Research On Image Edge Detection Based On Mallat Algorithm

Human beings perceive the external world through visual,auditory,tactile and other systems,so as to understand and understand the world.Among them,visual system is the most important way of human perception.Human perception of an object to a great extent,depends on the edge of the reason for this is that the edge is a region in the image and the other at the boundary of a region(junction),is the place that attribute mutations,is one of the largest local uncertainty in image,is also the largest concentration of image information,thus,the characteristics of the basic,image edge is the same,it can keep the shape of the objects in the image information in the premise of greatly reducing the information,so edge detection is the most basic and important in the field of image processing,Its solution is very important for high level feature extraction,feature description,target recognition and image understanding.In this paper,the basic theoretical knowledge of wavelet and wavelet transform is described,the application of edge detection in image processing is introduced,the edge detection methods are classified,and the classical edge detection operators are introduced and compared.In this paper,an image edge detection method based on Mallat algorithm and orthogonal wavelet transform is proposed.Second,the complex coefficient all-pass filter constructed in this paper is less affected by noise than the commonly used real coefficient filter,so that the detection results are more accurate.Simulation results show that this method is simple,fast and effective.