【正文】 respectively by using the method of morphological difference and calculating the information entropy of the edgedetected. Last,fuse together the edges to achieve the image the image with noise,the algorithm can filter out noise effectively, and the edge is continuous,smooth,clear experiment results show that the algorithm has good effect in edge detection,strong antinoise capacity,and the objective evaluation and visual effect are good, too. Keywords: image processing, edge detection, mathematical morphology, integration 1. Introduction Image edge is one of the most basic characteristic of the image。it not only contains a large amount of information,but also is an important basis about the image analysis and image ,image edge has a wide application detection is widely used in image segmentation,image matching, feature extraction,and other is also an important research direction of puter are many kinds of algorithms for image edge traditional algorithms for edge detection have Laplace operator,Roberts operator,So operator,Canny operator and so on [4],the main function of these operators is highpass filter,each of them has different characteristics and the traditional algorithms involve in orientation,so antinoise performance is generally poor,it is difficult to detect the edge of the plex image [5]. The basic idea of mathematics morphology is to use the structural element with a certain form to measure and to extract the corresponding shape in the image. Mathematics morphology can achieve the purposes of the image analysis and target recognition [6].It has unique advantages in image edge detection and could make the accuracy of edge detection and noise immunity to reach a certain balance [7].After 1985,mathematics morphology became a tool in the image processing,and shows a great application prospect order to improve the performance of xx 科技學(xué)院 2020 屆本科生畢業(yè)設(shè)計（論文）外文翻譯 2 morphological edge detection,the scholars mainly researched in the following three aspects during nearly thirty ,study the operators of the morphological edge of them use the plex type antinoise edge detection example,literalness put forward the improved operators about edge detection,but their results either cannot detect the details of the edge well,or cannot filter out the noise ,study how to select the structural element. Someone may use the muftistructure structural element,and someone may use muftiscale structural element [13, 14].But select a single type of structural element cannot detect the details of the image edge example,literature [13] only select singlescale structural element,the image edge is not ,study the image calculate weight coefficients according to the inverse proportion of image difference or according to the direct proportion of information entropy [15, 16].But when using the structural elements with different orientations to detect the image,the effect of edge fusion is not very ideal. The above studies have different defects order to improve the antinoise ability of the morphological edge detection and to make the edge with continuous,plete,without jagged,this paper proposes an adaptive muftistructure and muftiscale edge detection algorithm by improving the existing experimental results show that this algorithm is greater than the literalness . 2. The Basic Principle . The Basic Principle of Morphology Mathematics morphology is a nonlinear image processing based on set theory, now,it is applied to image analysis,pattern recognition,etc basic operations of mathematics morphology are:dilation,corrosion,opening and input Freyscale the grayscale morphology,the definition of dilation,erosion, opening and closing are shown as follows: Definition 1 Grayscale dilation operation is ( ) ( , ) m a x { ( 39。) ( 39。) | 39。 , 39。 }fsf s x y f x x y y s x y x x y y D x y D? ? ? ? ? ? ? ? ? Definition 2 Grayscale erosion operation is ( ) ( , ) m i n { ( 39。) ( 39。) | 39。 , 39。 }fsf s x y f x x y y s x y x x y y D x y D? ? ? ? ? ? ? ? ? Grayscale opening operation is erosion followed by dilation,it can smooth the contour of image and remove small extrudes. Grayscale closing operation is dilation followed by erosion,it can smooth the image outline and recover the holes and cracks. xx 科技學(xué)院 2020 屆本科生畢業(yè)設(shè)計（論文）外文翻譯 3 . The Existing Operators based on Morphology The edge detection based on morphology is to do the dilation and erosion operation by using structural element, then to establish certain mathematical operations and to get the relevant operator of image edge (x,y)Represents the function of the image edge, ( , )f xy Representative the input image, ( , )sxy and represents a structural element. According to the basic operation of morphology,there are some existing edge detection operators based on morphology. The edge detection operators which were put forward in literature [11] are listed as follows: 2 1 1 1m a x m a x { : : }d e deG G G G? 2 1 1 1m in m in { : : }d e deG G G G? 2 1 1 1m in2 1 1 1m a x2 2 2m a x m inm in{ : : }a x{ : : }d e ded e ded