期刊名称：Signal & Image Processing : An International Journal (SIPIJ)
印刷版ISSN：2229-3922
电子版ISSN：0976-710X
出版年度：2014
卷号：5
期号：5
页码：13
出版社：Academy & Industry Research Collaboration Center (AIRCC)
摘要：In this paper, we present feature-based technique for construction of mosaic image from underwater videosequence, which suffers from parallax distortion due to propagation properties of light in the underwaterenvironment. The most of the available mosaic tools and underwater image mosaicing techniques yieldsfinal result with some artifacts such as blurring, ghosting and seam due to presence of parallax in the inputimages. The removal of parallax from input images may not reduce its effects instead it must be correctedin successive steps of mosaicing. Thus, our approach minimizes the parallax effects by adopting an efficientlocal alignment technique after global registration. We extract texture features using Centre SymmetricLocal Binary Pattern (CS-LBP) descriptor in order to find feature correspondences, which are used furtherfor estimation of homography through RANSAC. In order to increase the accuracy of global registration,we perform preprocessing such as colour alignment between two selected frames based on colourdistribution adjustment. Because of existence of 100% overlap in consecutive frames of underwater video,we select frames with minimum overlap based on mutual offset in order to reduce the computation costduring mosaicing. Our approach minimizes the parallax effects considerably in final mosaic constructedusing our own underwater video sequences.
