A distributed Particle Swarm Optimization algorithm for block motion estimation using the strategies of diffusion adaptation

Inter-prediction motion estimation (ME) is solved in a distributed manner by a network of cooperating and learning agents. The block motion estimation problem is formulated as the optimization of a global cost function that is the sum of individual sub-problems. A network of agents is then employed to address this problem where each agent iteratively solves its own sub-problem using a modified Particle Swarm Optimization (PSO) algorithm. Diffusion strategies are employed to allow the agents to cooperate and diffuse information in real-time in order to reach the common minimizer of the global cost function.