A New Image Enhancement Algorithm Using Just Noticeable Difference and Histogram Equalization

Abstract

One common way to save the battery life of a portable device is to reduce the LCD backlight intensity, but result in poor quality image. The dim backlight image enhanced to bright image especially for the low-luminance image areas. It has been envisioned as the de-facto solution to the quality image costs of electronic devices such as smart phones. The enhancement algorithm proposed in this paper brightness and visibility prediction for dim image. Based on just noticeable difference (JND) theory and the Human Visual System response model, the algorithm effectively enhances the visibility of image details in dark regions without affecting the contrast of bright regions. Apply the just noticeable difference theory and find the histogram bin finding. This shows the bin table. Finally apply the histogram equalization method. The boosting and color restoration are performed. The purpose of using visibility prediction and brightness for dim backlight image without detail loss and color degradation.