Abstract
A general non-parametric technique is proposed for the analysis of a complex multimodal feature space and to delineate arbitrarily shaped clusters in it. The basic computational module of the technique is an old pattern recognition procedure: the mean shift. For discrete data, we prove the convergence of a recursive mean shift procedure to the nearest stationary point of the underlying density function and, thus, its utility in detecting the modes of the density. The relation of the mean shift procedure to the Nadaraya-Watson estimator from kernel regression and the robust M-estimators; of location is also established. Algorithms for two low-level vision tasks discontinuity-preserving smoothing and image segmentation - are described as applications. In these algorithms, the only user-set parameter is the resolution of the analysis, and either gray-level or color images are accepted as input. Extensive experimental results illustrate their excellent performance.
Keywords
Mean-shiftEstimatorSmoothingPattern recognition (psychology)Artificial intelligenceDensity estimationMathematicsKernel density estimationComputer scienceAlgorithmFeature vectorKernel (algebra)Feature (linguistics)Computer visionStatistics
Affiliated Institutions
Related Publications
Active appearance models
We describe a new method of matching statistical models of appearance to images. A set of model parameters control modes of shape and gray-level variation learned from a trainin...
Segmentation into Three Classes Using Gradients
Consider a three-dimensional "scene" in which a density f(x, y, z) is assigned to every point (x, y, z). In a discretized version of the scene the density D(i, j, k) assigned to...
Single-Image Super-Resolution Using Sparse Regression and Natural Image Prior
This paper proposes a framework for single-image super-resolution. The underlying idea is to learn a map from input low-resolution images to target high-resolution images based ...
Multi-Image Matching Using Multi-Scale Oriented Patches
This paper describes a novel multi-view matching framework based on a new type of invariant feature. Our features are located at Harris corners in discrete scale-space and orien...
Learning Hierarchical Features for Scene Labeling
Scene labeling consists of labeling each pixel in an image with the category of the object it belongs to. We propose a method that uses a multiscale convolutional network traine...
Publication Info
- Year
- 2002
- Type
- article
- Volume
- 24
- Issue
- 5
- Pages
- 603-619
- Citations
- 11224
- Access
- Closed
External Links
Social Impact
Altmetric
PlumX Metrics
Social media, news, blog, policy document mentions
Citation Metrics
11224
OpenAlex
Cite This
Dorin Comaniciu,
Peter Meer
(2002).
Mean shift: a robust approach toward feature space analysis.
IEEE Transactions on Pattern Analysis and Machine Intelligence
, 24
(5)
, 603-619.
https://doi.org/10.1109/34.1000236
Identifiers
- DOI
- 10.1109/34.1000236