Adaptive Quantization for Predicting Transform-Based Point Cloud Compression

The representation of three-dimensional objects with point clouds is attracting increasing interest from researchers and practitioners. Since this requires a huge data volume, effective cloud compression techniques are required. One the most powerful solutions Moving Picture Experts Group geometry-based (G-PCC) emerging standard. In G-PCC lifting transform coding technique, an adaptive quantization method used to improve efficiency. Instead assigning same step size all points, increased according level detail traversal order. way, attributes more important points receive finer have smaller error than less ones. paper, we adapt approach predicting propose hardware-friendly weighting for quantization. Experimental results show that compared current test model, proposed can achieve average Bjontegaard delta rate −6.7%, −14.7%, −15.4%, −10.0% luma, chroma Cb, Cr, reflectance components, respectively on MPEG Cat1-A, Cat1-B, Cat3-fused Cat3-frame datasets.