A New Satellite Image Segmentation Enhancement Technique for Weak Image Boundaries Based on Active Contours and Level Set Method

Level Set Method (LSM) and Snakes Method (SM) are two different procedures used in image processing to locate objects location and boundaries. Both of these methods have their own advantages and limitations. In this research, a new algorithm for image processing of the weak gradient features was developed to improve the overall image boundary detection system. This algorithm was based on the active contour model in conjunction with level set method to enhance the images detection approach. The algorithm presented a new technique to incorporate the advantages of both LSM and SM. First, different bands of satellite image from a region were extracted from the satellite scene and then by linear combination of these images, the obtained image was resulted enhanced image.After combination the bands of satellite image, the initial segmentation by LSM was transformed and used as an input for the SM and began its evolvement to the interested object boundary. The results showed that the algorithm can deal with low contrast images and features on them, demonstrates the segmentation accuracy under weak image boundaries, which responsible for lacking accuracy in image detecting techniques. Thus, better segmentation and boundary detecting for the satellite images were achieved and ability of the system to improve low contrast images and low gradient features increased and as a result, they can deal with over and under segmentation.