Learning to detect natural image boundaries using local brightness, color, and texture cues

@article{Martin04Learning, abstract = {The goal of this work is to accurately detect and localize boundaries in natural scenes using local image measurements. We formulate features that respond to characteristic changes in brightness, color, and texture associated with natural boundaries. In order to combine the information from these features in an optimal way, we train a classifier using human labeled images as ground truth. The output of this classifier provides the posterior probability of a boundary at each image location and orientation. We present precision-recall curves showing that the resulting detector significantly outperforms existing approaches. Our two main results are 1) that cue combination can be performed adequately with a simple linear model and 2) that a proper, explicit treatment of texture is required to detect boundaries in natural images.}, author = {Martin, D. R. and Fowlkes, C. C. and Malik, J. }, citeulike-article-id = {883988}, doi = {10.1109/TPAMI.2004.1273918}, journal = {Pattern Analysis and Machine Intelligence, IEEE Transactions on}, keywords = {ncuts, segmentation}, number = {5}, pages = {530--549}, posted-at = {2008-04-19 20:10:41}, priority = {2}, title = {Learning to detect natural image boundaries using local brightness, color, and texture cues}, url = {http://dx.doi.org/10.1109/TPAMI.2004.1273918}, volume = {26}, year = {2004} }

TY - JOUR ID - Martin04Learning L3 - citeulike-article-id:883988 TI - Learning to detect natural image boundaries using local brightness, color, and texture cues JF - Pattern Analysis and Machine Intelligence, IEEE Transactions on VL - 26 IS - 5 SP - 530 EP - 549 N2 - The goal of this work is to accurately detect and localize boundaries in natural scenes using local image measurements. We formulate features that respond to characteristic changes in brightness, color, and texture associated with natural boundaries. In order to combine the information from these features in an optimal way, we train a classifier using human labeled images as ground truth. The output of this classifier provides the posterior probability of a boundary at each image location and orientation. We present precision-recall curves showing that the resulting detector significantly outperforms existing approaches. Our two main results are 1) that cue combination can be performed adequately with a simple linear model and 2) that a proper, explicit treatment of texture is required to detect boundaries in natural images. KW - ncuts KW - segmentation AU - Martin, DR AU - Fowlkes, CC AU - Malik, J PY - 2004/// UR - http://dx.doi.org/10.1109/TPAMI.2004.1273918 ER -