Lidar-based ground topography mapping method and system

Abstract

The application relates to the technical field of topographic mapping, in particular to a ground terrain mapping method and system based on a laser radar, which comprises the following steps: acquiring point cloud data of a region to be mapped by using a laser radar, wherein the point cloud data comprises an elevation, a reflection intensity value and a number of echoes; acquiring an elevation noise characteristic value of each point cloud data, and correcting the elevation noise characteristic value through a discrete degree of the number of echoes of neighborhood point cloud data of each point cloud data, so as to screen possible ground points from all the point cloud data; classifying all the point cloud data; acquiring a ground feature degree of each class, and correcting the ground feature degree through a discrete degree of the reflection intensity value of all the point cloud data in each class, so as to extract candidate ground points from all the point cloud data; extracting real ground points from all the candidate ground points, and then performing ground terrain mapping on the region to be mapped. The application aims to improve the precision of ground terrain mapping by accurately extracting real ground points.

Description

Technical Field

[0001] This application relates to the field of topographic mapping technology, specifically to a ground topographic mapping method and system based on lidar. Background Technology

[0002] With the rapid development of modern engineering construction, urban planning, and Geographic Information Systems (GIS), the demand for high-precision ground topographic mapping is increasing. LiDAR technology, with its advantage of rapidly acquiring large amounts of three-dimensional spatial data, has been widely used in ground topographic mapping.

[0003] However, in actual surveying environments, when lidar scans the ground terrain, various objects in the surrounding environment, such as vegetation, buildings, and utility poles, reflect the laser light, introducing a large number of reflection points unrelated to the ground terrain into the acquired point cloud data. These points appear as noise in the point cloud data, severely interfering with the identification of the true ground terrain. Most existing noise filtering methods are based on a single feature for the identification and removal of noise points. For example, they judge whether a point is a noise point based solely on the elevation of the point cloud data. However, in complex environments, the elevation of noise points may be similar to that of real ground points, making it difficult to accurately distinguish between noise points and real ground points based solely on elevation features. Different materials reflect laser light in different intensities, but in reality, the reflection intensity values ​​of some noise points may overlap with those of ground points, making it impossible to accurately identify noise points and thus affecting the accuracy of ground terrain surveying. Summary of the Invention

[0004] In light of the above, it is necessary to provide a ground topographic mapping method and system based on lidar. Compared with traditional lidar-based ground topographic mapping methods, this method improves the accuracy of ground topographic mapping by accurately extracting real ground points. In a first aspect, embodiments of this application provide a ground topographic mapping method based on lidar, the method comprising the following steps: LiDAR is used to acquire point cloud data of the area to be mapped, which includes elevation, reflection intensity value and echo count; By measuring the elevation differences between each point cloud data and its neighboring point cloud data, and the degree to which the elevations of each point cloud data and its neighboring point cloud data deviate from the historical elevation range, the elevation noise characteristic values ​​of each point cloud data are obtained. The noise is then corrected by measuring the dispersion of the echo counts of the neighboring point cloud data of each point cloud data, so as to screen possible ground points from all point cloud data. All point cloud data are classified according to reflection intensity value and spatial location; the ground feature degree of each category is obtained by combining the proportion of possible ground points in each category, spatial distance, and distribution of correction values ​​of elevation noise feature values, and is corrected by the dispersion of reflection intensity values ​​of all point cloud data in each category, so as to extract candidate ground points from all point cloud data. By taking into account the number of candidate ground points within the nearest neighbor range of each candidate ground point, as well as the correction values ​​of the elevation noise characteristic value and the ground characteristic degree of the class of each candidate ground point, real ground points are extracted from all candidate ground points, and then ground topography is carried out in the area to be surveyed.

[0005] In one embodiment, the process of obtaining the elevation noise feature value is as follows: In the horizontal direction, taking any point cloud data as the origin, the nearest neighbor region of the any point cloud data is evenly divided into multiple sector regions with equal angles. By comparing the elevation of any point cloud data with the point cloud data in each sector region, and combining the average level of the extent to which the point cloud data in each sector region exceeds the historical elevation range, the elevation noise level of each sector region is obtained. Calculate the sum of the normalized value of the extent to which the elevation of any point cloud data exceeds the historical elevation range and 1; The average elevation noise level of all sector regions in the nearest region of any point cloud data is denoted as the noise mean. The elevation noise characteristic value of any point cloud data is the weighted sum of the sum value and the noise mean value.

