Dynamic multi-histogram equalisation for image contrast enhancement with improved brightness preservation

Histogram Equalization is the efficient method for enhancing the contrast of the image. However, it causes over-enhancement so it fails to preserve the natural appearance of an image. To overcome this drawback, Multi-HE methods are used to preserve both brightness and natural appearance of the image. In proposed technique, the histogram of an input image is divided into multiple sub-histograms based on mean and median values as thresholds. The narrow segments are identified using the factor which depends on the number of segments, dynamic range of each segment and intensity level. After identifying the narrow segments, it is scaled to the full dynamic range while the wider segments remain unaltered. The histogram equalization is applied to each segment individually. Finally the normalization is to be carried out for equalized histogram to avoid intensity saturation and uneven distribution of bins. When the number of segments is increased, the contrast of the image is increased with improved brightness preservation. But the natural appearance of the image is not preserved with loss of information. These characteristics are found by the parameters namely Universal Image Quality and Discrete Entropy parameter. The comparison of various histogram equalization techniques is performed using MedCalc statistical software.