Edge Detection Algorithm For Remote Sensing Image Based On Mathematical Morphology

Digital image processing is improving an image into another image. Edgedetection is an important direction of image processing. Edge image as the basicfeather of the image includes the feather of the image space and image contour.It can be used in the analysis and treatment of superior layer of the featuredescription, image segmentation, image enhancement, image restoration, patternrecognition and image compression. It is available to further comprehension andanalysis of the image. So many scholars dedicate to the research of the theoryand practice for image edge detection. Many feasible detect methods areformed.Image edge is the description of the remarkable change of gray betweenthe neighboring pixels in the image. The distributing of this change can bedescribed by mathematic grads. This paper study several traditional and newedge detection methods based on differential coefficient in detail. Theiradvantages and disadvantages are analyzed impersonally and the experienceresults are provided. In addition, further study of image edge detection usingmorphology theory is carried out. In this text, a detailed introduction about theorigin and development of the mathematical morphology is given. All kinds ofoperations and characters of the expansion, erosion, opening operation andclosing operation in mathematical morphology based on binary-morphology andgray-morphology are studied emphatically. Through the analysis, present theimportance of the structure element in mathematical morphology is present andconstruct a new arithmetic of the edge detection for remote sensing algorithm isconstructed. Simulation results by Matlab indicate that this new method isadvanced and applied.The study of this paper implement the filtering of the noise in the remotesensing image using mathematical morphological method and detect the edge ofthe remote sensing image precisely. Through the analysis of the simulationresults, the capability in many aspects and the problem of the algorithms issummarized. Many experiences are accumulated for further research.