Experimental Research on Signal Light Display Recognition System for Urban Railway Transportation
Accurate recognition for urban rail transit signal light display status is an important safeguard measure to ensure the safe operation of urban rail transit system.Recognition of signal status in existing urban rail transit relies on visual observation by train operators.The combination of subjective and objective factors,such as bad weather and driver fatigue,can lead to safety accidents due to signal status recognition problems.In this paper,a set of machine vision based on the principle of effectively recognizing 5～200 m trackside signal light display status recognition system has been designed and constructed.A trackside signal recognition method based on image processing technology is realized on this system.The field experiments combined with multi-scene virtual detection experimental results show that the method can effectively recognize the display state of trackside signals of urban rail transit.The recognition result accuracy can reach up to 95.4%,which can meet the requirements of urban rail transit on the performance and safety of signal status recognition.
NETL
NSTL
万方数据
