Automated Sign Detection in a Self-Driving Car

Self-driving cars are a recent trending area of research in the automation industry. The advantages of using self-driving cars are certainly on efficient resource utilization along with reduction in manual effort and danger to manual labour if any. Among other advantages, it also includes long working hours, supervision and repetition of the tasks. These cars, hence, can be used in factories and warehouses with large spaces and/ or geographical disparity where conditions are not suitable for humans to survive. The proposed work uses wireless software and hardware technologies such as a wireless transceiver module, Raspberry Pi as a microcontroller, H-bridge L293D IC for motor controller, and 4 DC motors used to move the automobile. The system uses the Open CV-Library for object detection and image processing. Also, the Firebase database and file storage service is used to send images from the car to the app. In order to make it hybrid, the system also proposes an IOS application that connects to the car so it can be moved manually. The system has been tested and the result shows that the car can move in any direction with the app-controlled remote and detect signs with an accuracy of 78%.