Application Of Multi-scale Morphology Algorithm In The Image Edge Detection

Edge detection is a very important image preprocessing method, the detectionresults directly determine the quality of subsequent processing, and so has been a hotissue in the field of digital image processing. So far, people have been proposed include:classic operator, wavelet transforms, fuzzy theory operator, morphological operators,and so on. But these operators could not solve the edge to locate and remove thiscontradiction of the noise process, how to improve the precision of the edge detection,and obtain good effect in the testing process that is a difficulty area, also this paperdirection.First, the paper describes the significance of the edge detection, the research statusat home and abroad. Then introduces several classic operators, and the simulationexperiments, analyses some problems exist in these types of operator. Chapter III、according to traditional classical operators of the existence of noise sensitive、can’t get amore complete edge, the amount of calculation and other defects, this paper propose amulti-scale morphological edge detection operator-MSM algorithm, experiment in thegray image with MSM algorithm and the traditional operators. The results showed thatcompared with the traditional operators, MSM algorithm has good noise immunity,robustness, and detects the edge of a more complete. Chapter IV, MSM algorithm willbe applied to the color images of the two colors in common model, Experimental resultsshow that MSM algorithm has a great advantage in color image edge detectioncompared with the traditional operator. Finally summary of the work done in this article,and the prospect of future edge detection technology development.