Position Uncertainty Reduction ofMobileRobot Based on DINDsinIntelligent Space
Tae-Seok Jin, Primož Podržaj, Hideki Hashimoto · 2006
Thispaperproposes a localization ofmobile robotusingtheimagesbydistributed intelligent networked devices (DINDs)inintelligent space(ISpace). Thisscheme combines datafromtheobserved position using dead-reckoning sensors andtheestimated position usingimagesofmoving object, suchasthose ofawalking human,usedtodetermine the movinglocation ofa mobilerobot. Themovingobject is assumed tobeapoint-object andprojected ontoanimageplane toformageometrical constraint equation thatprovides position dataoftheobject basedonthekinematics oftheintelligent space. Usingthea-priori knownpathofamovingobject anda perspective cameramodel, thegeometric constraint equations thatrepresent therelation between imageframecoordinates of a movingobject andtheestimated position oftherobotare derived. Theproposed approach isapplied foramobile robot in ISpace toshowthereduction ofuncertainty inthedetermining ofthelocation ofthemobile robot. IndexTerms-Mobile robot; Tracking; Intelligent space; Distributed sensors; Uncertainty.