A Wearable System Designed for Chinese Traffic Police Based on Gesture Recognition

Abstract

The body language and gestures of the traffic police can be difficult to identify accurately, especially in China, where there is often fog and haze. Poor judgment and slow recognition will reduce the efficiency of traffic management, and can even cause traffic accidents. To solve this problem, a wearable system has been designed for the traffic police. This wearable system consists of two main parts: a pair of data gloves and a vest with a light-emitting diode (LED) display. The data gloves were utilized to recognize and record the gestures of the traffic police. The gesture signals were converted into traffic instructions and displayed on the vest. The effectiveness of this wearable system was verified through a comparison experiment between only gesture and gesture plus displaying text on the vest. To achieve best effects of recognition, according to the results of further experiment, the simple instructions (stop, move straight, turn left waiting, turn right, and change lane) should be displayed in the form of an image, whereas the complex instructions (turn left, wait, slow down, and pull over) should be displayed in text form.