Object Detection via Boundary Structure Segmentation

We address the problem of object detection and segmentation using holistic properties of object shape. Global shape representations are highly susceptible to clutter inevitably present in realistic images, and can be robustly recognized only using a precise segmentation of the object. To this end, we propose a figure/ground segmentation method for extraction of perceptually salient regions in an image which resemble the global properties of a model boundary structure. Our boundary structure representation captures object shape and is based on geometric relationships of object boundary edges, while perceptual saliency favors coherent regions distinct from the background. We formulate the segmentation problem as an integer quadratic program and use a semidefinite programming relaxation to solve it. Obtained solutions provide both the segmentation of an object as well as a detection score used for object recognition. Our single-step approach improves over state of the art methods on seve...

• Type: Conference
• Year: 2010
• Where: CVPR
• Authors: Alexander Toshev, Ben Taskar, Kostas Daniilidis
• Keywords: Computer Vision, CVPR 2010, Object Boundary Edges, Object Detection, Object Shape