[0006] In one embodiment, the process of obtaining the elevation noise level is as follows: Calculate the arithmetic mean of the elevation differences between any given point cloud data and all other point cloud data in each sector region; Calculate the percentage of point cloud data in each sector whose elevation exceeds the historical elevation range in all point cloud data. For point cloud data in each sector where the elevation exceeds the historical elevation range, calculate the average value of the extent to which the elevation of all point cloud data in each sector exceeds the historical elevation range. The elevation noise level is positively correlated with the arithmetic mean, the quantity ratio, and the average value.

[0007] In one embodiment, the step of correcting the dispersion of echo counts from neighboring point cloud data to filter possible ground points from all point cloud data includes: When the number of echoes of each point cloud data is greater than a preset value, the dispersion of the number of echoes of the neighboring point cloud data of each point cloud data is recorded as the echo number dispersion; otherwise, the echo number dispersion is assigned a value of 0. The correction value of the elevation noise feature value is positively correlated with the echo number dispersion and the elevation noise feature value, respectively. By comparing the corrected value of the elevation noise feature with the preset elevation noise threshold, possible ground points are selected from all point cloud data.

[0008] In one embodiment, the process of obtaining the ground feature degree is as follows: The mean distance between all possible ground points in each category and their nearest possible ground point is denoted as the mean distance. The average of the corrected elevation noise characteristic values ​​of all possible ground points in each category is denoted as the comprehensive noise mean. The ground characteristic is negatively correlated with the mean distance and the mean comprehensive noise, and positively correlated with the quantity ratio.

[0009] In one embodiment, the ground characteristic is the normalized value of the product of the reciprocal of the distance mean, the reciprocal of the comprehensive noise mean, and the quantity ratio.

[0010] In one embodiment, the step of correcting for the dispersion of reflection intensity values ​​across all point cloud data to extract candidate ground points from all point cloud data includes: The dispersion of the reflection intensity values ​​of all point cloud data in each category is denoted as the reflection intensity dispersion. The correction value of the ground characteristic is positively correlated with the ground characteristic and negatively correlated with the reflection intensity dispersion; Point cloud data in each class whose ground feature value correction value is greater than or equal to a preset first ground threshold are used as candidate ground points.

[0011] In one embodiment, the extraction process of the real ground points is as follows: Calculate the ratio of the total number of candidate ground points within the nearest neighbor range of each candidate ground point to the total number of point cloud data points; The ground confidence value of each candidate ground point is obtained by using the ratio of each candidate ground point, the correction value of the ground characteristic, and the correction value of the elevation noise characteristic. Candidate ground points whose ground confidence values ​​are greater than or equal to a preset second ground threshold are taken as real ground points.

[0012] In one embodiment, the process of obtaining the ground confidence value is as follows: The corrected elevation noise characteristic values ​​of each candidate ground point are mapped to positive numbers; the ground confidence value is the normalized value of the product of the reciprocal of the positive number, the ratio, and the corrected ground characteristic value.

[0013] Secondly, embodiments of this application also provide a ground topographic mapping system based on lidar, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any of the lidar-based ground topographic mapping methods described above.

[0014] This application has at least the following beneficial effects: This application calculates elevation noise characteristic values ​​by comprehensively considering elevation differences and the degree of deviation of elevation from the historical elevation range. This can effectively identify point cloud data with elevation anomalies. Furthermore, by combining the characteristics that the echo count of normal ground points is usually relatively stable, while the echo count of noise points may be more discrete, the elevation noise characteristic values ​​are corrected by using the discreteness of the echo count, which helps to more accurately screen out possible ground points. Furthermore, classifying reflection intensity values ​​in conjunction with spatial location allows for better utilization of multi-dimensional information from point cloud data. By integrating various features such as the proportion of possible ground points and spatial distances, more reliable candidate ground points can be extracted, providing a more accurate basis for subsequent extraction of real ground points. Considering that real ground points are usually continuously distributed in space, this feature can be used to further improve the accuracy of real ground point extraction, effectively removing the interference of noise points. This ensures that the extracted real ground points accurately reflect the real ground topography, thereby improving the accuracy of ground topography mapping. Attached Figure Description

[0015] To more clearly illustrate the technical solutions and advantages in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 A flowchart illustrating the steps of a ground topographic mapping method based on lidar provided in one embodiment of this application; Figure 2 A schematic diagram of the process for obtaining ground confidence values; Figure 3 This is a schematic diagram of the process for extracting real ground points. Detailed Implementation

[0017] In the description of the embodiments in this application, the words "exemplary," "or," and "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design scheme described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design schemes. Specifically, the use of the words "exemplary," "or," and "for example" is intended to present the relevant concepts in a specific manner.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. It should be understood that, unless otherwise stated, " / " in this application means "or".

