A Method Of Edge Detection Based On Surfacelet Transform

Edge is one of the most basic characters of an image, which contains a lot of useful information that used for image processing including image identification and image analysis. It is one of the important parameters for people to describe or identify an object, or explain an image. Edge detection is one of the most classic operations in image processing, image analysis and computer vision. It is one of the basic methods for pattern recognition and image information extraction.There are lot enough methods of edge detection in the field nowadays, however, still have they flaws in some place which can be improved in theory. So there is still enough space for us to improve the methods of edge detection.Multi-scale geometric analysis has made very significant achievements and breakthroughs in the field of edge detection recent years. Because of the accuracy in searching edges, more and more scholars try to introduce multi-scale geometric analysis into methods of edge detection. Surfacelet transform, which has been proposed by Yue Lu and M.N. DO in 1997 is an excellent multi-scale analysis tool that has shown advantages in redundancy and multi-directions.In this paper, firstly, some classic methods of edge detection have been introduced and analyzed in the second sections. And, after doing a research in the theories of multi-rate system, NDFB and surfacelet, a method of edge detection based on surfacelet transform has been introduced, which have a good visual effect and a better performance in anti-noising than the classic methods.