Accession Number:

ADA599485

Title:

Leader/Follower Behaviour Using the SIFT Algorithm for Object Recognition

Descriptive Note:

Technical memo.

Corporate Author:

DEFENCE RESEARCH AND DEVELOPMENT SUFFIELD (ALBERTA)

Personal Author(s):

• Giesbrecht, J.

Report Date:

2006-06-01

Pagination or Media Count:

32.0

Abstract:

This paper presents an application of vision-based object recognition to create a leaderfollower behaviour in mobile robots. A system is developed which makes use of the Scale Invariant Feature Transform SIFT algorithm to recognize a leader robot or human. The follower robot then uses PID control to track the leaders movements while maintaining a fixed following distance.

Descriptors:

• *BEHAVIOR
• *LEADERSHIP
• *RECOGNITION
• ALGORITHMS
• MOBILE
• NAVIGATION
• ROBOTICS

Subject Categories:

• Administration and Management
• Psychology
• Cybernetics

Distribution Statement:

APPROVED FOR PUBLIC RELEASE