The Edge Information Analysis by Mathematical Geomorphology of Image

The edge enhancement and extraction are the important aspects for image analysis.There are many researches which have been performed for that purpose.The difficulties are that the edges are sometimes not very distinct and the noise may be extracted as long as the edges be-ing found.Several approaches have been proposed to resolve those problems.In this paper,the mathematical geomorphology is considered as a new way while being applied for remote sensed image.The details of the principles have been discussed.As the operation of dilation is sensitive to the edges,the effects are more distinguished when compared with the original image.Same effects could be found when using the operation of erosion,while the analysis for opening and closing has shown the better results with less influence of noise.The discussion about the combination of different operations has been also introduced for further applications.The image resulted from dilation and that from erosion can be put into a new one in which the edges could be enhanced with precise positions.Other combinations such as that for opening and closing can have the better effects than that by using dilation or erosion only.The structure elements of different size of windows have been discussed according to their effects.The experiments for comparison with conventional gradient operator show that the method of mathematical morphology has the advantages of edge enhancement and less influence of noise.