抄録
For the purpose of advancing Intelligent Transportation Systems (ITS), sensors capable of gathering precise and various traffic informations are needed. This paper describes a spatial image processing traffic flow sensor we have developed, which can measure wide range traffic flow informations, that is, volume, speed, vehicle type, vehicle presence and traffic congestion length simultaneously as well as detecting illegally parked vehicles and traffic abnormal conditions.
This sensor compresses images inputted from a camera as a JPEG format and then transmits to a traffic control center, where latest images or stored time-serial images before and after an incident detected by this sensor can be seen.
Recently, we have applied this sensor to both a traffic signal control and an image surveillance system, in which an ultrasonic vehicle detector is beneath its capabilities. And we have also constructed a collision avoidance warning system in a tunnel by providing directly most useful queue length measurement results to a variable message board.
