Multi-focus image fusion techniques: a survey

Multi-Focus Image Fusion (MFIF) is a method that combines two or more source images to obtain a single image which is focused, has improved quality and more information than the source images. Due to limited Depth-of-Field of the imagining system, extracting all the useful information from a single image is challenging. Thus two or more defocused source images are fused together to obtain a composite image. This paper provides a comprehensive overview of existing MFIF methods. A new classification scheme is developed for categorizing the existing MFIF methods. These methods are classified into four major categories: spatial domain, transform domain, deep leaning and their hybrids and have been discussed well along with their drawbacks and challenges. In addition to this, both the parametric evaluation metrics i.e. "with reference" and "without reference" have also discussed. Then, a comparative analysis for nine image fusion methods is performed based on 30 pairs of publicly available images. Finally, various challenges that remain unaddressed and future work is also discussed in this work.