Multifocus image fusion based on robust principal component analysis

• A multifocus image fusion scheme via robust principal component analysis is presented.
• Sparse features are computed to describe salient information within the images.
• The scheme is flexible to integrate different fusion strategies in the sparse domain.
• The method is simple, robust and able to effectively handle grayscale and color images.
• The new fusion method outperforms the state-of -the-art image fusion approaches.

Multifocus image fusion has emerged as a major topic in computer vision and image processing community since the optical lenses for most widely used imaging devices, such as auto-focus cameras, have a limiting focus range. Only objects at one particular depth will be truly in focus while out-of-focus objects will become blurry. The ability to create a single image where all scene areas appear sharp is desired not only in digital photography but also in various vision-related applications. We propose a novel image fusion scheme for combining two or multiple images with different focus points to generate an all-in-focus image. We formulate the problem of fusing multifocus images as choosing the most significant features from a sparse matrix obtained by a newly developed robust principal component analysis (RPCA) decomposition method to form a composite feature space. The local sparse features that represent salient information of the input images (i.e. sharp regions) are integrated to construct the resulting fused image. Experimental results have demonstrated that it is consistently superior to the other existing state-of-the-art fusion methods in terms of visual and quantitative evaluations.