Research And Implementation Of Image Enhancement Technology

About75%of the information humans obtained from outside world comes from visual sense. As the prime carrier of visual information, the quality of images will affect the amount of information people received. Image enhancement aims to improve the image visual effect. Appropriate image enhancement method can improve the amount of information in the image and make people’s work and life more convenient. As a result, the research of image enhancement has vital significance.The front half part of this paper depicts some traditional image enhancement methods, including:Histogram processing; Gray level transformation; Image smoothing; Image sharpening and the traditional method for the color image enhancement. Traditional image enhancement algorithms is the foundation of further research on image enhancement algorithm, so the image enhancement algorithm plays an important role.After the traditional image enhancement method, some improving methods of traditional image enhancement algorithm was introduced subsequently, including:the improvement of average filtering and median filtering method and histogram processing based on local histogram. Average filtering and median filtering have analogous working mechanism, so do their improved methods. Improved method can effectively restrain the fuzzy phenomenon caused by the average filtering and median filtering. Local histogram processing method based on histogram is suitable for the image histogram has certain characteristics of the image, the method does not change the overall characteristics of the image, but can enhance the visibility of the image details.Finally, medical endoscope color image enhancement algorithm was further studied. A simple and effective enhancement algorithm was employed according to the particularity of endoscope image. This algorithm includes two steps:gray image enhancement and image recolor. Firstly, convert endoscopy of the RGB color image to a spectral image, calculate the enhanced image with maximum entropy as the base images; Second,through the calculation of Euclidean distance of base image and standard image, according to the Euclidean distance to the desires of the base image to match the color;Finally convert them to RGB color images. Due to the relatively large amount of calculation of the algorithm, so the parallel computing approach was used in the experiment to speed up the implementation of the experiment.