Detecção e modelagem de rampas para órteses ativas usando uma câmera RGB-D

Abstract

: This paper presents a ramp detection and modeling algorithm applied in a active orthosis. This method uses depth sensing capacities of the RGB-D cameras to segment different detected planes that integrate the scene and then performs a classification that compares the slope and width of the detected planes with acessibility especifications shown in technical standards. Simulations were made using four different ramps and it proves that the system is able to appropriately classify these structures. Resumo de inclina¸c˜ao e largura distintas. Foi poss´ıvel constatar que a estrat´egia empregada foi capaz de classificar os objetos adequadamente.