基于IBBS-SIFT算法的PCB板Mark点定位方法
|
摘要
针对最佳伙伴相似性(Best-buddies similarity, BBS)算法和SIFT算法的不足,提出了一种基于IBBS-SIFT(Improved best-buddies similarity-scale invariant feature transform)算法的印刷电路板Mark点定位方法。首先对BBS算法进行改进,通过结合置信度图和权值,在采集的PCB板图像中得到与模板图像相似的多个区域;然后采用最佳相似点对匹配准则代替SIFT算法的欧氏距离匹配准则,用改进的SIFT算法对这些相似区域计算描述子并进行匹配;最后根据匹配的描述子得到主位置,通过主位置剔除误匹配点,精确定位到Mark点位置。实验结果表明:该方法可以有效提高Mark点定位效率,将耗时减至SIFT算法的10%～15%,同时减少误匹配点数目,实现Mark点精确定位。
Aiming at the shortcomings of best-buddies similarity(BBS) algorithm and SIFT algorithm, a Mark point positioning method of printed circuit board based on IBBS-SIFT(Improved best-buddies similarity scale-invariant feature transform) algorithm was proposed. Firstly, the algorithm of best-buddies similarity(BBS) was improved. By combining confidence map and weight, multiple regions similar to the template image were obtained in the PCB board image collected. Then, the Euclidean distance matching criterion of SIFT algorithm was replaced by the matching criterion of optimal similarity point pair, and the descriptors of these similar regions were calculated and matched by the improved SIFT algorithm. Finally, the main position was obtained according to the matched descriptor, and the mismatched points were removed by the main position to accurately locate the Mark point. Experimental results show that the algorithm proposed in this paper can effectively improve the efficiency of Mark point positioning, reduce the time consumption to 10%～15% of SIFT algorithm, and decrease the number of mismatched points and achieve accurate Mark point positioning.
Aiming at the shortcomings of best-buddies similarity(BBS) algorithm and SIFT algorithm, a Mark point positioning method of printed circuit board based on IBBS-SIFT(Improved best-buddies similarity scale-invariant feature transform) algorithm was proposed. Firstly, the algorithm of best-buddies similarity(BBS) was improved. By combining confidence map and weight, multiple regions similar to the template image were obtained in the PCB board image collected. Then, the Euclidean distance matching criterion of SIFT algorithm was replaced by the matching criterion of optimal similarity point pair, and the descriptors of these similar regions were calculated and matched by the improved SIFT algorithm. Finally, the main position was obtained according to the matched descriptor, and the mismatched points were removed by the main position to accurately locate the Mark point. Experimental results show that the algorithm proposed in this paper can effectively improve the efficiency of Mark point positioning, reduce the time consumption to 10%～15% of SIFT algorithm, and decrease the number of mismatched points and achieve accurate Mark point positioning.
引文
[1] Quyang W,Tombari F,Mattoccia S,et al.Performance evaluation of full search equivalent pattern matching algorithms[J].IEEE Transactions on Pattern Analysis & Machine Intelligence,2011,34 (1):127-143.
[2] 周欣,魏海坤,张侃健,等.PCB板Mark点的定位算法研究[J].工业控制计算机,2018,31(4):111-112.
[3] 陈小艳,王强,李柏林.改进的 Hough 变换检测圆方法[J].计算机系统应用,2015,24(8):197-201.
[4] 陈若珠,孙岳.基于最小二乘法的椭圆拟合改进算法研究[J].工业仪表与自动化装置,2017(2):35-38.
[5] 王万国,王仕荣,徐正飞.基于边界的最小二乘椭圆拟合改进算法[J].计算机技术与发展,2013(4):67-70.
[6] 熊风光,李希,韩燮.基于整体最小二乘的椭圆拟合方法[J].微电子学与计算机,2017,34(1):102-105.
[7] 刘政,熊昊,肖森林,等.基于PCA和分段RHT的PCB板圆Mark点定位[J].重庆理工大学学报(自然科学),2017,31(1):93-99.
[8] Lowe D G.Distinctive image features from scale invariant key points[J].International Journal of Computer Vision,2004,60(2):581-588.
[9] Ke Y,Sukthankar R.PCA-SIFT:A more distinctive representation for local image descriptors[C]//IEEE Compute Society Conference Vision and Pattern Recognition.IEEE Computer Society,2004:506-513.