[0019] It should also be noted that the terms "first" and "second" in this application are used to distinguish similar objects, rather than to describe a specific order or sequence.

[0020] The following description, in conjunction with the accompanying drawings, details the specific scheme of the lidar-based ground topographic mapping method and system provided in this application.

[0021] Please see Figure 1 The diagram illustrates a flowchart of a ground topography mapping method based on lidar according to an embodiment of this application. The method includes the following steps: Step 1: Use lidar to acquire point cloud data of the area to be mapped, which includes elevation, reflection intensity value and echo count.

[0022] To perform ground topographic mapping based on lidar, point cloud data of the area to be mapped needs to be collected. The specific data collection steps are as follows: Based on the requirements of the surveying task, a lidar with appropriate accuracy, scanning range, and frequency is selected. Before scanning, the lidar is calibrated and its parameters are set to ensure accurate acquisition of point cloud data of the area to be surveyed. During the scanning process, the area to be surveyed is fully covered according to a predetermined scanning path to obtain complete point cloud data of the area, including: planar coordinates, elevation, reflection intensity value, and echo count. The reflection intensity value reflects the reflectivity of an object's surface to laser light; different materials have different reflection intensities. The echo count helps distinguish different types of objects; for example, vegetation typically produces multiple echoes, while the ground generally produces fewer echoes.

[0023] Step 2: Comprehensively analyze the elevation, reflection intensity value and echo count of the point cloud data of the area to be surveyed, and extract the real ground points from the point cloud data of the area to be surveyed.

[0024] In point cloud data acquired by lidar, the elevation of normal ground points typically matches the terrain elevation of the area. The elevation range of normal ground points follows the overall altitude and undulation of the terrain. For example, in plains areas, the terrain is relatively flat with minimal undulation. Under normal circumstances, the elevation is highly concentrated with minimal fluctuations, and the reflection intensity values ​​are also relatively concentrated, with an echo count of mostly one to two. In contrast, noise points exhibit different characteristics. In plains areas, due to the absence of tall mountains or dense buildings, noise points mainly originate from low-altitude obstacles, non-terrain objects on the ground, or environmental interference. These noise points usually appear as isolated points or small clusters, with elevations deviating from the normal ground elevation range. The reflection intensity values ​​are more dispersed and fluctuate significantly, with a higher and more irregular echo count.

[0025] Step 2.1: Obtain the elevation noise characteristic value of each point cloud data by measuring the elevation difference between each point cloud data and its neighboring point cloud data, as well as the degree to which the elevation of each point cloud data and its neighboring point cloud data deviates from the historical elevation range.

[0026] Based on historical surveying data of the area to be surveyed, the undulation range of the area is determined and used as the historical elevation range of the area. It should be noted that: for newly surveyed areas where historical surveying data is unavailable, the method for obtaining the historical elevation range of the area to be surveyed is as follows: calculate the mean and standard deviation of the elevations of all point cloud data for the area to be surveyed; the historical elevation range of the area is then determined. ,in, This represents the average elevation of all point cloud data in the area to be surveyed; This represents the standard deviation of the elevation of all point cloud data in the area to be surveyed.

[0027] Assess whether the elevation of each point cloud data point in the area to be mapped is within the historical elevation range. If it is, it indicates that the location of each point cloud data point matches the terrain features of the area to be mapped and is likely a normal ground point. If it is not, it indicates that the location of each point cloud data point does not match the terrain features of the area to be mapped and is likely a noise point. At the same time, calculate the extent to which the elevation of each point cloud data point exceeds the historical elevation range. The greater the extent of the exceedance, the higher the degree to which the elevation of each point cloud data point deviates from the normal ground elevation.

[0028] Based on the above analysis, the elevation noise characteristic values ​​of each point cloud data are obtained by measuring the elevation differences between each point cloud data and its nearest neighbor point cloud data, as well as the degree to which the elevations of each point cloud data and its nearest neighbor point cloud data deviate from the historical elevation range. The specific process is as follows: Taking the i-th point cloud data of the area to be mapped as an example, the neighboring point cloud data of the i-th point cloud data are used to form the neighboring region of the i-th point cloud data. In the horizontal direction, with the i-th point cloud data as the origin, the neighboring region of the i-th point cloud data is evenly divided into multiple fan-shaped regions with equal angles. By comparing the elevation of the i-th point cloud data with that of the point cloud data in each sector region, and combining the average level of the extent to which the point cloud data in each sector region exceeds the historical elevation range, the elevation noise level of each sector region is obtained. Calculate the sum of the normalized value of the magnitude by which the elevation of the i-th point cloud data exceeds the historical elevation range and 1; denot the mean of the elevation noise level of all fan-shaped regions in the nearest region of the i-th point cloud data as the noise mean; the elevation noise characteristic value of the i-th point cloud data is the weighted sum of the sum and the noise mean.

