Effective Utilization of A Low-Cost Solution for Remote Sensing of Vehicles and Pedestrians

The field of remote sensing does not only confine to planetary and earth sciences but also in every field of science and technology. The hassled traffic patterns lead to fatalities and accidents and an efficient system comprising of detecting and analyzing the remotely sensed traffic components is of paramount essentiality. This study focuses at utilizing a downsized hardware along with deep learning algorithm for real-time detection of vehicles and pedestrians as per the Indian context of traffic and automotive markets. The system remotely senses the objects through computer vision and carries out the detection with high precision (0.98) and high frame rate (74 fps) necessary for intelligent transportation applications. The study provides a strong basis for utilization of a hardware system for assisting Advanced Driver Assistance Systems (ADAS) as per the Indian context of traffic and laying foundation for autonomous driving cars.