Adaptive linear fusion for image enhancement with hardware verification

In a previous paper in [I], we presented an approach to image enhancement where linear fusion scheme was employed. In the approach, only one input image is required and from the input image anther image is derived by histogram equalization. With the two images, a linear fusion is performed with a fixed weight. In this paper, a simple but effective mechanism is devised to determine an adaptive weight for the linear fusion. To justify the proposed approach, hardware verification is conducted and three scenarios are provided where low lux, high contrast, and normal cases are considered. Experimental results indicate that the proposed approach is feasible and promising in the real-world applications.