基于SURF改进算法的高分辨率岩心图像拼接  

Core Image Splicing with High Resolution Based on Improved SURF Algorithm

作　　者：顾宫 沈疆海[1] 

机构地区：[1]长江大学计算机科学学院,湖北荆州434023

出　　处：《长江大学学报（自然科学版）》2018年第9期46-50,共5页Journal of Yangtze University(Natural Science Edition)

摘　　要：由于相机自身视野的原因,在岩心图像扫描过程中,一幅岩心图像往往无法被完整扫描,使得对岩心图像分析增加了难度。为了解决这个问题,每行选择距离边界30个像素点来对左右2张图像进行局部特征点提取,提高了匹配速度。基于加速稳健特征(Speeded Up Robust Features,SURF)算法对岩心图像进行特征提取,利用随机抽样一致(Random Sample Consensus,RANSAC)算法对岩心图像特征点去噪。结合欧氏距离与特征向量筛选最优特征点,从而实现岩心图像高精度拼接,得到完整的岩心图像。该拼接方法减少了人为因素导致的错误,确保了地质资料的客观性和准确性,便于对岩心图像的专业分析与应用,具有良好的拼接效果和应用前景。Because of the camera's own visual field,a core image can't be scanned completely in the core image scanning process,which makes it difficult for core image analysis.In order to solve this problem,in this paper,30 pixels are selected from each line to extract the local feature points from the left and right images to improve the matching speed.Based on the SURF(Speeded Up Robust Features)algorithm,the core image is extracted.And the RANSAC(Random Sample Consensus)algorithm is used to denoise the feature points of the core image.In addition,combined with the Euclidean distance and the feature vector,the best feature points are screened,the high precision splicing of core images can be achieved,and a complete core image is obtained.This technique reduces the error caused by human factors,ensures the objectivity and accuracy of geological data,and facilitates the professional analysis and application of core images.The splicing method has good effect and application prospects.

关 键 词：岩心 SURF 特征匹配 图像拼接 特征点方向 

分 类 号：TP751.1[自动化与计算机技术—检测技术与自动化装置]