De-interlacing Algorithm Using Motion Adaptive and Edge-based Analysis

Until today, the interlaced scan format is widely used by the TV industry. The interlaced video is an easy way to halve the bandwidth, but exposes well known artifacts like line crawling, edge flicker, and jagged effects. Furthermore, an interlaced field is unsuitable for progressively scanned format used in LCD and plasma displays. De-interlacing is the process to convert interlaced format into progressive format. This work presents a novel edge-based motion adaptive de-interlacing algorithm to promote visual quality of video sequences. An interlaced field is divided and analyzed by utilizing block-based motion detection and edge detection. Different edge interpolation schemes are performed by considering the similarity of nearby edges and previous interpolation direction. The refinement is then utilized by referring to directional, temporal, and spatial information. Experiment results show that the proposed algorithm can promote image quality, especially in horizontal edges. The proposed algorithm also outperforms previous motion adaptive de-interlacing algorithms. This work should be useful to professionals in Image Processing and Consumer Electronics fields.

Chih-Huan Wu is pursuing the Ph.D. degree with Data Compression and Multimedia Communication Lab in electrical engineering at National Cheng Kung University, Tainan, Taiwan. His research interests include image processing and surveillance system. He is advised by Prof. Shen-Chuan Tai whose research interests include data compression and algorithms.