A full 3D occupied voxel terrain modeling method based on lidar

Abstract

The invention discloses a laser radar-based full 3D occupied voxel terrain modeling method, comprising: step 10) data acquisition and coordinate transformation; step 20) establishment of voxels: for the gap area from the starting point to the end point of light, calculate its The entry and exit points of the grid, and construct negative voxels to represent the idle state; for the end position area, associate it with several networks, and construct corresponding positive voxels to represent the state of obstacle occupancy; Step 30) Constraint-based volume Meta update: add the established positive voxel or negative voxel to the existing voxel linked list of the associated grid; then perform update processing; step 40) measure occupancy state: calculate the probability of the updated voxel occupancy state, Determine the state of occupation. This modeling method can solve the uncertainty of terrain modeling, the accuracy of spatial description and the efficiency of 3D terrain environment modeling, and realize accurate, efficient and real-time modeling of field terrain environment.

Description

technical field [0001] The invention belongs to the field of field operation robots, unmanned driving and other application fields, and in particular relates to a laser radar-based full 3D occupancy voxel terrain modeling method. Background technique [0002] In the field of autonomous navigation of unmanned platforms in outdoor environments, an accurate, efficient and robust 3D terrain model plays a vital role, directly determining the operating efficiency and overall economic benefits of the entire system. Different from the indoor environment, the outdoor environment has the characteristics of high dimensionality and complex structure, and the large-scale characteristics of the environment intensify the real-time contradiction of the terrain model. [0003] The perception of the spatial environment is the primary problem in 3D terrain modeling. Common perception methods include: binocular cameras, lidar, and ultrasonic ranging. During the working process of the binocula...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to propose a lidar-based full 3D occupancy voxel terrain modeling method to solve the uncertainty problem of terrain modeling in the full 3D occupancy voxel terrain modeling method based on uncertainty set representation, space Accuracy of description and efficiency of 3D terrain environment modeling to achieve accurate, efficient and real-time modeling of field terrain environment

Images