A robust approach to obtain a guidance directrix for a vision-based agricultural vehicle guidance system

A machine vision algorithm is presented to obtain a guidance directrix for an automatic vehicle guidance system. The vision algorithm is based on a binary transform by adaptive threshold, acnode filtering and an improved Hough transform. The outputs were the equations of the lines representing the rows. The acnode filtering was used to eliminate image noise. The least squares method and the Hough transform are the common methods to check the lines. However, the least square method was sensitive to noise points and the traditional Hough transform is very slow. To resolve this problem, an improved Hough transform is proposed. The algorithm was successfully implemented in a vision-based automatic guidance system for an experimental car. Real-time processing experiments showed that the algorithm met real-time processing requirements.