under construction ......
Image-based Relocalization is A Classification Problem
•Similar problems:
•Image Recognition (objects);
•Place Recognition (views);
•Scene Classification (categories);
•Image Indexing and Retrieval/Search;
•Image Near Duplicate Detection (NDD).
•Specialty in image-based relocalization: (loop closure)
•May use depth information (RGB-D);
•3-D model/camera pose information (Multiple View Stereo);
•Key frame chosen carefully (Structure from Motion or SLAM).
Figure 1. Example of 3d-to-2d relocalization
Image-Based Localization Pipeline
•Key Frame and its Camera Pose Stored;
•Frame id, image feature and camera pose.
•Option: Extract Local Features;
•Feature original or coded.
•Establish 2D-2D or 3D-to-2D Matches;
•Single camera: 3d-to-2d;
•Stereo or depth camera: 2d-to-2d;
•Camera Pose Estimation.
•Refine by 3-d point estimation;
•2-d features along with their depths.
Figure 2. Camera Pose Estimation
Relocalization Approaches
•Direct image matching:
•Gaussian blurred and downsampled templates for key frame matching;
•Feature matching-based:
•Corner (Harris, FAST, Good) + Descriptor (SIFT, SURF, BRIEF, FREAK, ORB, Daisy, …);
•BovW-based image matching:
•Build codebook or vocabulary, then map features to BovW;
•Learning the mapping from the features or BoVWs to camera pose:
•Camera pose estimation is formulated as a regression problem;
•Feature learning for keyframe matching:
•Hierarchical feature representation by deep learning;
•Metric learning for nonlinear distance optimization;
•Manifold learning for optimized dimension reduction of feature space.
Figure 3. An example of RGB-D image relocalization
Figure 4. Kinect fusion for scene modeling (in need of image-based relocalization)
Feature Matching
•Descriptor Extraction;
•Dimension reduction: PCA;
•Nearest Neighbor Search:
•Brute force: K-d tree or radius search;
•ANN;
•Outlier Removal:
•Ratio test;
•RANSAC for geometric consistency check
•Homography;
•Fundamental matrix;
SURF Feature Matching-based Relocalization
•Feature detection and descriptor for key frames or new frames which require relocalization:
•Harris corner;
•SURF (64-d vector);
•Feature matching for pose recovery:
•Brute force search;
•Pose difference is small enough.
•Key frame selection:
•Pose change enough big;
Figure 5. SURF feature based matching for relocalization
BoVW-based Relocalization
•Codebook building from training data:
•Feature extraction: FAST corner detection + SURF feature descriptor;
•Feature clustering for visual vocabulary: incremental K-means;
•Construct co-occurrence statistics of visual words by Chow Liu tree.
•A naïve Bayes approximation by a maximum weight spanning tree.
•Visual pattern from key frames or new frames for relocalization:
•Feature extraction: same as above;
•Mapping descriptors to bags of words by ANN-based (kd-tree) methods;
•Pattern matching for pose recovery: descriptor mapped to codebook;
•A sampling method in training data to find the best match.
Figure 6. A section Chow Liu tree computed to generate visual vocabulary.