In order to improve the user's travel convenience,especially for the visual impaired person,a multi-mode interactive intelli-gent terminal is designed based on semantic analysis and machine vision technology.The whole system is implemented in hardware,mainly with the Raspberry Pi as the core processor,together with a speech processing acquisition board,camera,speaker and GPS posi-tioning module.The software is designed by means of modular plug-ins,and makes comprehensive use of web crawler technology,speech recognition and synthesis technology,semantic understanding,intelligent dialogue,weather news broadcast,obstacle detection,route planning and so on.The system has a complete front-end and back-end design,with the front-end providing Web access,chat in-formation viewing,real-time obstacle viewing and map viewing,and the back-end providing back-end management.The result shows that the system has good stability,accuracy and practicality,which is of strong practical significance to serve people's daily travel.
deep learningtarget recognitionspeech processingintelligent terminals
NETL
NSTL
万方数据
