Computer Vision Guided RC Car

Self-driving cars that use computer vision to detect the presence of obstacles in their path and to do waypoint navigation are being actively developed by various companies around the world. The objective of this project is to develop a small set-up in which a student may implement various algorithms and test them iteratively. This project consists of the car, the IR communication link and image processing software on a laptop. A toy car is made from two DC geared motors mounted on a perforated board. The car turns using differential drive and receives signals from the computer using an IR receiver. An Arduino Nano is used to decode the received IR signals and controls the two DC motors. The IR signals can be given by the IR transmitter board connected to laptop or a normal TV remote control. A fixed overhead camera mounted with a top down view of the field acquires images that is transferred to the linux laptop through USB. The laptop running python with openCV libraries is used to do the image processing. In the current version, the python programme detects the location and orientation of the car with respect to the target and guides the car towards the target. This python programme can be upgraded with various advanced algorithms without major modification to the hardware. The Linux laptop communicates with RC car using IR transmitter (IR led) and receiver VS1838. The IR transmitter is made using an Arduino Uno mounted on a perforated board connected to the laptop with a USB cable. The toy car contains the IR receiver, also interfaced to the Arduino Uno, which drives the car. The car shown has two 9V geared DC motor wheels which are powered by two 1.5V AA batteries in series. The power to the motor is given through BC547 transistors. BC 547 is a NPN general purpose transistor. It is commonly used in audio amplifiers and signal processing circuits. In this application, the transistor is used as a low side switch. The VS1838B IR receiver is used to receive commands from a TV remote control or the transmitter connected to the laptop. It is compatible with common IR remote control data formats. The widely used IRremote.h Arduino library is used to decode the received IR signals and convert them to integers that can then be acted upon.