Implementation of an Map Following Vehicle (MFV) is a step towards making a ground vehicle that can work at its own, which can make decisions and do the tasks which we require from it to be done. It would be a vehicle which can process the data in real time and take the respective decisions on the bases of that processed data. The project is based on integration of different hardware and software modules and their interaction to produce an efficient outcome.
The basic idea of this project is to design a map following vehicle that would be autonomous. Meaning that, it can go to the places where sending a human would be a life threatening process and do the tasks which we specify it to do. It can take decisions at its own using the camera as an input device to acquire the video of the path, then this video is sent to the onboard computer system which processes the desired frame (image) of the video in real time and takes the decisions on the basis of that processed image that whether the space in front of the vehicle is a clear way or is it a hurdle. If the system finds it as a hurdle, then it changes the route of the vehicle and switch to a way where the track is clear for the vehicle to move forward and so on.
The vehicle is designed on a simple toy car of the size of approximately 2x2 feet, having a computer system onboard with a camera, a pair of batteries, microcontroller circuitry, motor driver circuitry, and motors to drive the car.
The project requires the programming skills to be utilized in the processing of the image and as well as for the micro controller. The programs for Digital Image Processing (DIP) are written in MATLAB. The C language and VMLAB software is used for the programming of the microcontroller. The project also requires the skills of circuit designing using electronic components; to design a biasing circuit of a microcontroller, the motor driver circuitry, and circuitry for interface the modules.
Our objective of this project is to develop a vision based map following vehicle which can detect obstacle on its way and avoid it. For that purpose we have design a test bed/prototype. To find the scope and take an effective step towards the betterment and optimization of the algorithm. To design a prototype of a map following vehicle using an ordinary single camera mounted on the top of map floor. In this we can lead to a simple and cheap solution to the problem of map following with obstacle detection and avoidance.
1.3 Project Report Structure
The following list will explain what will be covered in different chapters of this thesis: * In chapter 2, state of art is discussed the project scientific background. * In chapter 3, problem and proposed system is discussed in detail. * In chapter 4, the methodology of obstacle detection and avoidance of the map following vehicle is explained. * In chapter 5, the hardware and software is explained here. * In chapter 6, the future work is purposed.
State of Art
A number of interesting research results have been obtained on problems which are relevant to map following vehicles. A quick review of the salient systems provides a picture of the current state of the scientific art.
Planning a path based on a model is a problem that is fundamental to intelligent control of map following vehicles. Lozano-Pkrez  has developed a formal version of the general map following problem. This formalization is referred to as the "find-path" problem [62, 20]. In its most general form, the goal of find-path is to determine a continuous path for an object from an initial location to a goal location without colliding with an obstacle. Lozano-Pbrez provided a mathematical treatment of the find-path problem using the "configuration space" approach. The...