Stream processing of moment invariants for real-time classifiers

C. H. Messom, A. L C Barczak

Research output: Chapter in Book/Report/Conference proceedingConference PaperResearchpeer-review


This paper introduces a general purpose graphicsprocessing unit (GPGPU) stream processing implementation ofmoment invariants using an integral image or summed area tableapproach. Summed area tables have been used to help attainreal-time performance for some classifier systems, however dueto the computational complexity of moment invariants, a highthroughput computational platform is required to obtain real-time processing. The stream programming algorithm ispresented and its performance is evaluated and compared withalternate CPU based approaches. The significant performancegains means that moment invariant classifiers can beimplemented for real-time performance on a GPGPU that wouldnot be possible on current CPU platforms.

Original languageEnglish
Title of host publicationICARA 2009 - Proceedings of the 4th International Conference on Autonomous Robots and Agents
PublisherIEEE, Institute of Electrical and Electronics Engineers
Number of pages6
ISBN (Print)9781424427130
Publication statusPublished - 2009
Externally publishedYes
EventInternational Conference on Autonomous Robots and Agents 2009 - Massey University, Wellington, New Zealand
Duration: 10 Feb 200912 Feb 2009
Conference number: 4th


ConferenceInternational Conference on Autonomous Robots and Agents 2009
Abbreviated titleICARA 2009
Country/TerritoryNew Zealand

Cite this