Applying Wavelets and Evolutionary algorithms to automatic image enhancement

Getting clear high detailed images with high contrast is an important task in many spheres of science and engineering. However, it's not always possible because of imperfection of devices or environment conditions. This all leaded to development of different methods of image enhancement. In this article a developed two-phase full-color image enhancement algorithm is described. During the first phase the picture is denoised. Wavelet transformation has been chosen to perform it, because it allows easily remove high-frequency parts. Also, noise components, especially big random surges of signal, could be presented like set of local features of signals. Noise can be reduced by thresholding this features. During the second phase brightness and contrast are automatically tuned up using evolutionary algorithm. Evolutionary algorithms, which are effective methods of multidimensional optimization, allow quick selection of optimal values of transformation parameters, using objective optimization criterion.