Study On Algorithms Of Image Edge Detection And Enhancement Based On Mathematical Morphology

Mathematical morphology is developed as a representative non-linear signal processing and analysis theory in recent years and has been widely applied in digital image processing. Mathematical morphology is a powerful tool for the geometric characteristics of the image analysis and processing. The basic idea is to use an object to carry the characteristics of structural elements to detect images, the collection of image information. It is precisely because of the structural element has a unique advantage; morphological image processing has become a major area of research in digital image processing. At the same time, because of the treatment of a pixel in morphological methods only needing its adjacent pixel information, which makes the inherent parallelism of the algorithm, is easy hardware implementation, which has attracted extensive attention.Image edge detection involves the study of image feature extraction, namely, how to identify the contours of image objects, It has become the focus of many scholars. The edge exists between the target and the background, objectives and goals, region and region, being able to outline the geometric outline of the characteristics of objects, transmission of a variety of image information, description the object's important feature, For people to describe or identify the target as well as explaining the images provide valuable, important characteristic parameters.Image enhancement refers to the specific needs of an image to highlight some of the information, at the same time to weaken or remove some unwanted information, also a process to improve image quality. The purpose of image enhancement is to make the characteristics of the image clearer, prominent, in order to facilitate the implementation of more advanced image processing and analysis. In image processing systems, image enhancement technology as a pretreatment part of the basic technology, is the most important aspects in the system. Image enhancement image processing technology as the most active branch has been widely used in scientific research, medical and health, industrial production, communications and so on.In this paper we careful study the current situation and development trends of edge detection algorithms and image enhancement algorithm at home and abroad. According to the disadvantages of the current algorithms we carefully analysis the previous algorithms and determine the scheme of research programs. Combination the order morphology transformation we propose a new edge detection and image enhancement algorithms. Order Morphology is a general image morphology that injects the "order statistics" into the idea of mathematical morphology. It can make the edge detection effectively, and also put a distinction between signals and noise effectively. Order Morphology is a combination of statistical theory and morphology. Order morphology transformation is indeed a percentage of filters with the natural filtering properties, therefore the structural elements also often referred to as the filter window. The transformation method is to order the pixel value under the structural elements. And then according to a certain percentage select the corresponding pixel values in the sequences to correspond to the pixel that is in the center of the module. If we select the maximum percentage it will be the Maximum filter, it can be used to detect the brightest point in the images; if we select the minimum percentage it will be the Minimum filter, it can be used to detect the faintest point in the images. If we select the medium it will be the medium filter.Percentile based on morphological transformation of image edge detection methods are the development studies of traditional methods of morphological edge detection. At present the basic idea of edge detection applied many structural elements is to use multi-scale structural elements or multi-direction elements to do the image morphological operations, apply easy morphological operators for edge detection, then synthesis the detected result of each structure to gain the effect of edge detection. This paper presents a new edge detection method; this method decides the scale and type of the structure of the element and the percentile value of the Math Morphological Transformation through the average of local area, maintains the edge directional through the structural elements in different directions. The edge detection operator suppresses noise on maximal degree, improve in all directions on the edge of properties, maintain the details of the original image information, can effectively enhance the effect of image edge detection.Image's local entropy is contribution to many pixels in the partial image window, it is not sensitive for single-point noise, the size reflects the information contained in the partial image, The greater the local entropy the smaller the grays change;The smaller the local entropy the more drastic changes in grays. The edge of the image will correspond to smaller local image entropy. In this paper, we present an edge detection method based on Sequence Morphological and local entropy. The method constructs the edge detection operator. The edge detection operator selects the different percentile value according to the different sizes, different types of structural elements. And then compute the local entropy of the transform result; seek the percentile values and structural elements which the local entropy is smaller. It will sharp the edge of the image according to the characteristics of image entropy. The experimental results show that the method has a strong ability to suppress noise and good edge detection performance compared to the traditional edge detection methods.Tsallis entropy is the expansion of the Shannon entropy form; it is more generalized entropy of non-expansion and adjust its non-scalability through the non-expansion parameter q. In fact, the nature of images have their own long-range microscopic memory and dissipation characteristics, the use of non-extensive statistical point of view can be more appropriate to describe the image itself. According to most current methods of image enhancement with noise sensitive and large calculation we in-depth study the related concepts and the nature of algebraic geometry of Order Morphology Transformation and present an image enhancement method based on Tsallis entropy and the Order Morphology Transformation. We can see from the simulation experiment that the image contrast has been enhanced significantly at the same time filter out the high frequency noise. And we also give histogram and observe the effect of image enhancement intuitively.