Combined Use of Modified Hough Transformation, Random Sample Consensus and Linear Least Square to Extract the Normal Parameterization of a Straight Line: An Application for Cable Driven Parallel Robots

摘要

The current approach to determine the platform position of the Cable Driven Parallel Robot (CDPR) is by using camera system or forward kinematics. However both has their drawbacks. Alternatively, a laser scanner is introduced. Three popular algorithms to extract the straight line is applied to estimate two among three component of platform position vector. Then, the proposed method is applied on the physical prototype. Two estimated coordinates of the position vector are compared with the desired trajectory in order to evaluate the platform position accuracy. The comparison shows that the platform has deviation from the desired position.