摘要
为了解决传统的图像处理算法识别现场获得的工件图像速度慢且匹配效果较差等问题,通过对工件图像的识别方法进行研究,提出了一种改进的加速鲁棒特征(SURF)算法可以实现工件准确、实时的定位。该算法基于加速分割测试特征检测器(FAST)对SURF算法的特征提取方式进行改进,首先利用FAST提取特征点,然后通过SURF算法生成特征点描述子,使用主成分分析算法(PCA)对描述子进行降维。随后以欧式距离作为相似性度量进行粗匹配,再采用随机抽样一致算法(RANSAC)剔除误匹配点。最后结合双目视觉技术得到工件空间位置坐标。实验结果表明:本文提出的算法在运行时间上相比传统SURF算法减少80%,同时提高了匹配的精度。达到了准确、实时的工件定位目的。
In order to solve the problem that the traditional image processing algorithm recognizes that the workpiece image obtained on the spot is slow and the matching effect is poor,the recognition method of the workpiece image was studied,and an improved speeded up robust features( SURF) algorithm was proposed to realize accurate and real-time positioning of the workpiece. The algorithm improved the feature extraction method of SURF algorithm based on features from accelerated segment test detector( FAST). Firstly,feature points were extracted by FAST,then feature point descriptors were generated by SURF algorithm,and the principal component analysis algorithm( PCA) was used to reduce the dimension of the descriptor. Then the euclidean distance was used as the similarity measure for rough matching,and then the random sampling agreement algorithm( RANSAC) was used to eliminate the mismatched points. Finally,combined with binocular vision technology,the workpiece spatial position coordinates were obtained. The experimental results show that the proposed algorithm reduces the running time by 80%compared with the traditional SURF algorithm,and improves the matching accuracy,achieving accurate and realtime workpiece positioning.
引文
1蒋萌,王尧尧,陈柏.基于双目视觉的目标识别与定位研究[J].机电工程,2018,35(4):414-419Jiang Meng,Wang Yaoyao,Chen Bo. Recognition and orientation of object based on binocular vision[J]. Mechanical and Electrical Engineering,2018,35(4):414-419
2 周海英,王守义,杨阳.改进的核相关自适应目标跟踪算法及其实验验证[J].科学技术与工程,2018,18(14):190-196Zhou Haiying,Wang Shouyi,Yang Yang. Improved adaptive target tracking algorithm based on kernel correlation and its experimental verific[J]. Science Technology and Engineering,2018,18(14):190-196
3 Jaafari I E,Ansari M E,Koutti L. Fast edge-based stereo matching approach for road applications[J]. Signal Image&Video Processing,2016(11):1-8
4 Deng Z,Cao M,Rai L,et al. A two-stage classification method for borehole-wall images with support vector machine[J]. Plos One,2018,13(6):e0199749
5 Harris C,Stephens M. A combined corner and edge detector[C]//The Proceedings of 4th Alvey Vision Conference. United Kingdom:Plessey Research Roke Manor,1988:147-151
6 Lowe D G. Distinctive image features from scale-invariant keypoints[J]. International Journal of Computer Vision,2004,60(2):91-110
7 Bay H,Ess A,Tuytelaars T,et al. Speeded-up robust features[J].Computer Vision&Image Understanding,2008,110(3):404-417
8 Trujillo N,Chapuis R,Chausse F,et al. On road simultaneous vehicle recognition and localization by model based focused vision[C]//Iapr Conference on Machine Vision Applications. Tsukuba:DBLP,2009:388-391
9 Zhou J,Mao Y,Cheng T,et al. Research on routing algorithm for on-line monitoring of leakage current of insulators[C]//IEEE,Advanced Information Technology,Electronic and Automation Control Conference. New York:IEEE,2017:1822-1826
10 Wang Y Q,Liang B,Zhuang L L. Applied technology in unstructured road detection with road environment based on SIFT-HARRIS[J]. Advanced Materials Research,2014,1014:259-262
11 Rosten E,Porter R,Drummond T. Faster and better:A machine learning approach to corner detection[J]. IEEE Transactions on Pattern Analysis&Machine Intelligence,2008,32(1):105-119
12 张珊.基于PCA-SURF的单目视觉SLAM算法[D].青岛:中国海洋大学,2014Zhang Shan. Monocular vision SLAM algorithm based on PCASURF[D]. Qingdao:Ocean University of China,2014