Motion adaptive de-interlacing algorithm with edge-based low-complexity interpolation method

In this paper, we propose an efficient motion adaptive de-interlacing algorithm with edge-oriented low-complex interpolation (LCI) based methods. The proposed algorithm is composed of the improved 4-field motion detection and three de-interlacing modes, which are the inter-field average, the LCI with the expanded edge detection, and the 3-dimension (3-D) median LCI method. First, the proposed improved 4-field motion detection classifies the interlaced pixel into a static, a motion, or an uncertain status. Second, the corresponding de-interlacing method is applied to interpolate the de-interlaced pixel and it also preserves the original edge information. Simulation results show that our proposed edge-preservation LCI-based method provides better objective and subjective de-interlacing results than the previous non-motion compensated de-interlacing methods. Moreover, it achieves the same visually de-interlacing performance as the motion compensated based methods, which need large complexities for the de-interlacing process.