Automation Concept for 3D-Classification of Engineering-Shape Objects
This paper presents an automation concept for a test prototype for classification and retrieval of engineering-shape objects developed at RWTH Aachen University. A focus is the design to transfer classification results using simulated data to a hardware prototype that measures and classifies engineering shape. The hardware of the system is based on a simple stepper motor, two backlights, and two cameras. The software to classify the objects uses light-field descriptors—a Fourier-based shape-representation. It can be claimed that this prototype is the first implementation of light-field descriptors as a hardware prototype and application on physical measured engineering-shape objects.
Industrial Automations Intelligent Databases & Information Retrieval Image Processing Classification Pattern Recognition
Robert Schmitt Peter Fritz Gunter Nogueira Loch
Laboratory for Machine Tools and Production Engineering (WZL) RWTH Aachen University Aachen, Germany
国际会议
重庆
英文
424-428
2011-01-21（万方平台首次上网日期，不代表论文的发表时间）