Taipei：Asia-Pacific Education & Research Association
In recent years, the price of depth camera became low, so that researchers can use depth camera to do more application. For computer vision, depth images can provide more useful information. However, generally there are some problems in depth image, such as holes, incomplete edge, and temporal random fluctuations. Conventional inpainting approach must rely on color image and it cannot be processed in real time. Therefore, this paper proposes a real time depth image inpainting method. First, we use background subtraction and mask filter to patch up the no-measured pixels, and then using the relationship between successive depth images to remove temporal random fluctuations. Finally, using erosion and dilation smooth the edge. Experimental results outperform than traditional one.
Proceedings of International Research Conference on Information Technology and Computer Sciences, pp.15-23