Abstract:Recognizing the same place undervariant viewpoint differences is the fundamental capability for human beings and animals. However, such a strong place recognition ability in robotics is still an unsolved problem. Extracting local invariant descriptors from the same place under various viewpoint differences is difficult. This article seeks to provide robots with a human-like place recognition ability using a new 3-D feature learning method. This article proposes a novel lightweight 3-D place recognition and fast sequence matching to achieve robust 3-D place recognition, capable of recognizing places from a previous trajectory regardless of viewpoints and temporary observation differences. Specifically, we extracted the viewpoint-invariant place feature from 2-D spherical perspectives by leveraging spherical harmonics' orientation-equivalent property. To improve sequence-matching efficiency, we designed a coarse-to-fine fast sequence-matching mechanism to balance the matching efficiency and accuracy. Despite the apparent simplicity, our proposed approach outperforms the relative state of the art. In both public and self-gathered datasets with orientation/translation differences or noise observations, our method can achieve above 95% average recall for the best match with only 18% inference time of PointNet-based place recognition methods.

Viewpoint Space Partitioning Based on Affine Invariant Features

The Novel 3D/2D Constrain in Affine Projection

A Solution to Efficient Viewpoint Space Partition in 3D Object Recognition

An Improved Algorithm of Viewpoint Space Partition in 3d Object Recognition Application

Invariant object recognition by shape space analysis

Affine Invariant Feature Detection Based on Scale Space Analysis and Depth Information

Full-viewpoint 3D Space Object Recognition Based on Kernel Locality Preserving Projections

Affine-similar shape retrieval: application to multiview 3-D object recognition

Fast Object Recognition Based on Curvature Affine Invariant

A completely affine invariant image-matching method based on perspective projection

3D Face Recognition Using Affine Integral Invariants

Shape Features for Projected 2D Images Based 3D Object Recognition

A stable viewing strategy for rotation normalization free 3D model retrieval

Object recognition based on three-dimensional model

Invariant Descriptor for Multi-View 3D Facial Surface

Planar-Projective Summation Invariant Features For Camera Networks

Fast Sequence-Matching Enhanced Viewpoint-Invariant 3-D Place Recognition.

Invariants of the Space Point Element Structure and Their Applications

An affine invariant of parallelograms and its application to camera calibration and 3d reconstruction

A fuzzy associative approach for recognition of 3D objects in arbitrary pose

Multi-view 3-D Target Recognition Algorithm Based on Moment Invariant and Support Vector Machine Classification