Evolutionary approach and Spline Interpolation for contrast enhancement

In this paper, image contrast enhancement techniques are explored and two novel approaches are proposed. The first approach uses a Super Resolution based Convolution kernel in Lab colour space which is optimized using Genetic Algorithms, for the evolutionary generation of the kernel values. The employed fitness function tends to provide a balance between contrast and naturalness of images. The second approach is called Spline Interpolation which works by fitting a continuous three dimensional curve through (0,0,0), the current pixel´s RGB value and (255,255,255). These methods are compared with standard techniques on a database of bio-medical images and general images. The results indicate that the proposed methods overcome the drawbacks of existing methods.