Multi-Range Images Integration Algorithm Based on Normal Constraint
Juming Cao · Xi'an Jiaotong Daxue xuebao · 2009
Aiming at the problem of multi-view range images integration in the process of optical based 3D measurement and 3D modeling,a novel multi-view range images integration algorithm is proposed based on the normal constraint principle.The normal direction filtering is applied for each range images.The seed points are searched between two related range images.The corresponding k-nearest neighborhoods in both images are searched too.The weighted sum of each k-nearest point to seed points is calculated.The output points are the seeds which have been moved along the normal direction.Compared with the k-mean clustering multi-view range images integration algorithm,a uniform points distributions surface can be obtained which has a smoother and detailer feature,especially for the case with coarse registration error.