[0029] The process of obtaining the elevation noise level of each sector is as follows: calculate the arithmetic mean of the elevation difference between the i-th point cloud data and all other point cloud data in each sector. The larger the arithmetic mean is, the more different the elevation changes between the i-th point cloud data and the point cloud data in each sector are, and the more likely the i-th point cloud data is to be a noise point. Calculate the percentage of point cloud data in each sector whose elevation exceeds the historical elevation range in all point cloud data; for point cloud data in each sector whose elevation exceeds the historical elevation range, calculate the average value of the magnitude of elevation exceeding the historical elevation range for all point cloud data in each sector; the elevation noise level of each sector is positively correlated with the arithmetic mean, the percentage of points, and the average value, respectively.

[0030] It should be noted that: difference refers to the degree of distinction between data, which can be measured by calculating the absolute value of the difference, the square of the difference, the ratio, etc. This application does not impose any special restrictions on this.

[0031] It should be noted that positive correlation means that the variables have the same direction of change; when one variable increases, the other variable also increases, and when one variable decreases, the other variable also decreases.

[0032] In this embodiment, when calculating the elevation noise feature value of the i-th point cloud data, the weights of the sum and the mean noise value are 0.55 and 0.45, respectively, and the weights are derived from experimental data.

[0033] In this embodiment, the number of neighboring point cloud data of the i-th point cloud data is 100. The method for obtaining the neighboring point cloud data of the i-th point cloud data is as follows: all other point cloud data are arranged in ascending order according to their spatial distance to the i-th point cloud data, and the first 100 point cloud data are taken as the neighboring point cloud data of the i-th point cloud data. Here, 100 is only one embodiment of this application. The implementer can set it according to the actual situation. This application does not impose any special restrictions.

[0034] In this embodiment, the expression for the elevation noise level of each sector region is: In the formula, This represents the elevation noise level of the m-th sector. The arithmetic mean of the elevation differences between the i-th point cloud data and all other point cloud data in the m-th sector region; This represents the percentage of point cloud data in the m-th sector whose elevation exceeds the historical elevation range among all point cloud data. For the point cloud data in the m-th sector whose elevation exceeds the historical elevation range, calculate the average magnitude of the elevation deviation from the historical elevation range for all point cloud data in the m-th sector. This represents the average elevation range of all point cloud data in the m-th sector, where the elevation exceeds the historical elevation range; norm() represents the normalization operation. The difference between elevations is the absolute value of the difference.

[0035] In this application, unless otherwise specified, the Min-Max normalization method is used for normalization. The Min-Max normalization method is a well-known technique and will not be described in detail here.

[0036] It should be noted that by dividing the data into fan-shaped regions, directional analysis of the local terrain surrounding the i-th point cloud data can be performed. The elevation changes within each fan-shaped region can reflect the undulation characteristics of the local terrain. The greater the elevation difference between the i-th point cloud data and the m-th fan-shaped region, and the more point cloud data in the m-th fan-shaped region whose elevation exceeds the historical elevation range, and the greater the extent of exceeding the historical elevation range, the greater the calculated elevation noise level, indicating that the noise characteristics in the m-th fan-shaped region are more significant. Conversely, the greater the extent to which the elevation of the i-th point cloud data exceeds the historical range, and the more significant the noise characteristics of all fan-shaped regions in the i-th point cloud data's neighboring regions, the greater the calculated elevation noise characteristic value, indicating that the i-th point cloud data is more likely to be a noise point.

[0037] Step 2.2: By analyzing the dispersion of echo counts in the neighboring point cloud data of each point cloud data, the elevation noise characteristic value of each point cloud data is corrected in order to screen possible ground points from all point cloud data.

[0038] In normal ground areas, the number of echoes from a lidar sensor is typically 1 to 2, while noise points often produce more echoes. However, some objects may exhibit similar reflection patterns to normal ground. For example, small objects with smooth surfaces and strong laser absorption may only produce 1 to 2 echoes after laser irradiation, overlapping with the echo count range of normal ground points. This can easily lead to misjudgments when distinguishing between ground points and noise points based on the number of echoes.

[0039] Based on the above analysis, the elevation noise characteristic value of each point cloud data is corrected by the dispersion of the echo count of the neighboring point cloud data. The corrected elevation noise characteristic value of each point cloud data is then used as the elevation noise probability of each point cloud data. The specific process is as follows: When the number of echoes of each point cloud data exceeds a preset value, the dispersion of the number of echoes of the neighboring point cloud data of each point cloud data is recorded as the echo number dispersion; otherwise, the echo number dispersion is assigned a value of 0. The elevation noise probability of each point cloud data is positively correlated with the echo number dispersion and the elevation noise characteristic value of each point cloud data.

