Automatic Image Segmentation Using Pulse Coupled Neural Network and Independent Component Analysis

In order to determine the cyclic iteration times of Pulse Coupled Neural Network (PCNN) image segmentation effectively, and obtain the image segmentation result including regions of interest(ROI), an image segmentation method based on PCNN and Independent Component Analysis (ICA) is proposed in this paper. First, extract the independent signal sources corresponding to the image including ROI through ICA. Then, detect the signal sources corresponding to the segmentation result of the each iteration to achieve the output of target image including ROI. The experimental results demonstrate its validity, and the images including ROI correspond to unified independent signal sources. Evaluations of the proposed method are, the average cyclic iteration times N is 5.6, the average runtime is 0.08s, and the accuracy of target image outputs is 98.6%.