Application of Cellular Neural Network to Contour Detection in QuickBird Remotely Sensing Images Associated with Mathematical Morphology

For the purpose of extracting the features in high spatial resolution QuickBird panchromatic images, and of using the images into various fields, this paper presented a method to detect the contour of features in QuickBird remotely sensing images based on mathematical morphology (MM) integrated with cellular neural network (CNN). Firstly, remove the noise in images using open-closing morphological filter; secondly, utilize a CNN-based contour detection algorithm to detect the contour in the filtered images. The experimental results show that contour detection based on proposed approach is more effective than that of either morphological gradient algorithm-based or CNN contour detection algorithm based.