Point cloud data generation method and perception capability verification method of laser radar

The invention provides a point cloud data generation method and a perception capability verification method of a laser radar, and relates to the field of unmanned driving and unmanned vehicles. The point cloud data generation method comprises the following steps: placing a plurality of vertexes in a target three-dimensional space region in a simulation environment; related data of the multiple vertexes are obtained, the related data of the multiple vertexes are transmitted to a graphics processor for operation, and the graphics processor is used for simulating noise in an actual dust raising environment based on the related data and a preset noise function to obtain a noise value of each vertex; on the basis of comparison between the noise value of each vertex and a preset threshold value, screening out visible vertexes in the plurality of vertexes; rendering corresponding drawing points based on the visible vertexes; and simulating the to-be-tested laser radar to sample the drawing points to obtain simulated