Processo de reconhecimento não supervisionado de áreas de estacionamento

• Ano de defesa: 2014
• Autor(a) principal: Mexas, Antonio Henrique
• Orientador(a): Marengoni, Maurício
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Presbiteriana Mackenzie
• Palavras-chave em Português: processamento de imagens; segmentação de imagens; análise de imagens e reconhecimento de padrões
• Palavras-chave em Inglês: image processing; image segmentation; images analys is and recognition patterns
• Área do conhecimento CNPq: CNPQ::ENGENHARIAS::ENGENHARIA ELETRICA
• Link de acesso: http://dspace.mackenzie.br/handle/10899/24355

The detection of objects in an image is important for decision making in a remote sensing system. The interest areas in this kind of work are buildings, streets, bridges, airports, ports and parking. The latter will be addressed in this work. The detection of these objects using manual techniques can be laborious and result in detection errors. This paper examines algorithms to the identification and recognition of parking areas paved and marked. To define the processing the best parameters are analised, with the purpose to obtain more accurate results. This work proposes an algorithm composed of recognition of parking areas that uses important variables such as vehicle size, size of parking spaces, image resolution and camera zoom. The experimental results obtained with the proposed method have a higher hit rate than other academic approaches.