Improved Method for Enhancing Dark Images based on CLAHE and Morphological Reconstruction

Abstract

Abstract The image acquired are often flawed due to various defects from environmental scenarios such as indoor lighting, cloudy weather etc especially at night time. The dark image contains compressed dynamic range that can be enhanced for understanding the detailed information. In this investigative work, an improved method for enhancing extremely dark images is proposed utilizing the concept of illumination reflection model. This improved technique is based on Contrast Limited Adaptive Histogram Equalization (CLAHE) and reconstruction done using morphological processing with Top-hat transformation. The image is perceived in HSV color system and V component is estimated. This intensity (V) component is normalized and the inverse of the intensity component is computed. Then CLAHE algorithm is applied on the negative image. Then multiscale image enhancement is applied to the resultant image. Gamma enhancement is used to adaptively adjust the brightness component of an image. After gamma enhancement, the results of two gamma enhanced images with different gamma values are obtained. Principal Component Analysis extracts the significant information from these images that can be used for image fusion. During the PCA based image fusion, the weight value is adaptively calculated by using morphological Top-hat transformation to improve the quality of the image and illuminate the inconsistent background pixels. The proposed method focuses detail enhancement of extremely dark images by preserving the edges and structures. Experimental validations and results show the better performance of the proposed method than the existing method in terms of qualitative and quantitative measures.