Automatic unloading of piled boxes of unknown dimensions is undoubtedly of great importance to the industry. In this contribution a system addressing this problem is described: a l...
Most modern computer vision systems for high-level
tasks, such as image classification, object recognition and
segmentation, are based on learning algorithms that are
able to se...
A method for reconstruction of 3D polygonal models from multiple views is presented. The method uses sampling techniques to construct a texture-mapped semi-regular polygonal mesh ...
In this paper, we present an approach toward pedestrian detection and tracking from infrared imagery using joint shape and appearance cues. A layered representation is first intr...
Background: Microscopists are familiar with many blemishes that fluorescence images can have due to dust and debris, glass flaws, uneven distribution of fluids or surface coatings...