Vehicle Detection and Count in the Captured Stream Video Using Machine Learning

Abstract

The technology of vehicle detection in the captured video has implementation in the variety of fields. This emerging technology when implemented over the real-time video feeds can be beneficial for the missions of search and rescue in remote areas, where access is hampered by mountains and vast land areas without road networks. Areas afflicted by natural disaster (earthquake, flood) may be aided and improved by autonomous UAV systems, in military applications and civilians due to its ability to operate over large, difficult terrain. Rescue or surveillance missions include people and vehicle detection from aerial platforms. The supreme benefit of vehicle detection in the real-time streaming video feed is to track the terrorist intrusion over the borders and hit them. The recent terrorist attack occurred in the Indian soil at Uri, Pulwama, which could be avoided by using the enhanced vehicle detection techniques implied in the live stream video by machine learning. This chapter presents to recognize and identify the moving objects in the live video stream for specific interest. This chapter aims to explore the existing challenging issue in the area of unsupervised surveillance and security. The detection of vehicles is implemented with enhanced algorithms and machine learning libraries like OpenCV, TensorFlow, and others. The various approaches are used to identify and track the specific object through the trained model from the captured video.