In the purest sense, model-based recognition uses stored information about an object’s shape and/or appearance to determine the orientation and position of that object in an image. Once the pose of the object has been recovered, qualitative measurements are made to support or refute the belief that the object is actually present in the scene at the hypothesized location. Typically, a set of sequential steps, referred to as a pipeline, are used in the model-based recognition paradigm.
KeywordsObject Recognition Model Feature Training Image Data Feature Integrate Graphic
Unable to display preview. Download preview PDF.