| With the acceleration of aging in China,the disabled elderly population has increased rapidly.However,the number of professional nurses cannot meet the market demand.In the elderly care at home,the difficulty of dining is one of the urgent problems to be solved,so the dining robot came into being.However,at this stage,the intelligent dining robot still has many problems such as inconvenient use,high price,etc.,so it has not been popularized.Therefore,it is necessary for the development of intelligent food aid equipment to improve the use mode and interaction form of intelligent food aid robot through technical means and provide a new design scheme.In this thesis,a new interactive way based on computer vision to identify the user’s head posture and mouth state to control the catering robot is proposed,and a product prototype is designed and developed.The recognition of head posture is used to control the rotation of the dinner plate to select dishes;The recognition of mouth state is used to judge the eating intention and drive the mechanical arm to realize meal aid.Simple body movements have low requirements on the user’s physical ability,and can enable some upper limb disabled people to eat independently.The research contents are as follows:(1)The development context and research status of catering products are summarized.(2)The characteristics of user groups are analyzed,and a typical user model is built.(3)The data set is built.The LeNet-5 model structure is improved to realize real-time attitude recognition.(4)The design scheme of the intelligent catering robot is designed and the prototype is built for verification.The research results show that the interaction mode based on computer vision proposed in this thesis is feasible,and it can drive the catering robot to complete the catering action.Through the integration of disciplines,design and technology,a new scheme is provided for the development of intelligent catering robot. |
PDF Full Text Request