作者: Yi-Jen Chiu , Hyeong-Seok V. Ha
DOI:
关键词: Depth map 、 Computer science 、 Pixel 、 Stereo cameras 、 Video tracking 、 Computer vision 、 Multiview Video Coding 、 Video capture 、 Video processing 、 Artificial intelligence 、 Video denoising
摘要: Systems, apparatus and methods are related to real-time automatic conversion of 2-dimension images or video 3-dimensional stereo video. The method for converting a 2-dimensional (2D) image (3D) comprises: determining sharpness values individual pixels 2D input image; performing first region second region-based sampling the values; scaling pluralities smoothing scaled interpolated merging generate depth map.