Multimodal Segmentation of Dense Depth Maps and Associated Color Information
نویسندگان
چکیده
An integrated segmentation approach for color images and depth maps is proposed. The 3D pointclouds are characterized by normal vectors and then grouped into planar, concave or convex faces. The empty regions in the depth map are filled by segments of the associated color image. In the experimental part two types of depth maps are analysed: generated by the MS-Kinect sensor or by a stereo-pair of cameras.
منابع مشابه
Dense Structure-from-Motion: An Approach Based on Segment Matching
For 3-D video applications, dense depth maps are required. We present a segment-based structure-from-motion technique. After image segmentation, we estimate the motion of each segment. With knowledge of the camera motion, this can be translated into depth. The optimal depth is found by minimizing a suitable error norm, which can handle occlusions as well. This method combines the advantages of ...
متن کاملColor Segmentation Based Depth Image Filtering
We present a novel enhancement method that addresses the problem of corrupted edge information in depth maps. Corrupted depth information manifests itself in zigzag edges instead of straight ones. We extract the depth information from an associated color stream and use this information to enhance the original depth map. Besides the visual results, a quantitative analysis is conducted to prove t...
متن کامل3 D Recovery Using Calibrated Active
In this contribution we describe a system to generate dense depth maps from image sequences recorded from a camera mounted on a robot. Two color images of the sequence are treated as a stereo pair. A color edge detector is applied, lines are segmented from the edge images, and approximations as sequences of straight lines and circular arcs are computed. An initial depth map results from lines o...
متن کاملSemantic Segmentation of Urban Scenes Using Dense Depth Maps
In this paper we present a framework for semantic scene parsing and object recognition based on dense depth maps. Five viewindependent 3D features that vary with object class are extracted from dense depth maps at a superpixel level for training a classifier using randomized decision forest technique. Our formulation integrates multiple features in a Markov Random Field (MRF) framework to segme...
متن کاملDense Depth Maps by Active Color Illumination and Image Pyramids
An efficient hierarchical technique is presented for obtaining dense stereo correspondence using a color image pyramid. We show that matching results of high quality are obtained when using the hierarchical chromatic Block Matching algorithm. Furthermore, we present some results for parallel implementations of the algorithm. Most stereo matching algorithms can not compute correct dense depth ma...
متن کامل