About

Name: Semantic ICP

Summary: Novel point cloud registration algorithm that directly incorporates pixelated semantic measurements into the estimation of the relative transformation between two point clouds.

Description: The algorithm uses an Iterative Closest Point (ICP)-like scheme and performs joint semantic and geometric inference using the Expectation-Maximization technique in which semantic labels and point associations between two point clouds are treated as latent random variables. The minimization of the expected cost on the three-dimensional special Euclidean group, i.e., SE(3), yields the rigid body transformation between two point clouds.

License: BSD 3-Clause

Requirements

libraries

• Eigen
• PCL
• Ceres-Solver
• Sophus
• gtest
• Benchmark

Install

$ mkdir build
$ cd build
$ cmake ../
$ make

Maintainers

Steven Parkison ([email protected])

Paper

@INPROCEEDINGS { sparkison-2018a,
  AUTHOR = { Steven A Parkison and Lu Gan and Maani Ghaffari Jadidi and Ryan
    M. Eustice },
  TITLE = { Semantic Iterative Closest Point through
    Expectation-Maximization },
  BOOKTITLE = { Proceedings of the British Machine Vision Conference},
  YEAR = { 2018 },
  MONTH = { September },
  ADDRESS = { Newcastle, UK },
  PAGES = { 1--17 },
}