A novel wide field image mosaicing approach for the underwater pipelines

作者：Sheng, Mingwei; Zhang, Qiang; Zhang, Wen

来源：ICIC Express Letters, Part B: Applications , 2015, 6(7): 1851-1858.

• Contrast Enhancement
• Equalization algorithms
• Feature extraction methods
• Fusion algorithms
• Homography estimations
• Information amount
• Weighted averages
• Wide field

In order to meet the demands of constructing the wide field seabed image for the underwater pipelines exploration based on machine vision, a novel underwater image fusion and mosaicing algorithm was presented. Firstly, RGB automatic color equalization algorithm was proposed for the contrast enhancement of color underwater pipeline images. Secondly, precise matching points feature extraction methods were introduced to identify and remove the outliers, and the pixel-based registration was carried out with the homography estimation. Finally, a weighted average fusion method was proposed to reduce visible artifacts, that is, to hide the splice line of the component images. The proposed framework has been tested with underwater pipelines image sequences. The experimental results indicated that the algorithms can retain the integrality and continuity of the original underwater images, while keeping the information amount of the fused image.

• 出版日期2015
• 单位哈尔滨工程大学