Sample standard deviation measuring method based on random sampling

Abstract

The invention relates to the field of data processing of probability statistics, and particularly discloses a sample standard deviation measuring method based on random sampling, which comprises the following steps of: performing equal-proportion scaling on a positive photograph and a point cloud image generated by an IGV mapping algorithm in an elevation direction to form an overlook projection drawing as a sample image; performing irregular linear segmentation on the sample image to obtain a plurality of sample blocks; calculating a weight parameter of each sample block; randomly selecting a proper amount of point clouds from the sample blocks as sample points according to the weight parameters; according to differences of angular points and boundary points, the distance from a point to a line or from a point to a center is taken, and the standard deviation is calculated; according to the invention, the robustness and the accuracy of a detection result are improved. Meanwhile, the linear segmentation design provided by the invention is very ingenious, is not restricted by a specific map, and can be easily transplanted to other maps, so that the application range is very wide.

Description

Technical Field

[0001] The present invention relates to the field of probability statistics data processing, and in particular to a sample standard deviation measurement method based on random sampling. Background Art

[0002] Map accuracy verification technology is widely used in many fields such as autonomous driving, industrial automation, and 3D reconstruction. Map accuracy verification based on random sampling is one of the typical solutions. Although random sampling is prone to the problem of clustering points, its statistical range can cover most samples. Sampling schemes based on random samples often require a large amount of data as a basis. Generally speaking, the larger the amount of data, the better the sampling effect, and the final measurement result can be closer to the true value. However, the number of points in the actual point cloud map is too large, so the result of direct sampling will result in a very long inspection time each time.

[0003] When the size of the data set is limited, existing solutions usually alleviate the problem of uneven distribution of selected points through some voxel filtering and regularization. First, determine the size of the voxel. This size is selected based on the specific application scenario and the characteristics of the point cloud data. For example, if the point cloud represents a large outdoor scene, the voxel size can be set relatively large; if it is a detailed indoor object scan, the voxel size needs to be set smaller.

[0004] Space division:

[0005] The entire 3D point cloud space is divided into a regular cubic grid with the selected voxel size as the unit. Each voxel has a spatial range, for example, in the Cartesian coordinate system, a voxel may be from (x min ,y min ,z min ) to (x max ,y max ,z max ) is a small cube.

[0006] Point cloud aggregation and downsampling:

[0007] For each voxel, check the points it contains. Different strategies can be used to process these points.

[0008] A common strategy is to calculate the centroid (center of mass) of all points in a voxel. For example, if there are points P1(x1,y1,z1),P2(x2,y2,z2)…P n (x n ,y n ,z n ), then the centroid C(x c ,y c ,z c) is calculated as:

[0009]

[0010] This centroid point is then used to represent the points in the entire voxel, thereby achieving downsampling. Another strategy is to randomly select a point in the voxel to represent all the points in the voxel. This method has a relatively small amount of calculation, but may lose some detailed information of the point cloud. Summary of the invention

[0011] In view of this, the present invention provides a sample standard deviation measurement method based on random sampling, which can not only evenly distribute the sampling points by pre-linear segmentation, but also increase the reliability of the results by random sampling, thereby improving the robustness and accuracy of the results.

[0012] The present invention adopts the following technical solution:

[0013] The sample standard deviation measurement method based on random sampling includes the following steps:

[0014] S1: Using ArcMap, the orthophoto image (result.tif) and the point cloud image generated by the IGV mapping algorithm (mapping) are scaled in the elevation direction to form a bird's-eye view projection as a sample image;

[0015] S2: performing irregular linear segmentation on the sample image formed in step S1 to obtain a number of sample blocks;

[0016] S3: Calculate the weight parameter of each sample block;

[0017] S4: According to the weight parameter, randomly select an appropriate amount of point clouds in the sample block as sample points;

[0018] S5: According to the difference between corner points and boundary points, take the distance from point to line or from point to center and calculate its standard deviation;

[0019]

[0020] As a further technical solution of the present invention, in step S3, the weight parameter of each sample block is calculated using the following formula:

[0021]

[0022] Where D is the weight of the sample block, N is the number of samples, I is the reflection intensity of the lidar, and F is the impact factor.

[0023] As a further technical solution of the present invention, in step S5, there are two criteria for calculating the final error:

[0024] Method 1: Select a point x i , the selection range is the entire point cloud, and the average value of the entire point cloud is calculated. The calculation result is recorded as Calculate the difference between them. After doing the same operation for all points in the point cloud, add the difference and take the square root after dividing by n-1.

[0025] Method 2: Select a point x i Directly subtract the true value, calculate the error between the two, add the errors between multiple points, get the total error of the samples taken, and finally divide it by the number of samples to get the average value of the sample error.

