Implementing a Real-Time Hough Transform on a Mobile Robot

Robotic vision is a challenging problem due to the uncertain nature of real-world environments and the computational constraints of mobile platforms. Many computer vision algorithms have high computational requirements and are often seen as unsuitable for use in an embedded system. RoboCup, an international robotic soccer competition, presents a dynamic and mobile environment in which color based runlength encoding algorithms struggle. Many current approaches use static and fragile color lookup tables and heavily subsampled images to achieve an optimal processing rate of 30 frames per second. The aim of this project is to implement the Hough Transform for real-time line finding on the Aldebaran Nao robot. Using mostly color and lighting independent pixel gradient information, we detect soccer field lines. Using assembly language instructions to expose the processor’s full potential, this project implements a realtime Hough transform based object detection system on 320x240 pixel images.