Title: An automatic and robust algorithm for determining motion and structure from two perspective images
Abstract: A fully automatic and robust system is described in this paper for determining motion and structure from two perspective images. Classically, the problem is usually divided into two steps: (i) extract and match features between images; (ii) determine motion and structure from matched features. However, because only heuristics are used in the first step, several matches are not correctly established. These false matches, sometimes even one, may perturb completely the motion estimation process, making the result useless. In this paper, we use a very robust technique, namely the least-median-squares, to detect false matches (theoretically up to 50% of the whole data) in order to estimate correctly the motion and structure. The idea underlying this technique is to search for an estimate in the motion space with which most of matches satisfy the well-known coplanarity constraint. The proposed algorithm is completely automatic. It has been tested with a variety of real data: indoor, outdoor, rocks, etc. Good results have been obtained.
Publication Year: 1995
Publication Date: 1995-01-01
Language: en
Type: book-chapter
Indexed In: ['crossref']
Access and Citation
Cited By Count: 9
AI Researcher Chatbot
Get quick answers to your questions about the article from our AI researcher chatbot