A Self-driving car is no longer a distant idea.Soon the market is going to be hit by a car which drives itself. For a car to be able to drive itself it has to answer three important questions. The first question to be answered is where exactly is the car? Or what is the location of the car? This question can be answered with the help of GPS. GPS will tell the car where to go and where to stop? GPS also known as Global positioning System, provides the location and time information with the help of satellites. But GPS sometimes gives an inaccurate result, has a slower speed or is unreachable. So GPS is going to be used along with a computer vision. To discern the exact location of the car with an accuracy of a few meters, a single colour image is used. The side of the building, if any, in front of the car, can also be determined, even if all sides of the building are identical.
The Second question that needs to be answered is what is around the car? SegNet is the system that answers this question, using computer vision. The phone uses the images clicked by the camera of your smart phone. The computer assigns different colour codes to different objects like roads, bridges, street sign, pedestrian etc.
The third and most important question is what the car should do after analysing the first two question? A series of complex logarithms and software are designed to deal with this question.
The testing of the device has already started in Cambridge at the King’s Parade under the guidance of Prof. Roberto Cipolla. The main focus of the researchers is on the first two questions. As per their prediction self- driving car will debut in 2019. Self-driving car will also reduce the death toll in many major cities due to road accidents.