POINT CLOUD DENOISING USING NORMAL VECTOR-BASED GRAPH WAVELET SHRINKAGE

Many applications that use point clouds, such as 3D immersive tele-presence, suffer from geometric quality degradation. This noise may be caused by measurement errors of the capturing device or by the point cloud estimation method. In this paper, we propose a novel graph-based point cloud denoising approach using the spectral graph wavelet transform (SGWT) and graph wavelet shrinkage based on the normal vector of a point. Unlike conventional SGWT-based denoising methods, the proposed wavelet shrinkage thresholds are determined by the geometric structure of the point cloud and are different for each point. Excessive wavelet shrinkage leads to the loss of complex geometric structure and can be prevented by shrinking only as much as needed by considering the point cloud structure. Experimental results show that the proposed method achieves the best denoising accuracy compared with state-of-the-art point cloud denoising methods.