Bringing luminance histograms of low-contrast digital images to two-level piecewise uniform distribution

The paper describes the problems of processing digital images for enhancing contrast to increase the distinguishability of details, performed by histogram methods that approximate the reduction of the histogram of image brightness to a given distribution. Using the uniform distribution algorithm does not take much time, but it does not regulate the brightness and contrast of the image processed. It is possible to achieve better distinguishability of details using a normal distribution, as compared to the uniform distribution, but in this case the solution requires numerical methods which take much more time. There has been proposed the algorithm of controlling the contrast of a digital image by bringing the luma histogram to a distribution determined by a piecewise constant differential distribution function with two levels of values. This algorithm helps to control the average brightness and contrast, without resorting to numerical methods, very dark and very light low-contrast images having greater distinguishability of details. The mathematical foundations of the proposed algorithm are presented. There have been studied the possibilities of increasing the detail distinguishability in the processed low-contrast image, as compared to the popular method based on bringing the brightness histogram to a uniform distribution.