Both example-based and model-based approaches for classifying contour shapes can encounter difficulties when dealing with classes that have large nonlinear variability, especially...
Abstract. A novel approach for model set based object segmentation is described. The proposed method enables the using of a model set to guide the object segmentation. The object s...
Due to distortion, noise, segmentation errors, overlap, and occlusion of objects in digital images, it is usually impossible to extract complete object contours or to segment the ...
Segmentation of ancient documents is challenging. In the worst cases, text characters become fragmented as the results of strong degradation processes. New active contour methods ...
Shape matching has many applications in computer vision, such as shape classification, object recognition, object detection, and localization. In 2D cases, shape instances are 2D c...