A Mixed Particle Swarm Optimization Algorithm's Application in Image/Video Super-Resolution Reconstruction

Aiming at the pixel correlation of image sequences in super-resolution reconstruction of multi-frame video, an effective image/video super-resolution reconstruction model is established, and the problem of multi-frame video's super-resolution reconstruction is transformed from low resolution image to high resolution image. Aiming at the defects of basic particle swarm optimization algorithm (BPSO), the genetic principle and ant colony mechanism are introduced into the algorithm, an improved PSO algorithm, called mixed PSO (MPSO), is proposed. A new algorithm (MPSO), including its particle input, particle position, motion equations and fitness criterion, are designed. The experimental result of road surveillance video super-resolution reconstruction verifies the feasibility and effectiveness of the proposed model and algorithm.