FPGA-Based Real-Time Embedded Vision System for Autonomous Mobile Robots

Abstract

In this paper, we developed an embedded vision system for autonomous mobile robots. This system enable to help a robot to measure in real-time the distance of an obstacle that we have defined in advance and avoid it. The vision system uses a single front camera as sources of information on the surrounding environment and an FPGA as reconfigurable and flexible device for image processing. The whole system was implemented only in VHDL a parallel programming language on a Nexys 4 DDR development board with Artix FPGA device from Xilinx that operates at 100 MHz. This allowed obtaining a real-time image processing for a 640x480 @ 30 fps video stream.