Adaptive Algorithm For Edge Detection Research

Edge is one of the most fundamental and significant features. Edge detection is always one of the most classical studying projects of computer vision and image processing field. There are many edge detection algorithms including the algorithms of the basing on the gradient operator and the algorithms of the LoG operator. Following this perspective, the paper discusses the following two aspects.On the one hand in order to avoid the defects of the gradient operator, this paper presents a novel method for detection of the edges-A Self-adaptable Method of Edge-detection Based on the Gradient Magnitude. The gradient magnitude mean and the variance of every pixel in the image relative to the gradient magnitude mean are calculated according to the method proposed in the paper. Then, the best threshold value is calculated basing on the gradient magnitude mean and the variance of the image. Finally, the noise of initialization edge which is gained by according as the best threshold value is eliminated by the method of eliminating noise proposed in the paper. So the final edge is gained. The experimental results show that the novel method could enhance the edge and restrain the noise of the image at the same time.On the other hand in order to avoid the insufficient of the LOG operator, this paper presents a novel method for detection of the edge-A Self-adaptable Method Based on the edge detection using LOG. We gained the entropy of the Gray level co-occurrence matrix and the Gassian space coefficient from the image of the database. Then, the relation between the entropy of the Gray level co-occurrence matrix and the Gassian space coefficient is elicited using the curve fitting method. Optimum Gassian space coefficient Of LOG operator can be self-adaptable acquired basing on the entropy of the concrete image and the relation. The experimental results show that the LOG operator not only effective controls most noise of the images, but also locates the edge accurately using the self-adaptable Gassian space coefficient.The studying results introduced in this paper expanded the theory and applying practice in edge detection. It has value of theory and academia.