Algorithm and architecture design of illumination changes adaptive motion estimation

Multi-view video coding has been obtained significant attention in research and the technologies certainly enable the development of consumer electronics in the near future. The discrepancy of local illumination change, which is always generated in multi-view video capture, always impairs the coding efficiency. The illumination change adaptive motion estimation (ICAME) is applied to find the disparity correlation between neighboring camera views. In this paper, we propose a hardware-friendly algorithm, called illumination change progressive motion estimation (ICPME), to efficiently improve the synchronization of data usage for VLSI design and exactly provide ultra-similar performance with general ICAME. The architecture design is also presented to provide a reliable implementation for current application of multi-view video in consumer electronics.