RECURSIVE FILTER BASED DETAIL ENHANCED MULTI-EXPOSURE IMAGE FUSION
Keywords:
Image fusion, Multi exposure fusion, HDR image, recursive filterAbstract
The Multi-scale exposure fusion technique is an efficient image enhancement for a high dynamic range (HDR) scene. Several fusion algorithms have been introduced in recent years to combine significant data from various input images into a single fused image. In view of the problems of irregular exposure in the image acquisition and the severe loss of particulars in the conventional multi-exposure image fusion algorithms. The basic idea of the exposure fusion is to retain details in both quite dark and relatively bright regions without representation of HDR image and tone mapping step. The proposed algorithm better preserves the details in the most bright and dark regions of the HDR scene. This paper proposes a weighted technique to multi- exposure image fusion to better retain details, considering features like local contrast, exposure brightness and color information. To eliminate the noise and interference, using recursive filter. Moreover, the proposed method is quite fast and it can be used directly for most consumer cameras. Experimental results shows that the proposed method is superior in terms of subjective and objective analysis.