Multi-Exposure Image Fusion

Abstract

There is information in scene that normal camera is unable to capture using photography due to presence of under and over exposed regions in scene. Our goal is to capture all the information present in scene even the information that is present in over and under exposed region. Due to advancement of technology many techniques have been proposed, yet there is a need for efficient, cost effective and reliable technique that can produce perceptually appealing, sharp and more informative fused image. In current research multi-exposure image fusion technique is proposed for differently exposed images. Different scene are used for testing of proposed algorithm like indoor, outdoor, natural scene etc. Images may contain noise due to camera jittering or sensors used for capturing of multi-exposed images. Noise removal and Edge preservation filter is used to remove noise from input dataset. Features are extracted that are useful for reconstruction of image that have more information than any of the input image. Subjective and objective analysis is performed to verify the significance of proposed technique compared to already existing algorithms and results demonstrate that the produce technique outperform existing algorithm.