Robust analysis of feature spaces: color image segmentation

By Dorin Comaniciu, Peter Meer. 1997

Subfields: Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

Using Multiple Segmentations to Discover Objects and their Extent in Image Collections (2006, methodology): “It is naive to expect a segmentation algorithm to partition an image into its constituent objects – in the general case, you need to have solved the recognition problem already! In practice, some approaches, like Mean-shift [4], perform only a low-level over-segmentation of the image (superpixels).”
Background Modeling using Mixture of Gaussians for Foreground Detection - A Survey (2008, methodology): “[145] combine the foreground detection obtained by a block-wise MOG with a color image segmentation [163] using a projecting method [164].”
Segment-based stereo matching using graph cuts (2004, methodology): “In our current implementation, mean-shift color segmentation algorithm [3] is used.”
State of the art on automatic road extraction for GIS update: a novel classification (2003, background): “Comaniciu and Meer (2001) proposes a general technique for the recovery of significant image features based on the mean shift algorithm and density gradients analysis.”
Statistical cues for domain specific image segmentation with performance analysis (2000, background): “Although there has been recent progress in general purpose image segmentation, for example [2], [10], [12], it remains an extremely diﬃcult problem.”
Efficient Graph-Based Image Segmentation (2004, unknown): “One common approach to image segmentation is based on mapping each pixel to a point in some feature space, and then finding clusters of similar points (e.g., Comaniciu and Meer, 1997, 1999; Jain and Dubes, 1988).”
Statistical region merging (2004, methodology): “…be partially due to the fact that most approaches use assumptions on distributions, more or less restrictive, which would make any theoretical insight into how region merging works restricted to such settings and, therefore, of possibly moderate interest (see, e.g., [10] for related criticisms).”
SAIL-VOS: Semantic Amodal Instance Level Video Object Segmentation – A Synthetic Dataset and Baselines (2019, methodology): “Early segmentation methods group perceptually similar pixels, e.g ., color image segmentation by Comaniciu and Meer [19], normalized cuts by Shi and Malik [88] or implicit shape models [60].”
Tracking Using CamShift Algorithm and Multiple Quantized Feature Spaces (2004, methodology): “In order to assign higher weighting to pixels nearer to the region center, a weighted histogram may be used to compute the target histogram (Comaniciu et al ., 1996) [ ] Where the resulting histogram is scaled using Equation 2 for the discrete quantities we are using and is any convex,…”
Approximate graph edit distance computation by means of bipartite graph matching (2009, background): “2 the prototype graph and a graph instance for each distortion level representing the letters A, M, K, and Z are illustrated.”
Mean Shift: A Robust Approach Toward Feature Space Analysis (2002, methodology): “Our first image segmentation algorithm was a straightforward application of t he feature space analysis technique to anL u v representation of the color image [11].”
Scaling Open-Vocabulary Image Segmentation with Image-Level Labels (2021, background): “In machine perception, early works [11,42] group pixels based on local affinities.”
Stereo Matching Using Belief Propagation (2002, methodology): “In our experiments, the segmentation labels are produced by the Mean-Shift algorithm [9].”
Understanding Deep Learning Techniques for Image Segmentation (2019, methodology): “ense [119], forensics such as face [5], iris [51] and ngerprint [144] recognition. Generally traditional methods such as histogram thresholding [195], hybridization [193, 87] feature space clustering [40], regionbased approaches [59], edge detection approaches [184], fuzzy approaches [39], entropy-b”
FRVT 2006: Quo Vadis face quality (2010, methodology): “To identify contiguous regions in the resulting face image, the Rutgers image segmentation algorithm[21] was run on the normalized images to produce a labeled segmentation.”
Semantic object classes in video: A high-definition ground truth database (2009, methodology): “When a frame is loaded, three automatic segmentation algorithms are run: Felzenszwalb and Huttenlocher (2004), Mean Shift (Comaniciu and Meer, 1997) and the pyramid-based algorithm of Burt et al.”
IAM Graph Database Repository for Graph Based Pattern Recognition and Machine Learning (2008, methodology): “We ﬁrst segment images into regions of homogeneous color using a mean shift algorithm [24].”
Computer Vision Face Tracking For Use in a Perceptual User Interface (1998, background): “The work described in this paper is part of a larger effort aimed at giving computers the ability to segment, track, and understand the pose, gestures, and emotional expressions of humans and the tools they might be using in front of a computer or settop box.”
Weighted persistent homology (2017, background): “In certain scenarios, such as color detection in computer vision [6,20,22], each pixel may play different roles depending on its color.”