Real-Time Image Smoothing Based on True Edges

Xujie Li,Hanli Zhao,Xiaogang Jin,Xiaochun Qin

DOI: https://doi.org/10.1109/CAD/Graphics.2011.16

2011-01-01

Abstract:This paper propose a new real-time image smoothing algorithm based on true edges. We observed that the pixels marked as edge points by the well-known Canny operator are as close as possible to the center of true edges. This gives us an opportunity to judge how well edges are preserved for an edge-preserving smoothing filter. We therefore propose a nonlinear edge-preserving interpolation technique guided by the Canny true edges in an image. Our new edge-preserving smoothing algorithm smoothes high-contrast details and, at the same time, effectively preserves the true edges in the input image. In order to achieve a real-time performance, we further implement the algorithm using NVIDIA CUDA by taking advantage of the high parallelism of modern graphics hardware. The proposed algorithm smoothes low-contrast regions while effectively preserving true edges, and can also be used to produce vivid abstracted effects from realistic photographs.