Automatic Passenger Counting System Using Image Processing Based on Skin Colour Detection Approach

Public transportation plays a big role in transportation industry which provides mobility and creates better transport planning for example, electric train and buses. Latterly, the topic that researchers are always looking for is automatic passenger counting on public transport, for example in bus application. The number of people entering and leaving the bus is an important parameter to allocate the correct number of buses for each public transport service connection line. As far as the safety and comfort of the passengers is considered, the bus through which they are travelling should not be overloaded. Moreover, the distribution of existing buses on different routes is the basis for obtaining overall transport network optimization, and to reduce costs. In general, most of the available passenger counting system consists of sensor and counter to count the number of passengers on the public transport. However, the problem faced by this system is inaccurate counting due to fault in the transmission of data from sensor to counter. Therefore, this paper proposes an automatic passenger counting system for fleet management by Using image processing Based on skin colour detection approach. This study aims to increase the efficiency and reliability of bus management system in Universiti Malaysia Perlis (UniMAP). In order to obtain the number of passenger in the bus, the passenger images have been processed with several image processing techniques which consist of a few steps; (1) colour conversion of RGB to HSV colour image, (2) image segmentation Using thresholding technique, (3) removal of noise or unwanted object, (4) image smoothing and (5) passenger counting process. Overall, experimental results show that the proposed method has successfully segmented and perform the counting of 25 passenger images with average counting accuracy of 90.64%.