02-09-2017, 02:36 PM
The proposed models are specifically for enhancing images captured under an extremely dark / non-uniform lighting environment, as well as for a poor contrast environment. To perform the contrast enhancement, the proposed design adapts the window operation using the linear intensity scale and the cutout. By which the proposed models assign a partial range of the grayscale window to the full dynamic range. Therefore, it brings the contrast of the image into a particular grayscale window at the expense of saturating the pixels that fall out of the window. For the correction of illumination, the complete input image is improved with the proposed window model 1. And for low contrast enhancement, the input image is divided into four non-overlapping windows and each window is enhanced with the proposed window 2 model. mapped to get the output image. These dual models are designed and implemented in Spartan3E (XC3S500E-FG320) and Spartan6-LX9 (XC6SLX9-2CSG324) and their results are compared. Experimental results show that the proposed XSG-based designs adapt minimal resources as well as preserve image quality.