Image Segmentation with Simplified PCNN

Image segmentation is an important task for higher level image processing. A simplified pulse coupled neural network (PCNN) was proposed in this study. The comparative experiments were implemented to segment images by Otsu method, improved PCNN and our simplified PCNN algorithm. The mean values of Hausdorff distance and Tanimoto coefficient of our simplified PCNN algorithm were 5.41 +/- 0.03 and 0.944 +/- 0.008, respectively, which were in the same magnitude comparing with the results of other segmentation algorithms. However, the mean running time of our PCNN algorithm was only 3.38s, which was much less than those of other methods. The experimental results demonstrated that the proposed PCNN algorithm had the advantage of short running time of segmentation with satisfactory segmentation accuracy.