On-chip ego-motion estimation based on optical flow

Abstract

A novel system on chip (SoC) is presented for estimating the ego-motion of the vehicle based on optical flow cues. The main novelty of the system consists in its implementation as an on-chip hybrid hardware/software system. The improvements in the FPGA technology allows to have programmable logic resources accompanied by processors on the same chip. In this way, inherently sequential tasks are implemented as software modules executed in embedded general purpose processors while hardware friendly modules are implemented in specific purpose co-processing engines. The proposed SoC is capable of estimating ego-motion by using the PowerPC available in the FPGA XC4VFX60-10 of Xilinx. © 2011 Springer-Verlag.