Figure/Ground Assignment in Natural Images

Abstract. Figure/ground assignment is a key step in perceptual organization which assigns contours to one of the two abutting regions, providing information about occlusion and allowing high-level processing to focus on non-accidental shapes of figural regions. In this paper, we develop a computational model for figure/ground assignment in complex natural scenes. We utilize a large dataset of images annotated with human-marked segmentations and figure/ground labels for training and quantitative evaluation. We operationalize the concept of familiar configuration by constructing prototypical local shapes, i.e. shapemes, from image data. Shapemes automatically encode mid-level visual cues to figure/ground assignment such as convexity and parallelism. Based on the shapeme representation, we train a logistic classifier to locally predict figure/ground labels. We also consider a global model using a conditional random field (CRF) to enforce global figure/ground consistency at T-junctions. We...

• Type: Conference
• Year: 2006
• Where: ECCV
• Authors: Xiaofeng Ren, Charless Fowlkes, Jitendra Malik
• Keywords: Bottom-up Figure/ground Assignment, Computer Vision, Correct Figural Assignment, ECCV 2006, Global Crf Model, Global Model, Local Shapeme Model