[10] 冯文斌,刘宝华.改进的SIFT算法图像匹配研究[J].计算机工程与应用,2018,54(3):200-205.
[11] Dekel T,Oron S,Rubinstein M,et al.Best-buddies similarity for robust template matching[C]//IEEE Conference on Computer Vision and Pattern Recognition.IEEE Computer Society,2015:2021-2019.
[12] 王刚,孙晓亮,尚洋,等.一种基于最佳相似点对的稳健模板匹配算法[J].光学学报,2017,37 (3):274-280.
[13] 刘佳,傅卫平,王雯,等.基于改进SIFT算法的图像匹配[J].仪器仪表学报,2013,34(5):1107-1112.
[14] 吴丽君,文吉成,陈志聪,等.改进SIFT算法实现图像的快速匹配[J].福州大学学报(自然科学版),2017,45(6):801-809.
[15] 完文韬,杨成禹.改进的SIFT算法在图像特征点匹配中的应用[J].长春理工大学学报(自然科学版),2018,41(1):44-47.
[16] 贺一楠,耿娟,秦军,等.结合Forstner与NCC约束的UAV图像配准研究[J].国土资源遥感,2015,27(1):48-54.
[17] 孙卜郊,周东华.基于NCC的快速匹配算法[J].传感器与微系统,2007(9):104-106.
[18] 曹一溪,王斌.基于颜色直方图匹配的半透明材质估计[J].计算机辅助设计与图形学学报,2015,27(3):441-450.
[2] 周欣,魏海坤,张侃健,等.PCB板Mark点的定位算法研究[J].工业控制计算机,2018,31(4):111-112.
[3] 陈小艳,王强,李柏林.改进的 Hough 变换检测圆方法[J].计算机系统应用,2015,24(8):197-201.
[4] 陈若珠,孙岳.基于最小二乘法的椭圆拟合改进算法研究[J].工业仪表与自动化装置,2017(2):35-38.
[5] 王万国,王仕荣,徐正飞.基于边界的最小二乘椭圆拟合改进算法[J].计算机技术与发展,2013(4):67-70.
[6] 熊风光,李希,韩燮.基于整体最小二乘的椭圆拟合方法[J].微电子学与计算机,2017,34(1):102-105.
[7] 刘政,熊昊,肖森林,等.基于PCA和分段RHT的PCB板圆Mark点定位[J].重庆理工大学学报(自然科学),2017,31(1):93-99.
[8] Lowe D G.Distinctive image features from scale invariant key points[J].International Journal of Computer Vision,2004,60(2):581-588.
[9] Ke Y,Sukthankar R.PCA-SIFT:A more distinctive representation for local image descriptors[C]//IEEE Compute Society Conference Vision and Pattern Recognition.IEEE Computer Society,2004:506-513.
[10] 冯文斌,刘宝华.改进的SIFT算法图像匹配研究[J].计算机工程与应用,2018,54(3):200-205.
[11] Dekel T,Oron S,Rubinstein M,et al.Best-buddies similarity for robust template matching[C]//IEEE Conference on Computer Vision and Pattern Recognition.IEEE Computer Society,2015:2021-2019.
[12] 王刚,孙晓亮,尚洋,等.一种基于最佳相似点对的稳健模板匹配算法[J].光学学报,2017,37 (3):274-280.
[13] 刘佳,傅卫平,王雯,等.基于改进SIFT算法的图像匹配[J].仪器仪表学报,2013,34(5):1107-1112.
[14] 吴丽君,文吉成,陈志聪,等.改进SIFT算法实现图像的快速匹配[J].福州大学学报(自然科学版),2017,45(6):801-809.
[15] 完文韬,杨成禹.改进的SIFT算法在图像特征点匹配中的应用[J].长春理工大学学报(自然科学版),2018,41(1):44-47.
[16] 贺一楠,耿娟,秦军,等.结合Forstner与NCC约束的UAV图像配准研究[J].国土资源遥感,2015,27(1):48-54.
[17] 孙卜郊,周东华.基于NCC的快速匹配算法[J].传感器与微系统,2007(9):104-106.
[18] 曹一溪,王斌.基于颜色直方图匹配的半透明材质估计[J].计算机辅助设计与图形学学报,2015,27(3):441-450.