Point cloud denoising method, device, medium and product

Abstract

Embodiments of the present application provide a point cloud denoising method and device, a medium and a product, and relate to the technical field of artificial intelligence. The method comprises: obtaining target point cloud data for abnormal point elimination; constructing a denoising network based on a density-based hierarchical clustering algorithm and a business selection strategy according to the target point cloud data, and generating a denoising model based on the denoising network; inputting to-be-denoised point cloud data into the denoising model, and outputting denoised point cloud data. The scheme of the present application filters out isolated noise that may mislead path decision in advance, and clears the interference for subsequent path selection; the denoising network based on the density-based hierarchical clustering algorithm and the business selection strategy is constructed, the point cloud structure is accurately divided, the representative elite points are screened in combination with the business selection strategy, the path is expanded with the elite points as the core in an extreme case, the noise points are avoided from dominating the decision, and the problem of poor adaptability of a single path in an extreme scenario in the prior art is solved, and the denoising accuracy is improved.

Images