Due to the relatively occluded environment in the mining area, the driving route is fixed and the driving speed is slow, but the working conditions are complex and the environment is harsh. Therefore, the unmanned transportation of mining trucks in the mining area has been rapidly promoted and applied on a large scale in the mining scene.
It is understood that 2020-2025 is the key construction period for autonomous driving in Chinese mining areas, and the market volume is as high as 100 billion.
As the “eyes” of unmanned mining trucks, LiDAR has the advantages of high precision, long distance and high stability. It provides highly stable perception and accurate detection for unmanned mining trucks under harsh working conditions, and ensures the safety of vehicles in complex working conditions.
However, when working in the mining environment, dust is everywhere, not only gradually covering the surface of the LiDAR, but also forming a dust mist in front, as if a curtain covers the “eyes” of the unmanned mine trucks, affecting the detection quality of the LiDAR, and even lead to misjudgment by unmanned vehicles. Therefore, in the actual environmental operation of unmanned mining trucks, the hard impact of dust on LiDAR perception is a difficult problem faced by the industry.
LSLiDAR engineers have been deeply involved in the operation site of unmanned mining trucks for a long time. After years of exploration, they have conducted a large number of comparative experiments and calculations on the characteristics of dust and point cloud algorithms, and formed a set of LiDAR dust filtering algorithms suitable for mining truck operating conditions. This set of algorithms can be combined with any LiDAR of LSLiDAR to be mounted on unmanned vehicles to solve the problem of perception accuracy of unmanned driving in dusty environments such as mining areas, improve the operating efficiency and safety of unmanned driving, and first time break through domestic LiDAR unmanned application problems in mining areas.
Before the optimization of dust algorithm
Dust, Mine truck Dust point cloud can be detected
Coal ash can be detected
After the optimization of dust algorithm
Dust is optimized for removal Coal ash is identified and removed
After the dust dissipates, the ground is detected and the ground line gradually comes out
It can be seen from the point cloud image that the LSLiDAR LiDAR can identify and remove large pieces of dust, reducing the burden of filtering by the host computer; when the objects in the middle and far distances are not blocked by soot, the objects in the middle and far distances can still be detected.
