Multi-view registration is a fundamental but challenging problem in 3Dreconstruction and robot vision. Although the original motion averagingalgorithm has been introduced as an effective means to solve the multi-viewregistration problem, it does not consider the reliability and accuracy of eachrelative motion. Accordingly, this paper proposes a novel motion averagingalgorithm for multi-view registration. Firstly, it utilizes the pair-wiseregistration algorithm to estimate the relative motion and overlappingpercentage of each scan pair with a certain degree of overlap. With theoverlapping percentage available, it views the overlapping percentage as thecorresponding weight of each scan pair and proposes the weight motion averagingalgorithm, which can pay more attention to reliable and accurate relativemotions. By treating each relative motion distinctively, more accurateregistration can be achieved by applying the weighted motion averaging tomulti-view range scans. Experimental results demonstrate the superiority of ourproposed approach compared with the state-of-the-art methods in terms ofaccuracy, robustness and efficiency.
展开▼
