Online Back-Lit Image Enhancement Based on Learning Restoration

This paper presents a method for online image restoration and enhancement via learning SVM model. The proposed method first obtains the images with the camera's laptop, then, process this image with learning approach via machine learning model (i.e. SVM model), where each image is divided into front-lit and back-lit segments. After that, the front-lit segment is expanded with the exposure of each pixel. Finally, both segments are merged together to produce the final restored image. Experiments and comparisons are performed on the proposed method, and shows that it outperformed standard restoration method.