Image Edge Detection And Application

In image processing ,as a basic characteristic, the edge of the image is often applied to high level domain, and the same time which is widely used in the recognition,segmentation,intensification and compress of the image. Edge detection is one important subject in the field of computer vision and image measuring.In the thesis, we introduce some traditional detection algorithms. And discuss the representative algorithm of image edge detection, such as Sobel operator,Prewitt operator Roberts operator,LOG operator,Laplacian operator and also analyzes the advantages and disadvantages of the operators. Based on the same standard image picture, experimental results of each algorithms are given, then analyzes the results.This thesis introduces fundamental theory of Canny operator, and carries on its analysis and evaluation. On this foundation, an improved Canny edge detection algorithm which is to be put forward. We improved the Canny algorithm by replacing the Gaussian convolution with Adaptive Edge Preserving Filtering. According to the characteristic of revulsion every point gray scale, this algorithm adaptively changes coefficients of the filter, and also can effectively protect the corner edge.And then, a new algorithm is proposed instead of the double threshold one, It divides image into sub images, and which detects and connects the edges in accordance to the difference of the gradient direction between the edge and the noise, so as to protect efficiently the details of all low contrast edges with the noise suppressed simultaneously. The new approach is superior to the conventional one. The results of the experiment picture indicate that improved algorithm has the better accuracy and precision in the edge detection.At last, it accounts for the generation of title, and summarizes works achieved. The study direction in the near future is pointed out in the paper.