High Speed Image Registration Algorithm Based on ORB and Corner Angle Constraint
Shicong ZENG,Tailiang GUO,Zhixian LIN
DOI: https://doi.org/10.16280/j.videoe.2015.09.018
2015-01-01
Abstract:For the image registration is difficult under the condition of the different deformation,illumination and others,the registration algorithm based on scale invariant features transfonn(SIFT)is complex,and it doesu0027 t meet the real-time requirement.So,a high speed registration algorithm is proposed based on Oriented FAST and Rotated Brief(ORB) and corner angle constraint.Firstly,ORB feature points are extracted in two partitioned images,then the feature points are matched with double thresholdu0027s hamming distance,for random sample consensus(RANSAC) canu0027t reject the mismatched feature points.Corner angle constraint is proposed to reject the mismatched feature points,and then calculate the optimal transfonn matrix with RANSAC algorithm to finish the image registration.Experimental results show that,with three groups of images with different deformation and others,the algorithm can rectify the image,and its average time to rectify the three groups of images is 80.399 ms,less than 1/20 of the SIFTs average time,it matches both the validity and time.