Automatic neutral section passing control device based on image recognition for electric locomotives

Conventionally, a neutral section is operated manually by drivers according to break/close markers along railways, when an electric locomotive is passing the section. Because of the increase in trains' speed, neutral section passing control devices with magnets on ground and onboard detection coils are used in some countries. This paper presents a new automatic onboard neutral section passing control device based on image recognition. It is designed to reduce the costs of installation and maintenance compared with the existing devices. An industrial smart camera is used as a kernel processor, which has high reliability and a high processing speed. The new device can recognize the break/close markers along a railway by detection of image edges. The break/close time is optimized by calculation of image size of break/close markers. With this device not only the distance between the train and break/close markers can be determined but also the train's speed can be calculated.