Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

An Efficient Method for Accurate Pose Estimation and Error Correction of Cuboidal Objects

Utsav Rai, Hardik Mehta, Vismay Vakharia, Aditya Choudhary, Amit Parmar, Rolif Lima, Kaushik Das

TL;DR

An efficient method for precise pose estimation of cuboid-shaped objects, which aims to reduce errors in target pose in a time-efficient manner, is presented.

Abstract

The proposed system outlined in this paper is a solution to a use case that requires the autonomous picking of cuboidal objects from an organized or unorganized pile with high precision. This paper presents an efficient method for precise pose estimation of cuboid-shaped objects, which aims to reduce errors in target pose in a time-efficient manner. Typical pose estimation methods like global point cloud registrations are prone to minor pose errors for which local registration algorithms are generally used to improve pose accuracy. However, due to the execution time overhead and uncertainty in the error of the final achieved pose, an alternate, linear time approach is proposed for pose error estimation and correction. This paper presents an overview of the solution followed by a detailed description of individual modules of the proposed algorithm.

An Efficient Method for Accurate Pose Estimation and Error Correction of Cuboidal Objects

TL;DR

An efficient method for precise pose estimation of cuboid-shaped objects, which aims to reduce errors in target pose in a time-efficient manner, is presented.

Abstract

The proposed system outlined in this paper is a solution to a use case that requires the autonomous picking of cuboidal objects from an organized or unorganized pile with high precision. This paper presents an efficient method for precise pose estimation of cuboid-shaped objects, which aims to reduce errors in target pose in a time-efficient manner. Typical pose estimation methods like global point cloud registrations are prone to minor pose errors for which local registration algorithms are generally used to improve pose accuracy. However, due to the execution time overhead and uncertainty in the error of the final achieved pose, an alternate, linear time approach is proposed for pose error estimation and correction. This paper presents an overview of the solution followed by a detailed description of individual modules of the proposed algorithm.
Paper Structure (11 sections, 6 equations, 4 figures, 1 table)

This paper contains 11 sections, 6 equations, 4 figures, 1 table.

Table of Contents

  1. Introduction
  2. Related Work
  3. Proposed Approach
  4. Sensor Data Acquisition
  5. Sensor data processing
  6. Segmentation
  7. ROI Filtration
  8. Pose Estimation
  9. Pose Error Estimation and Correction
  10. Results
  11. Conclusion

Figures (4)

  • Figure 1: (a) RGB image of the scene aligned to the point cloud, (b) ROI thresholding to reduce search space, (c) T1, T2 and corner point selection and extraction of corresponding 3D point from point cloud using convexity defect and inverse projection, (d) Point cloud segmentation using non zero pixels from (b) and fitting oriented bounding box to cross verify dimensions of ROI
  • Figure 2: Proposed algorithm in action: Accurate Pose estimation of cuboidal object and motion planning during Mohamed bin Zayed International Robotics Challenge (MBZIRC) 2020 MBZIRC-2020
  • Figure 3: (a) Rotational error of $3.26^{\circ}$ between artificial point cloud (red) and target point cloud (gray). (b) Transitional error of 0.19cm between artificial point cloud (red) and target point cloud (gray) after compensating rotational error.
  • Figure 4: (a) Artificial point cloud (red) is registered on target point cloud (white) with rotational error of 2.47° in yaw and translation error of 0.8 mm in x, 3.1 mm in y and -0.2 mm in z (b) Rotational error compensated (c) Followed by translation error correction and giving final accurate pose with 0.23° accuracy in rotation and 0.3mm in transition (Euclidean distance from centroid)