Research On Edge Detection Of Color Image

There are many ways of detecting to the edge of color image. Using the method of PCA and edge detection of gray image, the edge of color image can be detected effectively. But it would still missing information for image, and undermining the integrity of the picture on the edge. To this problem, there are some improvement and extension of image with color PCA approach, and find a way to improving the integrity of which the result of the detecting to the edge of the color image.From the analysis of the way that the edge detecting of color image, there are some principle can be find that if want to keep the integrity of the information for the images, it is need to loss few of information for images to each step of the image processing. There are two steps of the edge detecting of the color image:conversion of the color image and the process of the detecting of image.The first step is that the improving of the old way of conversion of the color picture. In the past, using PCA for conversion of the color image, only the first major ingredient be used to replacing the color image. Some information lost, even the further way that using the contribution to decide which major ingredient should be selected to replacing the color image. From the result of the analysis of using PCA to solve the image, it find that there are some information of edge of the image in every major ingredients, processing and comprehensive to that, the integrity of information for images can be kept Effectively.The second step is the researching of the methods of edge detection. In this step, the major works is that improve the way of edge detecting of small wave and mathematics of morphology, the target is improve the ability of detecting the weak edges of two methods, do that, more information of edge can be kept in the step of dividing image by threshold. Meanwhile, a method of edge clarify of picture be find which is based on the way of gradient features of image and mathematics of morphology.Finally, there is an analysis to the improved edge detecting of the color image. After that, it is a discussion to the scope of the PCA and the characteristics of the new small wave and new mathematics of morphology. A simulation is done to the comprehensive handling to the color image. Describing how to uniting the new PCA and the new methods of edge detection.In the other way, some solutions don't be done for the subject, and there are some Inadequate in the subject, all these need to be study and improve in future.