| EPSRC Reference: |
GR/H52962/01 |
| Title: |
INTERDISCIPLINARY STUDIES IN OBJECT RECOGNITION |
| Principal Investigator: |
Harris, Professor J |
| Other Investigators: |
|
| Researcher Co-Investigators: |
|
| Project Partners: |
|
| Department: |
Computer Science |
| Organisation: |
University of Reading |
| Scheme: |
Standard Research (Pre-FEC) |
| Starts: |
01 June 1993 |
Ends: |
30 June 1996 |
Value (£): |
124,337
|
| EPSRC Research Topic Classifications: |
| Vision & Senses - ICT appl. |
|
|
| EPSRC Industrial Sector Classifications: |
|
| Related Grants: |
|
| Panel History: |
|
|
Summary on Grant Application Form |
|
(i) Establish an interdisciplinary test facility for experiments in visual object recognition.(ii) Investigate the 'geon' theory of object recognition by humans.(iii) Appraise the use of 'geons' in object recognition by machines.Progress:Despite national advertisements, no suitable staff could be employed at this salary. Permission was consequently obtained from SERC to reduce the duration of the RA post to 30 months, at a higher salary level.Following a literature review, research concentrated on two issues: (i) the relative effectiveness of intensity images and line-drawings, as cues in the recognition of the geonic structure of objects, and (ii) the pre-attentive nature of geonic cuing. An investigation of mechanisms of early vision as revealed by pop-out phenomena was adopted. Software was developed to run psychological experiments on UNIX workstations under X-windows, to provide best forward compatibility with machine vision developments. There were some difficulties with real-time issues, but the system now runs well. Initial experiments to verify reported pop-out demonstrations were run under three conditions: grey-scale, line-drawings, and line+grey images. The results suggest that intensity images are processed faster than line-drawings and demonstrate parallel pop-out after fewer practice trials. This work is in preparation for journal publication.Our observations contradict Biederman's account of the derivation of geon structure (through an intermediate stage of line-drawing), and suggest that geon extraction might use surface characteristics directly, rather than infer surfaces indirectly from edge information. This goes against current practice in machine vision, though the implications for the latter have yet to be explored.
|
|
Key Findings |
|
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
|
|
Potential use in non-academic contexts |
|
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
|
|
Impacts |
|
| Description |
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk |
| Summary |
|
| Date Materialised |
|
|
|
|
Sectors submitted by the Researcher |
|
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
|
| Project URL: |
|
| Further Information: |
|
| Organisation Website: |
http://www.rdg.ac.uk |