Construeix el teu propi joc de mòbil amb visió per ordinador!

Abstract

Few years ago the computer vision was relegated to research, education and industry areas, but in recent years it has been approaching to the common user. Some new technological developments such as the Microsoft Kinect sensor, the Android smartphones unlocking system through facial recognition or the capacity to read road signs that were introduced in some new car are some examples of the introduction of the computer vision into the daily routine of the common user. This project aims to contribute to this introduction of the computer vision by developing a mobile application for the mobile operating system Android using computer vision. The main objective of this application is to create a platform game where the user can draw on paper the desired screen beforehand, then make photography and finally play it on his mobile device. The application uses latest software such as the development environment and language Processing and the development environment Eclipse. The game performance begins with the user drawing the desired screen on a piece of paper and using different colours to identify the different game concepts such as walls and floor, the starting point for the player in the game, the final position where the game ends and some obstacles that can kill the player. Then the user takes a picture of the drawn screen with the device camera and the picture is processed. After that the system analyzes the colours painted by the user throughout the HSV colour model in order to interpret what kind of game element is each colour and the location of each element. These saved locations create a new game screen placing the elements where the user has indicated in his paper and finally the user can play his own screen. This project uses the Java programming language and the Eclipse environment with the plugging ADT (Android Developer Tools) to develop all the code for the image processing and the colour identification of the elements of the game. It has uses language and environment Processing to develop all the construction and the generation of the game based on the picture taken by the user.