A Novel Approach for Contrast Enhancement of Gray Scale Images using Multiscale Morphology

Generally image enhancement techniques are applied for improvement of gray scale images as well as on medical and scientific images like x-ray images, MRI images, ultrasound images and satellite images etc. The present work is based on the enhancement of gray scale images using multiscale morphological techniques with different structuring elements at different scales. The structuring elements like disk, line, square, diamond are selected to calculate top-hat transformation, bot-hat transformation respectively on different scales. The proposed algorithm is used to extract the white and black region of an image and then calculate their sum of extracted regions. Subsequently, black and white regions are calculated at the neighboring scales also with their calculated sum. The summation of the results reveals the enhanced image. From the enhanced image, again black and white regions are extracted and the sum is obtained. The results of the white and black regions and at its neighboring scales are combined with the first enhanced image to obtain the final image. The results are compared with the other conventional techniques like un-sharp masking, contrast stretching, with the help of Background variance (BV), Detail Variance (DV) and Entropy as performance metrics. Thus the final enhanced image shows better results both quantitatively and qualitatively.