Range image registration based on photometry 光学解析に基づく距離画像の位置合わせ
Range image registration based on photometry
3D modeling of a real scene stands for constructing a virtual representationof the scene, generally simplified that can be used or modified atour will. Constructing such a 3D model by hand is a laborious and timeconsuming task, and automating the whole process has attracted growing interestin the computer vision field. In particular, the task of registering (i.e.aligning) different parts of the scene (called range images) acquired fromdifferent viewpoints is of crucial importance when constructing 3D models.During the last decades, researchers have concentrated their efforts on thisproblem and proposed several methodologies to automatically register rangeimages. Thereby, key-point detectors and descriptors have been utilized tomatch points across different range images using geometric features or texturalfeatures. Several similarity metrics have also been proposed to identifythe overlapping regions. In spite of the advantages of the current methods,several limitation cases have been reported. In particular, when the scenelacks in discriminative geometric features, the difficulty of accounting forthe changes in appearance of the scene observed in different poses, or fromdifferent viewpoints, significantly degrades the performance of the currentmethods. We address this issue by investigating the use of photometry (i.e.the relationship between geometry, reflectance properties and illumination)for range image registration. First, we propose a robust descriptor usingalbedo that is permissive to errors in the illumination estimation. Second,we propose an albedo extraction technique for specular surfaces that enlargesthe range of materials we can deal with. Third, we propose a photometricmetric under unknown lighting that allows registration of range images withoutany assumptions on the illumination. With these proposed methods, wesignificantly enlarge the practicability and range of applications of rangeimage registration.