Parameter learning for the livewire image segmentation by particle swarm optimization

Livewire is an interactive segmentation tool can extract the boundary of a region with a mouse. The segmentation is based on the features of the pixels in the image. In the traditional livewire, the features have been assigned with fixed weights. In this paper, we design a learning phrase before the segmentation, where the particle swarm optimization(PSO) is applied to find more suitable weights. To make the PSO more effective, the initialization of the population are special designed, the iteration and the convergence are visualized, the start and stop of PSO are human-controlled. Experiments show that the PSO learning livewire has better performance than the livewire with fixed feature weights.