获取目标姿态的图像匹配方法研究
|
摘要
在研发工业机械手引导系统中,提出了一种基于轮廓特征匹配对工件姿态进行定位的方法。首先对模板和待测图像利用金字塔向下采样法剔除图像细节特征并确定塔层,再以图像关键外轮廓点集为匹配特征,依据模板自动计算旋转步长,采用相似度量法则进行匹配判断,引入贪婪值控制匹配速度和精度,最后采用最大值法和均值法筛选消除重配和误配得到图像坐标系下的姿态。然后对单目相机进行标定获得相机参数,利用最小二乘法确立图像坐标系与世界坐标系的关系,实现目标在世界坐标系中准确定位。实验测试表明:对目标遮挡、背景复杂、对比度低等鲁棒性强,匹配率高,可实现任意坐标、角度情况下的目标识别和定位,毫秒级耗时,可以满足引导系统要求。
引文
[1]张银银.图像匹配技术的算法研究[D].华中科技大学,2010.
[2]Lowe D G.Distinctive image features from scale-invariant keypoints[J].International Journal of Computer Vision,2004,60(2):91-110.
[3]Kumar Y.Automatic Misalignment Defects Detection Correction in PCB using SURF(Speed up Robust Features)Technique[J].International Journal of Engineering&Technical Research,2017,V6(1).
[4]Talmi I,Mechrez R,Zelnikmanor L.Template Matching with Deformable Diversity Similarity[J].2017:1311-1319.
[5]Buniatyan D,Macrina T,Ih D,et al.Deep Learning Improves Template Matching by Normalized Cross Correlation[J].2017.
[6]Hofhauser A,Steger C,Navab N.Harmonic Deformation Model for Edge Based Template Matching[A].Visapp 2008:Proceedings of the Third International Conference on Computer Vision Theory and Applications,Funchal,Madeira,Portugal,January.DBLP[C].2008:75-82.
[7]Tuytelaars T,Mikolajczyk K.Local Invariant Feature Detectors[J].Foundations&Trends?in Computer Graphics&Vision,2007,3(3):177-280.
[8]王魁生,刘蒙.浅谈图像仿射变换的应用[J].信息技术与信息化,2015(1):155-156.
[9]倪健,白瑞林,李英,等.采用轮廓向量特征的嵌入式图像匹配方法[J].计算机工程与应用,2014,50(13):168-172.
[10]李莉.OpenCV耦合改进张正友算法的相机标定算法[J].轻工机械.2015,(4):60-63.
[11]赵双,杨慕升,等.结合小波变换的Shi-Tomasi算法遮挡图像匹配研究[J].机械设计与制造,2017(11):118-121.
[2]Lowe D G.Distinctive image features from scale-invariant keypoints[J].International Journal of Computer Vision,2004,60(2):91-110.
[3]Kumar Y.Automatic Misalignment Defects Detection Correction in PCB using SURF(Speed up Robust Features)Technique[J].International Journal of Engineering&Technical Research,2017,V6(1).
[4]Talmi I,Mechrez R,Zelnikmanor L.Template Matching with Deformable Diversity Similarity[J].2017:1311-1319.
[5]Buniatyan D,Macrina T,Ih D,et al.Deep Learning Improves Template Matching by Normalized Cross Correlation[J].2017.
[6]Hofhauser A,Steger C,Navab N.Harmonic Deformation Model for Edge Based Template Matching[A].Visapp 2008:Proceedings of the Third International Conference on Computer Vision Theory and Applications,Funchal,Madeira,Portugal,January.DBLP[C].2008:75-82.
[7]Tuytelaars T,Mikolajczyk K.Local Invariant Feature Detectors[J].Foundations&Trends?in Computer Graphics&Vision,2007,3(3):177-280.
[8]王魁生,刘蒙.浅谈图像仿射变换的应用[J].信息技术与信息化,2015(1):155-156.
[9]倪健,白瑞林,李英,等.采用轮廓向量特征的嵌入式图像匹配方法[J].计算机工程与应用,2014,50(13):168-172.
[10]李莉.OpenCV耦合改进张正友算法的相机标定算法[J].轻工机械.2015,(4):60-63.
[11]赵双,杨慕升,等.结合小波变换的Shi-Tomasi算法遮挡图像匹配研究[J].机械设计与制造,2017(11):118-121.