Addressing the non-functional requirements of computer vision systems: a case study

Abstract

Computer vision plays a major role in most autonomous systems and is particularly fundamental within the robotics industry, where vision data are the main input to all navigation and high-level decision making. Although there is significant research into developing and optimising algorithms for feature detection and environment reconstruction, there is a comparative lack of emphasis on how best to map these abstract concepts onto an appropriate software architecture. In this study, we distinguish between functional and non-functional requirements of a computer vision system. Using a RoboCup humanoid robot system as a case study, we propose and develop a software architecture that fulfills the latter criteria. To demonstrate the modifiability of the proposed architecture, we detail a number of examples of feature detection algorithms that were modified to capture the rapidly evolving RoboCup requirements, with emphasis on which aspects of the underlying framework required modification to support their integration. To demonstrate portability, we port our vision system (designed for an application-specific DARwIn-OP humanoid robot) to a general-purpose, Raspberry Pi computer. We evaluate the processing time on both hardware platforms for several image streams under different conditions and compare relative to a vision system optimised for functional requirements only. The architecture and implementation presented in this study provide a highly generalisable framework for computer vision system design that is of particular benefit in research and development, competition and other environments in which rapid system evolution is necessary to adapt to domain-specific requirements.