Implementing Fuzzy Logic in Planar Motion Tracking

Khoo Boon How, Alex See Kok Bin

School of Engineering Monash University Malaysia,

No. 2, Jalan Kolej, Bandar Sunway, 46150, Petaling Jaya, Selangor Darul Ehsan, Malaysia.

Abstract:

Robotics applications require high dexterity and dynamic response. Dynamic of a moving object is related to unknown motion path. The motion of the target will be varying with respect to time. It is difficult to describe dynamic motion with a single equation. This paper proposes a fuzzy logic prediction system improve the performance of motion tracking system. The dynamic models of a moving target are replaced with set of fuzzy inference rules. A Fuzzy Prediction System (FPS) is introduced. This Fuzzy Logic Controller comprises of two layers in a cascaded form. The first layer is responsible for coarse step prediction. The second layer is responsible for fine step correction. This proposed Fuzzy Logic controller was tested with computer simulation. It was implemented and evaluated by a real time imaging system utilizing an inexpensive webcam. The proposed system was tested with an object moving at about 400 mm/sec. Correlation between the measured path and predicted path has been investigated. From this study, it was found that that system is capable of predicting a target motion path; a correlation result of approximately 0.97 was achieved.

Keywords: Fuzzy Logic, Motion Tracking, Motion Prediction, Vision System.

(Full Paper, 672KB)

BibTex:

@ARTICLE{P1120548005,

AUTHOR = {Khoo Boon How, Alex See Kok Bin},

TITLE = {Implementing Fuzzy Logic in Planar Motion Tracking},

}

(Full Paper, 672KB)