Study Of Digital Image Processing Algorithms Based On The .net Framework

In our information society of the 21st century, image contains a lot of information. Image is the main media through which we can understand and exchange information. The image processing application field involves almost every aspect of human beings, and with the development of technology, the application of digital image processing will continue to expand. Compared with analog images, digital images have many advantages:(1) The quality of digital images will remain constant during storage and transmission; (2) Digital image processing can achieve better effect, an analog image can be digitalized to a right size of two dimensions arrays, and the image processing program is of generality. (3) Extensive applicability:the images can come from many sources, and it can be a visible or invisible image. (4) Highly flexible:the digital images can not only be processed by linear operations but also be processed by nonlinear operations. Therefore, this paper compares the classic algorithms in the field of image processing, and also uses C# language to implement part of the algorithm base.Net, as follows:(1)Achieve the basic algorithms of image processing, such as histogram equalization, local smoothing method, overrun pixel smoothing method. These algorithms improve the image effects. The experimental results show that, when an image background and foreground are too bright or too dark, histogram equalization can achieves obvious effects, especially better in the X-ray images with bone structure, and underexposed photo’s sound rendering details.(2) Implement image segmentation related algorithms, such as Canny edge detection, Roberts operator, Sobel operator and Priwwit operator. These algorithms are the basic components of the image recognition and analysis. These algorithms are important technology in image processing. They can extract relevant features of image. The experimental results show that the above algorithms can identify the actual edges of the image as much as possible and the location of the point.(3) Implement Hough transformation, edge extraction for image description algorithm. This kind of algorithm represents and describes the divided regions. So the "state of nature" pixels are more suitable for computer processing. The experimental results show that the image description will not be affected by graphics’ rotate and transform. According to the needs, we can use external of the target to represent the shape feature. Internal of the target can also be used to represent the attributes, such as the color or texture of the region.(4) Implement the Fourier transformation of the image transformation algorithms. These algorithms can make image processing simpler and more efficient, and it is likely to see certain "things" which may be not easy to see in the airspace of the image. The experimental results show that, after the Fourier transformation we can get magnitude image and phase image. The frequency of the image is characterized by the indicators of the gray-scale variation in the intensity of the image, and the low frequency portion of the image spectrum substantially determines the gray-scale variation, while the high frequency portion of the image spectrum determines the details of the changes in the image. The phase spectrum of the image determines the look of the image. If another image of the amplitude combines with the phase of the image, we can see the look of the image; only gray-scale variation is changed, because the phase decides the direction of the change of amplitude, and determines the approximate appearance of the image.With those algorithms above, we also build a digital image processing platform based on.Net, providing a simple and intuitive operation interface. The experimental results show that the platform is designed and implemented with usability, intuitive, scalability and maintainability. The platform facilitates the subsequent image processing, and can also be directly applied to the teaching of comparative experiments. In addition, we can also conduct secondary development; there is a good prospect of application.