Research Of Smart Wheelchair Control System Based On Facial Expression

For the most disabled people, smart wheelchair, as a sort of auxiliary means, plays animportant role in their daily life. But for the people who are seriously disabled or withparalysis, even the electric wheelchairs are hard or impossible to apply; in order to make thesepeople “walk” by themselves, oversea and domestic researchers are dedicated to the researchesof relying on the body control rather than trational wheelchair. Considering the hightechnological content and price of most smart wheelchairs in the market which is also withgreat autonomy and interactivity, this topic designs a kind of smart wheelchair control systemwith low price and easy operation based on facial expression and activity in accordance withthe cahracteristics of smart wheelchair.The structure of the system is divided into three parts: Information collection, informationanalysis, information output and device control. The content of information collection andinformation analysis is facial expression recognition; The content of the information outputand device control is design of intelligent wheelchair drive system.Facial expression definition and recognition solutions have been put forward to controlwheelchairs. Firstly, image preprocessingfor the image to be detected shall be done, includinggrey level transformation, median filtering, image enhancement, histogram equalization, etc.;secondly, face detection shall be done after the image preprocessing. In this regard, aharrisdetecting algorithm based on a2D geometric features shall be put forward with astepwise refinement mark thought of “coarse positioning--fine matching--feature pointsautomatic tagging”. Facial features will be coarse positioned at first, and then the features ofbrows, eyes, pupil, nostril and mouth will be fine matched based on the haar-like features, andall the features will be automatically labeled at the same time. A fast recognition solution offacial expression definition based on the geometry relationship of facial features has beenpresented, thus realizing the recognition of facial expressions, establishing a communicationmechanism based on facial expression and drive circuit, developing an interactive interface.The smart wheelchair drive system adopts DSP chip TMS320F2812as the main controlcore, design hardware,software and monitoring systems.