[0040] It should be noted that dispersion refers to the degree of unevenness in the distribution of data, which can be achieved by calculating variance, standard deviation, coefficient of variation, etc. This application does not impose any special restrictions on it.

[0041] In this embodiment, the preset value is 2. The preset value is preset by a person and the implementer can set it according to the actual situation. This application does not impose any special restrictions.

[0042] In this embodiment, the number of neighboring point cloud data of the i-th point cloud data is 10. The method for obtaining the neighboring point cloud data of the i-th point cloud data is as follows: all other point cloud data are arranged in ascending order according to their spatial distance to the i-th point cloud data, and the first 10 point cloud data are taken as the neighboring point cloud data of the i-th point cloud data. Here, 10 is only one embodiment of this application. The implementer can set it according to the actual situation. This application does not impose any special restrictions.

[0043] In this embodiment, the dispersion of the number of echoes is the variance.

[0044] In this embodiment, the expression for the elevation noise probability of each point cloud data is: In the formula, Indicates the elevation noise probability of the i-th point cloud data; This represents the elevation noise feature value of the i-th point cloud data; The echo frequency dispersion of the neighborhood point cloud data of the i-th point cloud data is represented by norm(), which represents the normalization operation.

[0045] It should be noted that: the greater the change in the number of echoes of the neighboring point cloud data of the i-th point cloud data, the greater the probability that the i-th point cloud data belongs to a noise point; the greater the calculated elevation noise probability, the more likely the i-th point cloud data is to be a noise point.

[0046] Furthermore, by comparing the elevation noise probability of each point cloud data with a preset elevation noise threshold, possible ground points are selected from all point cloud data, specifically as follows: Point cloud data with an elevation noise probability greater than or equal to the preset elevation noise threshold are taken as possible noise points, and the remaining point cloud data are taken as possible ground points.

[0047] In this embodiment, the preset elevation noise threshold is set to 0.7, and the preset elevation noise threshold is calculated from experimental data.

[0048] Step 2.3: Classify all point cloud data according to reflection intensity value and spatial location; combine the proportion of possible ground points in each category, spatial distance, and distribution of correction values ​​for elevation noise feature values ​​to obtain the ground feature degree of each category, and correct it by the dispersion of reflection intensity values ​​of all point cloud data in each category, so as to extract candidate ground points from all point cloud data.

[0049] In practical applications, due to the diversity and reflectivity of ground features, the reflected signals of non-ground objects or the elevation of transition areas between ground features and the ground may exhibit anomalies. Furthermore, measurement accuracy issues can also lead to deviations in the elevation measurements of normal ground points. These factors can all cause noise points to be misidentified as target ground points. Therefore, to more accurately distinguish between real ground points and noise points, it is necessary to further consider characteristics such as the reflection intensity, location, and distances between all point cloud data.

[0050] The ground is typically composed of a relatively homogeneous material, and these identical materials have similar laser reflection characteristics, resulting in a high degree of consistency in reflection intensity among ground points. Furthermore, the spatial distribution of ground points is relatively concentrated, with very few discrete, isolated points. In contrast, noise points are mostly isolated points, lacking significant clustering.

[0051] Based on the above analysis, all point cloud data are classified according to reflection intensity value and spatial location. Specifically, to avoid the influence of different units on subsequent processing, different features are normalized, including x-coordinate, y-coordinate, z-coordinate, and reflection intensity value, where z-coordinate is the elevation. Then, the x-coordinate, y-coordinate, z-coordinate, and reflection intensity value of each point cloud data are used to form the feature vector of each point cloud data. The point cloud data are classified according to the similarity between the feature vectors of the point cloud data.

[0052] In this embodiment, K-Means is used to classify point cloud data. The distance metric is the Euclidean distance between feature vectors, and the number of classifications is determined by the elbow rule. Both K-Means and the elbow rule are well-known technologies, and will not be described in detail here. As other implementation methods, implementers may use other existing feasible technologies to classify point cloud data, and this application does not impose any special restrictions.

[0053] Furthermore, by combining the proportion of possible ground points, spatial distances, and the distribution of elevation noise probability among various categories, the ground characteristic degree of each category is obtained. The specific process is as follows: The average distance between all possible ground points in each category and their nearest possible ground point is denoted as the distance mean; the average elevation noise probability of all possible ground points in each category is denoted as the comprehensive noise mean. The ground feature degree of each type is negatively correlated with the mean distance and the mean comprehensive noise, and positively correlated with the proportion of possible ground points in each type.

