CN-121482611-B - Method for identifying mao-bamboo forest snow disaster damage of airborne laser radar based on differential point cloud encryption strategy

CN121482611BCN 121482611 BCN121482611 BCN 121482611BCN-121482611-B

Abstract

The invention relates to the technical field of forestry disaster monitoring, in particular to an identification method of an airborne laser radar phyllostachys pubescens disaster damage based on a differential point cloud encryption strategy; according to the method, the three-dimensional structure of the bamboo forest is reconstructed through fitting the axis of the bamboo trunk and image post-processing, different types of disaster damage are identified, virtual point clouds are generated to fill the scanning missing area, the accuracy of the axis of the bamboo trunk is effectively improved, the image expression of the bamboo trunk and the tree crowns is enhanced through layered reconstruction and principal component analysis technology, a finer disaster damage characteristic diagram is provided, and accurate disaster assessment is facilitated. In addition, the accuracy and reliability of disaster damage analysis are further improved through targeted image processing based on disaster damage types. In the whole, the invention provides high-efficiency and accurate technical support for post-disaster evaluation and ecological restoration of the bamboo forests.

Inventors

GUAN FENGYING
FENG PENGFEI
WANG YUETING

Assignees

国际竹藤中心

Dates

Publication Date: 20260508
Application Date: 20251215

Claims (7)

