Reconhecimento de Placas de Veículos Através da Aplicação de Técnicas de Processamento de Imagens e Redes Neurais Artificiais

Abstract

Digital images are increasingly present in people's lives and can be used for work or leisure. With the popularization of the Internet, especially social networks, has become increasingly practical to store and share images with people from different parts of the world. All these facilities afforded by digital images have increased a need for mechanisms of automation and decision making about this subject. Electronic component industries already use digital image processing to be defective electronic circuit boards, and different areas, such as medicine, it uses software features that help detect tumors or other diseases by analyzing radiographic images. In order to meet this need, focusing on a particular problem, this paper aims to develop software able to detect, extract and recognize alphanumeric characters contained in vehicles plates. At first, you will see what has been researched and developed over the work, citing some concepts and characteristics of digital images and neural networks. Then will be commented the techniques used to implement the proposed software, following the steps for image recognition. Finally, we will present the results and training that served to seek the best model to be used for the recognition of letters and numbers.