一种稳健的低照度岩心样本图像拼接算法的实现  

The realization of a robust stitching algorithm for low illuminated core sample images

作　　者：胡明[1] 王正勇[1] 卿粼波[1] 刘跃[1] 

机构地区：[1]四川大学电子信息学院,成都610065

出　　处：《四川大学学报（自然科学版）》2015年第3期547-556,共10页Journal of Sichuan University(Natural Science Edition)

基　　金：国家自然科学基金项目(61372174)

摘　　要：以往的基于SIFT-BBF-RANSAC算法对岩心图像进行拼接的过程在白光岩心扫描工程应用中取得较好的效果,但当面对低照度、特征不明显的岩心荧光图像、以及受消光性和干涉色影响的岩石薄片正交偏光图像时,传统的BBF算法无法很好地粗剔除误匹配特征点对,RANSAC算法就无法实现图像的正确拼接.本文提出了一种改进算法,首先根据实际应用中待匹配岩心荧光、岩石薄片正交偏光图像的特点对图像进行模糊增强,然后对图像利用SIFT算法提取特征点,最后基于改进的BBF-RANSAC剔除误匹配算法,实现了低照度、重合量很小的岩心荧光图像以及岩石薄片正交偏光图像拼接.The past core images stitching algorithm based on SIFT-BBF-RANSAC obtained good results in the white light core scan engineering applications, but when faced with low illumination core images with less obvious characteristics, such as fluorescence core images, or orthogonal polarization core ima- ges affected by extinction and interference, the traditional BBF algorithm cannot eliminate false matches well, which results in the wrong stitching result by RANSAC algorithm. The paper proposed an im- proved algorithm. First make the images enhanced by using fuzzy enhancement algorithm according to the characters of fluorescence core images or orthogonal polarization core images, and then to extract the feature points by using SIFT algorithm, and last to realize the stitching of fluorescence core images which are low illumination and low coincide parts or orthogonal polarization core images based on the im- proved BBF - RANSAC eliminate false matches algorithm.

关 键 词：岩心样本图像拼接 图像增强 剔除误匹配 

分 类 号：TP315[自动化与计算机技术—计算机软件与理论]