[0054] It should be noted that negative correlation means that the variables change in opposite directions; when one variable increases, the other decreases, and vice versa.

[0055] In this embodiment, the expressions for the ground feature degree of each type are as follows: In the formula, Represents the ground feature degree of the k-th class; This represents the mean distance in the k-th class; This represents the proportion of possible ground points in the k-th class; This represents the mean of the combined noise in the k-th class; norm() represents the normalization operation. It should be added that: if If the value is 0, calculate first. The sum of the sum with a preset constant greater than 0 is then used for subsequent calculations. The value of the preset constant greater than 0 is preset by the user and can be set by the implementer according to the actual situation. In this embodiment, the value of the preset constant greater than 0 is 0.01.

[0056] In another embodiment, the expression for the ground feature degree of each type is: In the formula, Represents the ground feature degree of the k-th class; This represents the mean distance in the k-th class; This represents the proportion of possible ground points in the k-th class; represents the mean of the combined noise in the k-th class; norm() represents the normalization operation; e represents the natural constant.

[0057] It should be noted that: the smaller the mean distance value in each category, the more clustered the distribution of all possible ground points in each category, and the greater the probability that the point cloud data in each category belongs to ground points; the greater the proportion of possible ground points in each category, the greater the probability that the point cloud data in each category belongs to ground points; and the greater the calculated ground feature degree, the more likely the point cloud data in each category belongs to ground points.

[0058] Furthermore, by analyzing the dispersion of the reflection intensity values ​​of all point cloud data in each category, the ground feature degree of each category is corrected to obtain the corrected ground feature degree for each category, which is then used as the ground feature fit degree for each category. The specific process is as follows: The dispersion of the reflection intensity values ​​of all point cloud data in each category is denoted as the reflection intensity dispersion. The ground feature fit of each category is positively correlated with the ground feature degree of each category and negatively correlated with the reflection intensity dispersion.

[0059] In this embodiment, the dispersion of the reflection intensity value is the interquartile range.

[0060] In this embodiment, the expression for the fit of various ground features is: In the formula, This represents the degree of fit of the ground features in the k-th class; This represents the dispersion of the reflection intensity in the k-th class; Represents the ground feature degree of the k-th class; norm() represents the normalization operation; This indicates a preset value greater than 0, used to avoid a denominator of 0. The value is preset by a person, and the implementer can set it according to the actual situation. In this embodiment, The value is 0.01.

[0061] In another embodiment, the expression for the ground feature fit of each type is: In the formula, This represents the degree of fit of the ground features in the k-th class; This represents the dispersion of the reflection intensity in the k-th class; represents the ground characteristic of the k-th class; norm() represents the normalization operation; e represents the natural constant.

[0062] It should be noted that: the reflection intensity dispersion is used to reflect the consistency of the reflection intensity values ​​of point cloud data in various categories. The smaller the calculated reflection intensity dispersion, the more consistent the reflection intensity values ​​of point cloud data in various categories, and the more likely the point cloud data in each category is to be ground points; the larger the calculated ground feature fit, the more likely the point cloud data in each category is to be ground points.

[0063] Furthermore, point cloud data in each class with a ground feature fit greater than or equal to a preset first ground threshold are used as candidate ground points.

[0064] In this embodiment, the preset first ground threshold is set to 0.6, and the preset first ground threshold is calculated from experimental data.

[0065] Step 2.4: Extract real ground points from all candidate ground points by using the number of candidate ground points in the nearest neighbor range of each candidate ground point, as well as the correction values ​​of the elevation noise characteristic value and the ground characteristic degree of the class to which each candidate ground point belongs.

[0066] Since normal ground points are spatially continuous, for each candidate ground point, the ratio of the total number of candidate ground points in the nearest neighbor range to the total number of point cloud data is calculated and recorded as the nearest neighbor ratio of each candidate ground point. The larger the calculated nearest neighbor ratio, the more candidate ground points are distributed around each candidate ground point, the more continuous the distribution of candidate ground points in the nearest neighbor range is, and the greater the probability that each candidate ground point belongs to a ground point.

[0067] In this embodiment, the nearest neighbor range of each point cloud data consists of all the neighboring point cloud data of each point cloud data.

