Object Detection in Autonomous Vehicles

Object detection in self-driving cars is a prime component because there is no human control over the cars’ driving. Self-driving cars must detect an object 360 degrees around the car for the car to steer or handle the driving on the correct track. The market for autonomous vehicles is rapidly growing, and as all people now know about the level of technology in the developed countries, which are the United States, the United Kingdom, Germany, England, and Japan, those countries are going all out in the creation of electric and autonomous vehicles. However, most of the current object detection methods are expensive and ineffective. Therefore, in this paper, the new, efficient, and faster design of object detection in autonomous vehicles using the YOLOv5 neural network is investigated, explored, designed, implemented, and evaluated. YOLOv5 is the best model because it is built on the Python framework, is available in a variety of sizes, and is constantly updated and improved. The Python platform, Roboflow tool, and Google Colab were adopted to design the model. Roboflow is great for object detection because of smart polygons, label assist, commenting, and exporting in a variety of formats. Meanwhile, Google Colab is free to use (allows you to run Python code in the cloud), easy to use (user-friendly interface), and offers access to powerful GPUs. The adoption of these with the YOLOv5 neural network is a new attempt in autonomous vehicles. As a result, a better and more efficient object detection method yet affordable for autonomous vehicles was produced and is presented in this paper.