloud data in the fused point cloud, a projection area map is generated. The projected area map is inspected to determine the point cloud quality inspection information.

2. The method according to claim 1, characterized in that, The step of generating a projection area map based on the point cloud data and the corresponding map coordinates of each point cloud in the fused point cloud includes: The map matrix is ​​determined based on the map location information corresponding to the area map; For each point cloud data in the fused point cloud, the map matrix is ​​updated according to the corresponding map coordinates of the point cloud data to obtain the updated map matrix. Based on the map matrix and the reflection values ​​in each point cloud data, a projection area map is generated.

3. The method according to claim 2, characterized in that, The step of generating a projection area map based on the map matrix and the reflection values ​​in each point cloud data includes: At least one point cloud location and at least one non-point cloud location are found in the map matrix; For each of the aforementioned point cloud locations, the reflection value in the point cloud data corresponding to the point cloud location is determined as the image pixel value; For each of the aforementioned non-point cloud locations, a preset pixel value is determined as the image pixel value corresponding to the non-point cloud location; A projection area map is generated based on the image pixel values ​​corresponding to each point cloud location and the image pixel values ​​corresponding to each non-point cloud location.

4. The method according to claim 2, characterized in that, The step of updating the map matrix based on the corresponding map coordinates of the point cloud data to obtain the updated map matrix includes: Based on the corresponding map coordinates of the point cloud data, the matrix position corresponding to the map coordinates is found in the map matrix; Obtain the preset point cloud identifiers; The updated map matrix is ​​obtained by updating the matrix value corresponding to the matrix position using the preset point cloud identifier.

5. The method according to claim 1, characterized in that, The step of detecting the projected area map and determining the point cloud quality detection information includes: The projected area map is divided according to the obtained raster division information to determine at least one image raster. For each of the image gratings, count the number of point clouds corresponding to the image gratings; The number of point clouds corresponding to the image raster is detected, and first detection information is generated; The pixel values ​​of each image in the projection area are detected to determine the second detection information; Based on the first detection information and the second detection information, point cloud quality detection information is determined.

6. The method according to claim 5, characterized in that, The step of detecting the number of point clouds corresponding to the image raster and generating first detection information includes: The number of point clouds corresponding to the image raster is compared with a number threshold to obtain the number comparison result; When the number comparison result is that the number of point clouds is less than or equal to the number threshold, the update speed is determined according to the current scanning speed, and the first detection information is determined according to the update speed. When the number comparison result is that the number of point clouds is greater than the number threshold, the first detection information is determined to be empty, and the detection of each image pixel value in the projection area map is continued to determine the second detection information.

7. The method according to claim 5, characterized in that, The step of detecting the pixel values ​​of each image in the projection area map to determine the second detection information includes: The image pixel values ​​in the projection area are identified to obtain the image pixel position corresponding to at least one preset pixel value; The image region is determined based on the position of each image pixel; Based on the image region and the region map, determine the rescan region; The second detection information is determined based on the rescanned area.

8. A point cloud quality inspection information generation device, characterized in that, The device includes: The point cloud frame acquisition module is used to acquire at least one point cloud frame corresponding to the acquisition area, wherein the point cloud frame includes at least one point cloud data. The data matching module is used to determine the map coordinates corresponding to each point cloud data by matching the point cloud data corresponding to each point cloud frame with the regional map. The point cloud fusion module is used to fuse the point cloud data corresponding to each point cloud frame according to the corresponding map coordinates to determine the fused point cloud corresponding to the collection area. The projection map acquisition module is used to generate a projection area map based on the point cloud data and the corresponding map coordinates of each point cloud data in the fused point cloud. The image detection module is used to detect the projected area map and determine the point cloud quality detection information.

9. A point cloud quality inspection information generation device, characterized in that, The point cloud quality inspection information generation 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 point cloud quality inspection information generation method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the point cloud quality detection information generation method according to any one of claims 1-7.

Images