---
title: "3-D Inference From the Motion Parallax of a Conic Arc and a Point in Two Perspective Views"
authors:
  - "Roger Y. Tsai"
doi: ""
url: "https://ijcai.org/proceedings/1983/240"
pdf: "https://ijcai.org/proceedings/1983/0240.pdf"
venue: "Proceedings of the Eighth International Joint Conference on Artificial Intelligence"
year: 1983
pages: "1038-1041"
track: "Vision"
keywords:
---

# 3-D Inference From the Motion Parallax of a Conic Arc and a Point in Two Perspective Views

Roger Y. Tsai

*Proceedings of the Eighth International Joint Conference on Artificial Intelligence*, pages 1038-1041.

## Abstract

We present an efficient method for 3-D inference from the motio n parallax of an edge in the form of a conic arc and a coplonar point feature in two perspective views. The edge on the object can be a boundary line along which the image intensity has high contrast, or object surface discontinuity , while the point feature could be corner point , vertex point or any point with distinguishing feature. We prove a theorem regarding the uniqueness of 3-D motion parameters and object surface structure determined from the projective correspondences of a conic arc and a point. The theorem gives rise to an efficient method that requires very few arithmetic operations to compute the solutions. We prove a corollary of the theorem which states that given the motion parallax of a conic arc without knowning the correspondences of the two end points, and two coplanar points, the solution is unique. Since the correspondences of the two end points are not used, the method works even if part of the conic edge could be occluded in either view. The theoretical findings have been supported by computer simulation tests.