Proc. SPIE. 8784, Fifth International Conference on Machine Vision (ICMV 2012): Algorithms, Pattern Recognition, and Basic Technologies
KEYWORDS: Image processing algorithms and systems, Detection and tracking algorithms, Image segmentation, Image processing, Particles, Machine vision, Human vision and color perception, Particle swarm optimization, Optimization (mathematics), Image information entropy
An improved watershed image segmentation algorithm is proposed to solve the problem of over-segmentation by
classical watershed algorithm. The new algorithm combines region growing with classical watershed algorithm. The key
to region growing lies in choosing a growing threshold to reach a desired result of image segmentation. An entropy
evaluation criterion is constructed to determine the optimal threshold. Considering the entropy evaluation criterion as an
objective function, the particle swarm optimization algorithm is employed to search global optimization of the objective
function. Experimental results show that this new algorithm can solve the problem of over-segmentation effectively.