1. The method for identifying the disaster damage of the phyllostachys pubescens forest of the airborne laser radar based on the differential point cloud encryption strategy is characterized by comprising the following steps of: step one, data acquisition and input; receiving airborne laser radar point cloud data and ground investigation data; Step two, data preprocessing; preprocessing the acquired point cloud data to obtain a normalization processing result of laser reflection intensity; Fitting the axis of the bamboo trunk; fitting the axis of the bamboo trunk by layered reconstruction and principal component analysis based on the normalization processing result of the laser reflection intensity, and filling the missing area; step four, segmentation feature fusion and image post-processing; post-processing is carried out based on the fitted bamboo trunk axis, and image expression containing disaster damage type characteristics is enhanced; the specific process for carrying out post-treatment based on fitting of the bamboo dry axis comprises the following steps: If the fact that the axis of the fitting bamboo trunk is not branched is identified, and the included angle between the highest height layer in the axis of the fitting bamboo trunk and the ground xoy surface is smaller than a first preset threshold rotation angle, judging that the bamboo is bent; if the axis of the fitted bamboo trunk is identified to be not branched, and the axis is absent, the bamboo is judged to be broken; If the axis branches exist on the axis of the fitting bamboo trunk and the angle between the axis branches is larger than a second preset threshold rotation angle, the bamboo split is judged to be broken; if the position of the root is identified, the included angle between the axis of the fitted bamboo trunk and the surface xoy of the ground is smaller than a third preset threshold rotation angle, judging that the stump turning bamboo is carried out; The specific process for enhancing the image expression containing the disaster damage type characteristics is as follows: the method comprises the steps of carrying out axis extension on bent bamboo, aiming at a fracture surface with the bending curvature larger than a preset threshold value of the broken bamboo, extending along the normal direction, adding random noise, improving the point cloud density of the fracture surface to strengthen a serrated edge, aiming at broken bamboo strips, generating bifurcation point cloud at the bifurcation position of the bamboo strips, reconstructing the broken form of the bamboo strips and fibers, and aiming at stump-turning bamboo strips, generating compensation point cloud with the inclination angle multiplied by a preset density coefficient at the root.
2. The method for identifying the damage of the phyllostachys pubescens and snowfall on the basis of the differential point cloud encryption strategy, which is characterized in that the point cloud data of the airborne laser radar are specifically as follows: The method comprises the steps of scanning point coordinates and laser reflection intensity of the surface of the moso bamboo, wherein each scanning point coordinate of the surface of the moso bamboo has the laser reflection intensity which is uniquely corresponding to the scanning point coordinate of the surface of the moso bamboo, and the ground investigation data specifically comprise average ground elevation.
3. The method for identifying the disaster damage of the phyllostachys pubescens by the airborne laser radar based on the differential point cloud encryption strategy as claimed in claim 1, wherein the specific process of preprocessing the point cloud data is as follows: carrying out Gaussian filtering on all acquired scanning point coordinates on the surface of the moso bamboo, and then carrying out threshold separation and normalization processing on the laser reflection intensity of each Mao Zhubiao face points; and aiming at the acquired phyllostachys pubescens surface scanning point coordinates, calculating Gaussian filtering output of each phyllostachys pubescens surface scanning point coordinate through Gaussian filtering.
4. The method for identifying the disaster damage of the phyllostachys pubescens by the airborne laser radar based on the differential point cloud encryption strategy as claimed in claim 1, wherein the specific process of preprocessing the collected point cloud data is as follows: Carrying out elevation threshold separation on Gaussian filter output of each moso bamboo surface scanning point coordinate to obtain an overground part and an underground part, carrying out normalization processing on laser reflection intensity of the overground part, marking the part with the z-axis component larger than the average ground height Cheng Wei as the overground part, and marking the part with the z-axis component smaller than or equal to the average ground height Cheng Wei as the underground part; Aiming at the overground part, obtaining the laser reflection intensity corresponding to the original coordinate point output by Gaussian filtering of each moso bamboo surface scanning point coordinate, and mapping the laser reflection intensity to a fixed range to obtain a laser reflection intensity normalization result.
5. The method for identifying the damage of the phyllostachys pubescens and snowfall on the basis of the differential point cloud encryption strategy, which is characterized by comprising the following specific process of fitting the axis of the dried bamboo: Layering at preset height intervals, generating point cloud data matrixes of each height layer based on laser reflection intensity normalization results, performing Principal Component Analysis (PCA) on the point cloud data matrix results of each height layer, calculating eigenvectors of a point cloud covariance matrix, generating discrete bamboo segments based on the eigenvectors of the covariance matrix, The method comprises the steps of obtaining the maximum value of the z-axis component of the overground part in Gaussian filtering output of all the moso bamboo surface scanning point coordinates, and calculating the height range of each height layer; And selecting the characteristic vector corresponding to the maximum characteristic value as the axis direction of the bamboo trunk to obtain bamboo trunk fragments corresponding to each height layer, and carrying out fitting connection on the bamboo trunk fragments.
6. The method for identifying the damage of the phyllostachys pubescens and snowfall on the basis of the differential point cloud encryption strategy, which is characterized by comprising the following specific processes of fitting and connecting the dried bamboo segments: The method comprises the steps of analyzing and detecting adjacent segments in space through connected components, obtaining the end point of a bamboo trunk segment corresponding to a previous high layer and the starting point of a bamboo trunk segment corresponding to a next high layer in the interface of each high layer, marking the upper and lower segments with x-axis components or y-axis components smaller than a preset threshold value of the start point and the end point as adjacent segments, locating the end point of the bamboo trunk segment with the smallest distance in the adjacent segments and the starting point of the bamboo trunk segment with the next high layer, taking the middle point as the fitting point of a bamboo trunk axis, and connecting the fitting points of all the bamboo trunk axes to obtain fitting bamboo trunk axes; and generating virtual cloud points with preset density based on the fitting bamboo trunk axis.
7. The method for identifying the disaster damage of the phyllostachys pubescens by the airborne laser radar based on the differential point cloud encryption strategy as claimed in claim 5, wherein the point cloud data matrix is specifically: The first three columns of the point cloud data matrix respectively represent x-axis components, y-axis components and z-axis components in coordinates corresponding to Gaussian filter output of all the moso bamboo surface scanning point coordinates contained in the height layer, and the fourth column of the point cloud data matrix represents laser reflection intensity normalization results corresponding to the coordinates.

