Generalized Fuzzy Enhancement Adaptively For Image Based On Improved Nonlinear Gain And Bilateral Filter

In the field of image processing, image enhancement technology plays an important role because of its strong practical value and flexibility. On one hand, through the technology, images can be processed according to the interests of observers so as to meet their requirements; On the other hand, it is necessary to pre process images to meet the recognition ability of computers when using it to analyze images accurately. In addition, when images imaging, it is inevitable to be interfered by some objective factors such as weather, light and so on, which makes the quality of images deteriorate and affects their use value. Therefore in order to obtain high quality images, enhance measures must be taken.In this paper, a new algorithm of images enhancement is proposed based on the study of image enhancement techniques. To enhance details, remove noise and improve contrast at the same time, images are separated to high-frequency and low -frequency sub bands by wavelet transform. The former is largely made up of noise and details information, the latter is mainly include background information. Owing to the different contents of the two parts, there are different ways to deal with them. The high-frequency coefficients are divided into the noise and the details by the Bias shrinkage threshold and the nonlinear gain function is used to deal with them separately in order to enlarge the image details and reduce the noise. Due to the small part of noise included in the low-frequency sub band, a bilateral filter should be applied for noise suppression firstly and then the generalized fuzzy operator is made to improve the contrast of images.In order to improve the performance and adaptability of the algorithm, the nonlinear gain function is ameliorated by adding regulatory parameter that can make the image more clear and rich in details. Meanwhile a method to calculate the optimal value of the parameter c in the nonlinear gain function is provided based on the entropy. The bilateral filter used in low-frequency sub band is optimized which proposed an adaptive way to determine the value of range parameter through a large number of experiments and fast method for bilateral filtering is realized.Last, through the genetic algorithm, the generalized fuzzy enhancement model is simplified as the model of finding optimal value of parameters.Simulating the algorithm and comparing it with other algorithms, the results show that it has higher contrast, definition, entropy, peak signal to noise ratio, which can improve the image quality greatly.