Research On Automatic Stake Test System Based On Vision Of Autonomous Mobile Robot

Automatic Mobile Robot (AMR) is the most popular question for research in robots domain in recent years, it has great prosperous application in intelligent human-machine interactive systems, service robots, space robots and unmanned driving systems. In order to improve the equality and accuracy of Automatic Stake Test System in Driving License Examination, this thesis does deep research in application of AMR to Automatic Stake Test System based on machine vision theory and pattern recognition theory. Main work of this thesis is as follows:(1)According to the character of Automatic Stake Test System, this thesis applies Finite States Machine (FSM) theory to the module design of decision in AMR. By abstracting the whole procedure of Automatic Stake Test System, imitating the states and events which real person could meet during the Automatic Stake Test, using state diagram to denote states and events, minimizing them and deleting redundancy states contained by other states, the thesis finally gain the FSM of Automatic Stake Test System. Using FSM not only makes design course more easy and understandable but also accelerates design cycle.(2)AMR judges the states change through vision and external events gained by other sensors, makes transfer among diversified states and attains the purpose of using robots to simulate manual work in Automatic Stake Test.(3) A moving-target detective identification and automatic tracing control system for AMR is designed. Considering identification & control processing, this system fractionizes assignments received by AMR, achieves prompt surveillance and real time trace control strategy to objective in the structured & unstructured; familiar & unfamiliar environment, improves real time character and accuracy of robot.This system applies in Automatic Stake Test System of Driving License Examination, satisfies the real time character and reliability demands of Automatic Stake Test System .