[0026] Compared with the prior art, the present invention has the following beneficial effects:

[0027] The present invention provides a sample standard deviation measurement method based on random sampling. On the basis of the existing methods at this stage, linear segmentation and random sampling schemes are added. That is, the selected points can be evenly distributed on the entire image by linear segmentation, and the overall randomness can be increased by random sampling, which plays a role in improving the robustness and accuracy of the detection results. At the same time, the linear segmentation design proposed by the present invention is very clever, not constrained by a specific map, and can be easily transplanted to other maps, so it has a wide range of applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 It is a top view projection diagram in an embodiment of the present invention.

[0029] Figure 2 It is a linear segmentation diagram in an embodiment of the present invention.

[0030] Figure 3 It is a point cloud random sampling diagram in an embodiment of the present invention.

[0031] Figure 4 This is the algorithm flow chart of the sample standard deviation measurement method based on random sampling. DETAILED DESCRIPTION

[0032] The following is a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with specific embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0033] The present invention divides the image into several blocks by linear segmentation. By random sampling, 1,000 points are randomly selected from each segmented image, for a total of 10,000 sample points. We add weight parameters to each point according to the type of point and the importance of the image it is in, so that it has different effects on the final accuracy test.

[0034] See attached Figure 1-4 As shown, the embodiment of the present invention proposes a sample standard deviation measurement method based on random sampling, comprising the following steps:

[0035] S1: Using ArcMap, the orthophoto image of Wuhu Port (result.tif) and the point cloud image generated by the IGV mapping algorithm (mapping) are scaled in the elevation direction to form a bird's-eye view projection as a sample image. Figure 1 As shown;

[0036] S2: Perform irregular linear segmentation on the sample image formed in step S1 to obtain several sample blocks, such as Figure 2 As shown;

[0037] S3: Calculate the weight parameter of each sample block using the following formula:

[0038]

[0039] Where D is the weight of the sample block, N is the number of samples, I is the reflection intensity of the lidar, and F is the impact factor.

[0040] S4: According to the weight parameter, randomly select an appropriate amount of point clouds in the sample block as sample points, such as Figure 3 As shown;

[0041] S5: According to the difference between corner points and boundary points, take the distance from point to line or from point to center and calculate its standard deviation;

[0042] In this embodiment, in step S5, there are two criteria for calculating the final error:

[0043] Method 1: Select a point x i , the selection range is the entire point cloud, and the average value of the entire point cloud is calculated. The calculation result is recorded as Calculate the difference between them. After doing the same operation for all points in the point cloud, add the difference and take the square, divide by n-1, and take the square root, as shown in the following formula:

[0044]

[0045] Method 2: Select a point x iDirectly subtract the true value, calculate the error between the two, add the errors between multiple points, get the total error of the samples taken, and finally divide it by the number of samples to get the average value of the sample error.

[0046] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above and that the present invention can be implemented in other specific forms without departing from the spirit or essential features of the present invention. Therefore, the embodiments should be considered exemplary and non-restrictive in all respects, and the scope of the present invention is defined by the appended claims rather than the above description, and it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims be included in the present invention.

[0047] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. A sample standard deviation measurement method based on random sampling, characterized in that: The following steps are involved: S1: Using ArcMap, the orthophoto image and the point cloud image generated by the IGV mapping algorithm are scaled in the elevation direction to form a bird’s-eye view projection image as a sample image; S2: performing irregular linear segmentation on the sample image formed in step S1 to obtain a number of sample blocks; S3: Calculate the weight parameter of each sample block; S4: According to the weight parameter, randomly select an appropriate amount of point clouds in the sample block as sample points; S5: According to the difference between corner points and boundary points, take the distance from point to line or from point to center and calculate its standard deviation.

2. The method for measuring sample standard deviation based on random sampling according to claim 1, characterized in that: In step S3, the weight parameter of each sample block is calculated using the following formula: Where D is the weight of the sample block, N is the number of samples, I is the reflection intensity of the lidar, and F is the impact factor.

3. The method for measuring sample standard deviation based on random sampling according to claim 2, characterized in that: In step S5, there are two criteria for calculating the final error: Method 1: Select a point x i , the selection range is the entire point cloud, and the average value of the entire point cloud is calculated. The calculation result is recorded as Calculate the difference between them. After doing the same operation for all points in the point cloud, add the difference and take the square root after dividing by n-1. Method 2: Select a point x i Directly subtract the true value from the true value, calculate the error between the two, add the errors between multiple points, get the total error of the samples taken, and finally divide it by the number of samples to get the average of the sample errors.

Images