[0068] Furthermore, by using the ratio of the number of nearest neighbors of each candidate ground point, and the degree of fit between the elevation noise probability of each candidate ground point and the ground features of its class, the ground confidence value of each candidate ground point is obtained. The specific process is as follows: The elevation noise probability of each candidate ground point is mapped to a positive number; the ground confidence value of each candidate ground point is the normalized value of the product of the reciprocal of the positive number, the ratio of the number of nearest neighbors of each candidate ground point, and the ground feature fit of the class to which each candidate ground point belongs. The purpose of mapping the elevation noise probability to a positive number is to avoid a denominator of 0. There are many methods for mapping data to positive numbers, and implementers can choose feasible methods at their own discretion; this application does not impose any special restrictions. A schematic diagram of the process for obtaining the ground confidence value is shown below. Figure 2 As shown.

[0069] In this embodiment, the elevation noise probability is mapped to a positive number by calculating the sum of the elevation noise probability and a constant greater than 0. The constant greater than 0 is 0.01. The value of the constant greater than 0 is set by the implementer according to the actual situation, and this application does not impose any special restrictions.

[0070] It should be noted that the higher the calculated ground confidence value, the greater the probability that each candidate ground point is a ground point.

[0071] Furthermore, candidate ground points with ground confidence values ​​greater than or equal to a preset second ground threshold are designated as real ground points, while the remaining candidate ground points are designated as noise points. A schematic diagram of the real ground point extraction process is shown below. Figure 3 As shown.

[0072] In this embodiment, the preset second ground threshold is set to 0.5, and the preset second ground threshold is calculated from experimental data.

[0073] Step 3: Using the extracted real ground points, perform ground topographic mapping of the area to be surveyed.

[0074] Furthermore, using all real ground points in the area to be surveyed as the core data for ground topographic mapping can remove the interference of noise points, such as noise points reflected by non-ground objects like buildings and utility poles, ensuring that subsequent ground topographic mapping is based on data that accurately reflects the actual ground conditions.

[0075] The three-dimensional coordinate information of all real ground points in the area to be surveyed is processed and converted into a regular grid. The value of each grid cell represents the ground elevation at its location, thereby constructing a digital elevation model. Based on this, a contour map is drawn, and corresponding terrain analysis is performed, including calculating the slope, determining the aspect, and analyzing the curvature of the terrain, thus realizing the ground topographic mapping of the area to be surveyed.

[0076] Based on the same inventive concept as the above methods, this application also provides a ground topographic mapping system based on lidar, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any of the above-described ground topographic mapping methods based on lidar.

[0077] In summary, this application calculates elevation noise characteristic values ​​by comprehensively considering elevation differences and the degree of deviation of elevation from historical elevation ranges. This can effectively identify point cloud data with elevation anomalies. Furthermore, by combining the characteristic that the echo count of normal ground points is usually relatively stable, while the echo count of noise points may be more discrete, the elevation noise characteristic values ​​are corrected using the discreteness of the echo count, which helps to more accurately screen out possible ground points. Furthermore, classifying reflection intensity values ​​in conjunction with spatial location allows for better utilization of multi-dimensional information from point cloud data. By integrating various features such as the proportion of possible ground points and spatial distances, more reliable candidate ground points can be extracted, providing a more accurate basis for subsequent extraction of real ground points. Considering that real ground points are usually continuously distributed in space, this feature can be used to further improve the accuracy of real ground point extraction, effectively removing the interference of noise points. This ensures that the extracted real ground points accurately reflect the real ground topography, thereby improving the accuracy of ground topography mapping.

[0078] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than that shown in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. In the descriptions corresponding to the flowcharts and block diagrams in the accompanying drawings, the operations or steps corresponding to different blocks may also occur in a different order than disclosed in the description, and sometimes there is no specific order between different operations or steps. For example, two consecutive operations or steps may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. Each block in a block diagram and / or flowchart, and combinations of blocks in a block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0079] It will be apparent to those skilled in the art that this application is not limited to the details of the exemplary embodiments described above, and that this application can be implemented in other specific forms without departing from its essential characteristics. Therefore, the embodiments described above should be considered exemplary and non-limiting in all respects.

