一种用于图像拼接的改进BRISK算法  

Improved BRISK algorithm used in image mosaic

作　　者：陈思聪[1,2] 刘晶红[1] 何林阳 周前飞[1,2] 

机构地区：[1]中国科学院长春光学精密机械与物理研究所航空光学成像与测量重点实验室,吉林长春130033 [2]中国科学院大学,北京100049

出　　处：《液晶与显示》2016年第3期324-330,共7页Chinese Journal of Liquid Crystals and Displays

基　　金：吉林省重大科技攻关项目(No.11ZDGG001);装备预研项目(No.402040203);国家林业公益性行业科研专项(No.201204515)~~

摘　　要：为了获得精准的航空拼接图像,本文提出了一种结合BRISK算法与互相关模块匹配算法的新算法。传统的BRISK算法在拼接平移方向上存在较大的误差。针对该问题,首先使用BRISK算法实现尺度和旋转的校正,再引入模块匹配方法完成平移校正,同时在BRISK算法中加入RANSANC算法实现精准拼接。实验结果表明,本文算法是一种运算时间短、精确度高、拼接效果良好的图像拼接方法。In order to obtain accurate aerial stitching images,a new algorithm combined with Binary Robust Invariant Scalable Keypoints（BRISK）and cross-correlation template matching algorithm is proposed in this thesis.There are some errors in the splicing translational direction on using traditional BRISK algorithm in image mosaic.Aiming at the problem,the BRISK algorithm is firstly used for the scale and rotation correction of the matching images.Then the template matching method is used for the translation correction. Meanwhile,the random sample consensus（RANSANC）algorithm is added to the BRISK algorithm to achieve precise mosaic.The experiment results show that the proposed algorithm has short operation time,high precision,and can achieve good image mosaic results.

关 键 词：图像拼接 BRISK特征 RANSANC算法 模板匹配 

分 类 号：TP722[自动化与计算机技术—检测技术与自动化装置]