Gray Scale Image Enhancement Based On Pcnn And Pso

As one of the visual objects，image plays an important role in the messagefrom which the people can get information，so how to process the imagebecomes more and more important. With the development of science andtechnology，Digital Image Processing technology has widely used in the fieldsof biomedical，remote sensing，aerospace，military，public security，industrialapplications and video conferencing and so on. It has brought a great deal ofeconomic and social benefits to people.When gray image generated，transferred and converted，it could be becomeblurring，visibility loss，contrast reduction due to the light，the device，as wellas external noise and other reasons. So we need to take method to deal with thegray image to improve image quality and image visibility，only this we could goon further work such as identification and analysis. For gray imageenhancement，we can use one or more methods according to certain purpose orneeds. We don’t care about the reason of the degraded images，only care abouthow to improve contrast and sharpness of the image. Currently，in the selectionof method of the gray image enhancement has not uniform rules. In practice，wecan choose several enhancement method at the same time to process the image.The gray image enhancement algorithm has conducted in-depth research inthis paper. Main research work of the dissertation are shown as following：（1）Gray Transformation Function which based on Particle SwarmOptimization (PSO) and also called Beta Function has been proved. The BetaFunction is a gray-scale transformation function put forward by Tubbs in1987.There are two parameters needed to determine in the function. By using PSO tooptimize them has been an effective method. In this paper，the Beta Function isdescribed， and simulation experiments have been conduct to verify theeffectiveness of the function.（2）Pulse Coupled Neural Network (PCNN) model has been improved.According to analyze the working principle of PCNN model，this article gets ridof some parameters which has less influence on the operational mechanism ofPCNN model. At the same time， β-parameter is improved to achieve self-adaption setting. The experimental results show that improved PCNN isbetter than the conventional method in filtering effect，the edges of processedimage are more distinct.（3）In order to prepare for subsequent work (such as the identification andanalysis etc)，the paper has took improved PCNN and Beta function based on thePSO together to process images with noise. The experimental results show thatthe method not only removes the noise，but also extends the image gradation，and enhances edge sharpness of image，and enhances the overall effect of theimage，it is an effective method of image enhancement.