Description

Method for identifying mao-bamboo forest snow disaster damage of airborne laser radar based on differential point cloud encryption strategy Technical Field The invention relates to the technical field of forestry disaster monitoring, in particular to an identification method of an airborne laser radar phyllostachys pubescens disaster damage based on a differential point cloud encryption strategy. Background With the increase of climate change and extreme weather events, damage of phyllostachys pubescens in natural disasters such as snow disaster, wind disaster and the like is gradually serious. The existing method based on point cloud processing has obvious limitation in processing the bamboo trunk missing area, so that the disaster damage assessment result is incomplete. An airborne laser radar (LiDAR) is used as an efficient remote sensing technology, can rapidly acquire three-dimensional point cloud data of large-area terrains and vegetation, has higher spatial resolution and accuracy, and is widely applied to the fields of forest resource monitoring, disaster assessment and the like. However, when the laser radar scans in the phyllostachys pubescens forest, point cloud data of the bamboo trunk part is often lost due to shielding of bamboo leaves and tree crowns. The existing point cloud processing-based bamboo forest disaster damage assessment method often omits filling and accurate reconstruction of a bamboo trunk missing area, and results in incomplete disaster damage assessment results or larger errors. Therefore, how to effectively utilize the point cloud data of the airborne laser radar, specifically identify different types of bamboo forest disaster damage, and solve the problem of point cloud data loss has become a key technical challenge in the current bamboo forest disaster assessment. Aiming at the problems, it is necessary to provide an identification method of the vehicle-mounted laser radar phyllostachys pubescens snow disaster damage based on a differential point cloud encryption strategy. Disclosure of Invention The invention aims to solve the problems in the background art and provides an identification method of the vehicle-mounted laser radar phyllostachys pubescens snow disaster damage based on a differential point cloud encryption strategy (namely, different point cloud generation methods are adopted according to different disaster damage types). The method has the innovation points that the problems of the loss of the cloud data of the points of the bamboo forest and the low disaster recognition accuracy are solved by combining a targeted image processing method of disaster types through layered reconstruction and principal component analysis technology. The aim of the invention can be achieved by the following technical scheme: an identification method of the mao bamboo forest snow disaster damage of the airborne laser radar based on a differential point cloud encryption strategy comprises the following steps: step one, data acquisition and input; Receiving airborne laser radar point cloud data; the airborne laser radar point cloud data comprises moso bamboo surface scanning point coordinates and laser reflection intensity, and average ground elevation is recorded, wherein each moso bamboo surface scanning point coordinate has a laser reflection intensity which is uniquely corresponding to each moso bamboo surface scanning point coordinate. Step two, data preprocessing; Preprocessing the acquired point cloud data, including data noise reduction, ground threshold separation and normalization processing, to obtain a normalization processing result of laser reflection intensity; And carrying out Gaussian filtering on all acquired scanning point coordinates on the surface of the moso bamboo, and then carrying out threshold separation and normalization processing on the laser reflection intensity of each Mao Zhubiao face points. Aiming at the collected phyllostachys pubescens surface scanning point coordinates, calculating Gaussian filtering output of each phyllostachys pubescens surface scanning point coordinate through Gaussian filtering; In a preferred mode of the invention, the Gaussian filter output of the scanning point coordinates of the surface of each moso bamboo is subjected to elevation threshold separation to obtain an overground part and an underground part, the laser reflection intensity of the overground part is normalized, the part with the z-axis component larger than the average ground height Cheng Wei is marked as the overground part, and the part with the z-axis component smaller than or equal to the average ground height Cheng Wei is marked as the underground part. Aiming at the overground part, obtaining the laser reflection intensity corresponding to the original coordinate point output by Gaussian filtering of each moso bamboo surface scanning point coordinate, and mapping the laser reflection intensity to a fixed range to obtain a laser reflection intensity norm