Towards Manipulation-Driven Vision
Abstract
For the purposes of manipulation, the authors would like to know what parts of the environment are physically coherent ensembles, that is, which parts will move together, and which parts are more or less independent. It takes a great deal of experience before this judgement can be made from purely visual information. This paper develops active strategies for acquiring that experience through experimental manipulation, using tight correlations between arm motion and optic flow to detect both the art itself and the boundaries of objects with which it comes into contact. The number of papers written on techniques for visual segmentation is vast. Methods for characterizing the shape of an object through tactile information also are being developed, such as shape from probing or pushing. But while it has long been known that motor strategies can aid vision, work on active vision has focused almost exclusively on moving cameras. There is much to be said about bringing a manipulator into the equation, as the authors have shown in this paper. Many variants and extensions to the experimental "poking" strategy explored here are possible.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 2001
- Accession Number
- ADA434691
Entities
People
- Giorgio Metta
- Paul M. Fitzpatrick
Organizations
- Massachusetts Institute of Technology