An Extensive Survey on IoT Architecture and Machine Learning Algorithms for Crowd Detection in Public Transportation

Internet of Things links real-world objects to virtual world. Large amount of devices are connected to the internet and generates large amount of data. The data collected is to be controlled and converted to information for developing an efficient system. In this paper we focus on transportation system. Excessive crowds in buses often leads to loss of precious life. So, a survey is proposed using embedded system for automatic crowd monitoring in buses to avoid unnecessary accidents due to overcrowding. For counting people in the bus, the survey contains sensors such as Infrared sensor [IR], horizontal beam sensor, passenger counter sensor, video processing sensor, people counter. The gateways in this survey are PIC Microcontroller, Arduino, Raspberry Pi. The communication protocols are GSM, Bluetooth, ZigBee and Wi-Fi. The system is mainly designed to avoid accidents due to overcrowding.