基于改进 ORB 的低照度岩心荧光图像拼接技术

摘要

The traditional ORB algorithm for image stitching has good effect .But for core image of petroleum geology exploration ,especially low illuminated core fluorescence image ,existing ORB algorithms has higher rate of mistake matc‐hing ,resulting in not valid for subsequent image stitching .For this problem ,this paper analyzes the shortcomings of the o‐riginal algorithm ORB ,and improves the original algorithm ORB by using FREAK description operators what based on the human visual system ,and it is named IORB(improved ORB) algorithm .Experimental results show that improved algorithm for image stitching fluorescence gets the ideal correct matching rate in the guaranteed time on the basis of the cost and en ‐hances the effect of the subsequent stitching process of stitching .%传统的 ORB 算法对于特征信息丰富的图像拼接已经有了不错的效果，但是对于石油地质勘探中的岩心图像，特别是低照度岩心荧光图像，现有 ORB 算法误匹配率较高，从而导致不能有效地进行后续的图像拼接。针对该问题，论文分析了原 ORB 算法的缺陷，采用基于人眼视觉系统的 FREAK 描述算子改进了原 ORB 算法，并命名为 IORB（Improved ORB）算法。实验结果表明，改进算法对于荧光图像的拼接在保证时间代价的基础上得到了理想的正确匹配率，并提升了后续拼接过程的拼接效果。