Region-dependent vehicle classification using PCA features

Video-based vehicle detection is the focus of increasing interest due to its potential towards collision avoidance. In particular, vehicle verification is especially challenging due to the enormous variability of vehicles in size, color, pose, etc. In this paper, a new approach based on supervised learning using Principal Component Analysis (PCA) is proposed that addresses the main limitations of existing methods. Namely, in contrast to classical approaches which train a single classifier regardless of the relative position of the candidate (thus ignoring valuable pose information), a region-dependent analysis is performed by considering four different areas. In addition, a study on the evolution of the classification performance according to the dimensionality of the principal subspace is carried out using PCA features within a SVM-based classification scheme. Indeed, the experiments performed on a publicly available database prove that PCA dimensionality requirements are region-dependent. Hence, in this work, the optimal configuration is adapted to each of them, rendering very good vehicle verification results.

Método de desvio de obstáculos aplicado em veículo autônomo.

A operação de veículos autônomos necessita de meios para evitar colisões quando obstáculos não conhecidos previamente são interpostos em sua trajetória. Algoritmos para executar o desvio e sensores apropriados para a detecção destes obstáculos são essenciais para a operação destes veículos. Esta dissertação apresenta estudos sobre quatro algoritmos de desvio de obstáculos e tecnologia de três tipos de sensores aplicáveis à operação de veículos autônomos. Após os estudos teóricos, um dos algoritmos foi testado para a comprovação da aplicabilidade ao veículo de teste. A etapa experimental foi a realização de um programa, escrito em linguagem de programação Java, que aplicou o algoritmo Inseto 2 para o desvio de obstáculos em uma plataforma robótica (Robodeck) com o uso de sensores ultrassônicos embarcados na referida plataforma. Os experimentos foram conduzidos em ambiente fechado (indoor), bidimensional e horizontal (plano), fazendo o Robodeck executar uma trajetória. Para os testes, obstáculos foram colocados para simular situações variadas e avaliar a eficácia do algoritmo nestas configurações de caminho. O algoritmo executou o desvio dos obstáculos com sucesso e, quando havia solução para a trajetória, ela foi encontrada. Quando não havia solução, o algoritmo detectou esta situação e parou o veículo.

Intelligent vehicle initiative: business plan.

Mode of access: Internet.

Evaluation of the Intelligent Cruise Control System. Volume I: study results.

National Highway Traffic Safety Administration, Washington, D.C.

Reducing highway crashes through administrative license revocation.

Mode of access: Internet.

Examination of lane change crashes and potential IVHS countermeasures. Final report.

National Highway Traffic Safety Administration, Washington, D.C.

Validation of the reduction of rear-end collisions by a high-mounted auxiliary stoplamp. Final report.

National Highway Traffic Safety Administration, Washington, D.C.

Guidelines for enhancement of visual conspicuity of the trailing end of trains. Final report.

Federal Railroad Administration, Office of Research, Development and Demonstrations, Washington, D.C.

Field test evaluation of rear lighting systems. Final report.

National Highway Traffic Safety Administration, Washington, D.C.

Examination of reduced visibility crashes and potential IVHS countermeasures. Final report.

National Highway Traffic Safety Administration, Washington, D.C.

Integration of in-vehicle electronics for IVHS and the electronics for other vehicle systems. Summary report.

National Highway Traffic Safety Administration, Washington, D.C.

Examination of intersection, left turn across path crashes and potential IVHS countermeasures. Final report.

National Highway Traffic Safety Administration, Washington, D.C.

Statistical evaluation of the effectiveness of Federal Motor Vehicle Safety Standard 108: Side Marker Lamps (Only). Final report.

National Highway Traffic Safety Administration, Washington, D.C.

Integration of in-vehicle electronics for IVHS and the electronics for other vehicle systems.

National Highway Traffic Safety Administration, Washington, D.C.

Experimental safety vehicle (phase two) developed by General Motors. Final report.

National Highway Traffic Safety Administration, Washington, D.C.