Claims

1. A ground topographic mapping method based on lidar, characterized in that, The method includes the following steps: LiDAR is used to acquire point cloud data of the area to be mapped, which includes elevation, reflection intensity value and echo count; By measuring the elevation differences between each point cloud data and its neighboring point cloud data, and the degree to which the elevations of each point cloud data and its neighboring point cloud data deviate from the historical elevation range, the elevation noise characteristic values ​​of each point cloud data are obtained. The noise is then corrected by measuring the dispersion of the echo counts of the neighboring point cloud data of each point cloud data, so as to screen possible ground points from all point cloud data. All point cloud data are classified according to reflection intensity value and spatial location; the ground feature degree of each category is obtained by combining the proportion of possible ground points in each category, spatial distance, and distribution of correction values ​​of elevation noise feature values, and is corrected by the dispersion of reflection intensity values ​​of all point cloud data in each category, so as to extract candidate ground points from all point cloud data. By taking the number of candidate ground points in the neighborhood of each candidate ground point, as well as the correction values ​​of the elevation noise characteristic value and the ground characteristic value of the class of each candidate ground point, the real ground points are extracted from all candidate ground points, and then the ground topography of the area to be surveyed is carried out. The process for obtaining the elevation noise feature values ​​is as follows: In the horizontal direction, taking any point cloud data as the origin, the nearest neighbor region of the any point cloud data is evenly divided into multiple sector regions with equal angles. By comparing the elevation of any point cloud data with the point cloud data in each sector region, and combining the average level of the extent to which the point cloud data in each sector region exceeds the historical elevation range, the elevation noise level of each sector region is obtained. Calculate the sum of the normalized value of the extent to which the elevation of any point cloud data exceeds the historical elevation range and 1; The average elevation noise level of all sector regions in the nearest region of any point cloud data is denoted as the noise mean. The elevation noise characteristic value of any point cloud data is the weighted sum of the sum value and the noise mean value.

2. The ground topographic mapping method based on lidar as described in claim 1, characterized in that, The process of obtaining the elevation noise level is as follows: Calculate the arithmetic mean of the elevation differences between any given point cloud data and all other point cloud data in each sector region; Calculate the percentage of point cloud data in each sector whose elevation exceeds the historical elevation range in all point cloud data. For point cloud data in each sector where the elevation exceeds the historical elevation range, calculate the average value of the extent to which the elevation of all point cloud data in each sector exceeds the historical elevation range. The elevation noise level is positively correlated with the arithmetic mean, the quantity ratio, and the average value.

3. The ground topographic mapping method based on lidar as described in claim 1, characterized in that, The step of correcting for the dispersion of echo counts from neighboring point cloud data to filter possible ground points from all point cloud data includes: When the number of echoes of each point cloud data is greater than a preset value, the dispersion of the number of echoes of the neighboring point cloud data of each point cloud data is recorded as the echo number dispersion; otherwise, the echo number dispersion is assigned a value of 0. The correction value of the elevation noise feature value is positively correlated with the echo number dispersion and the elevation noise feature value, respectively. By comparing the corrected value of the elevation noise feature with the preset elevation noise threshold, possible ground points are selected from all point cloud data.

4. The ground topographic mapping method based on lidar as described in claim 1, characterized in that, The process of obtaining the ground feature degree is as follows: The mean distance between all possible ground points in each category and their nearest possible ground point is denoted as the mean distance. The average of the corrected elevation noise characteristic values ​​of all possible ground points in each category is denoted as the comprehensive noise mean. The ground characteristic is negatively correlated with the mean distance and the mean comprehensive noise, and positively correlated with the quantity ratio.

5. The ground topography mapping method based on lidar as described in claim 4, characterized in that, The ground characteristic is the normalized value of the product of the reciprocal of the distance mean, the reciprocal of the comprehensive noise mean, and the quantity ratio.

6. The ground topographic mapping method based on lidar as described in claim 1, characterized in that, The step of correcting for the dispersion of reflection intensity values ​​across all point cloud data types to extract candidate ground points from all point cloud data includes: The dispersion of the reflection intensity values ​​of all point cloud data in each category is denoted as the reflection intensity dispersion. The correction value of the ground characteristic is positively correlated with the ground characteristic and negatively correlated with the reflection intensity dispersion; Point cloud data in each class whose ground feature value correction value is greater than or equal to a preset first ground threshold are used as candidate ground points.

7. The ground topographic mapping method based on lidar as described in claim 1, characterized in that, The process for extracting the real ground points is as follows: Calculate the ratio of the total number of candidate ground points within the nearest neighbor range of each candidate ground point to the total number of point cloud data points; The ground confidence value of each candidate ground point is obtained by using the ratio of each candidate ground point, the correction value of the ground characteristic, and the correction value of the elevation noise characteristic. Candidate ground points whose ground confidence values ​​are greater than or equal to a preset second ground threshold are taken as real ground points.

8. The ground topography mapping method based on lidar as described in claim 7, characterized in that, The process for obtaining the ground confidence value is as follows: The corrected elevation noise characteristic values ​​of each candidate ground point are mapped to positive numbers; the ground confidence value is the normalized value of the product of the reciprocal of the positive number, the ratio, and the corrected ground characteristic value.

9. A ground topographic mapping system based on lidar, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the ground terrain mapping method based on lidar as described in any one of claims 1-8.

Images