Object Detection Framework Using OpenCV for Low Cost and High Performance

In contemporary times, advancements in technology have led to a renewed interest in the development of autonomous vehicles, commonly referred to as self-driving cars. These vehicles are equipped with sensors and computer vision tools that enable them to detect and respond to their environment, reducing the need for human intervention in the navigation process. Given the transformative potential of this technology, there has been a growing emphasis on researching and developing autonomous vehicles as a way of revolutionizing the transportation industry. The goal of this research work is to code highly optimized plus parameterized Python algorithms that would drive the object detection systems like Canny edge, pedestrian, vehicle, etc., with the help of pre-existing object models with greater emphasis on reducing the cost-performance trade-off. Various object models like EfficientNet, YOLOv4, R-CNN, etc. were deployed and tested for creating comparison in terms of ease of deployment, efficiency, accuracy, computational, and hardware costs to filter out the most efficient model for the coded computer vision tools. These innovative systems will possess the capability to detect lane markings, accurately identify pedestrians and other vehicles, and continually monitor the vehicle's immediate surroundings, thereby providing real-time information to the autonomous driving system through highly customizable parameters that can be changed according to the needs. By default, the codes in the research work will be customized for Indian roads, which are known for their unique challenges, such as lack of lane markings, heavy traffic, and poor road conditions. However, challenges remain, including legal and regulatory issues, code robustness, hardware integration, public perception, acceptance, and ensuring the safety and security of the autonomous driving system. Despite these challenges, the potential of autonomous vehicles to revolutionize transportation is significant.