Robust and Efficient 3D Recognition by Alignment

Abstract

Alignment is a prevalent approach for recognizing three-dimensional objects in two-dimensional images. Current implementations handle errors that are inherent in images in ad hoc ways. This thesis shows that these errors can propagate and magnify through the alignment computations, such that the ad hoc approaches may not work. In addition, a technique is given for tightly bounding the propagated error, which can be used to make the recognition robust while still being efficient. Further, the error bounds can be used to formally compute the likelihood that a set of hypothesized matches between model and image features is correct. The technique for bounding the propagated error makes use of a new solution to a fundamental problem in computer recognition, namely, the solution for 3D pose from three corresponding points under weak-perspective projection. The new solution is intended to provide a fast means of computing the error bounds. In deriving the new solution, this thesis gives a geometrical interpretation to the problem, from which the situations are inferred where the solution does not exist and is unstable.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Sep 01, 1992
Accession Number
ADA270839

Entities

People

  • Tao D. Alter

Organizations

  • Massachusetts Institute of Technology

Tags

Communities of Interest

  • Air Platforms
  • C4I
  • Materials and Manufacturing Processes

DTIC Thesaurus Topics

  • Algorithms
  • Artificial Intelligence
  • Computations
  • Computer Vision
  • Computers
  • Data Sets
  • Detectors
  • Geometry
  • Image Recognition
  • Information Processing
  • Machine Perception
  • Object Recognition
  • Probability
  • Recognition
  • Three Dimensional
  • Two Dimensional

Readers

  • Applied Combinatorial Optimization and Logic Circuit Design.
  • Calculus or Mathematical Analysis
  • Computer Vision.

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference
  • AI & ML - Machine Learning Algorithms