Changes between Version 5 and Version 6 of GSoC/2016/Additional_segmentation_algorithms/mean_shift


Ignore:
Timestamp:
Sep 8, 2016, 1:04:54 PM (8 years ago)
Author:
hao2309
Comment:

--

Legend:

Unmodified
Added
Removed
Modified

  • GSoC/2016/Additional_segmentation_algorithms/mean_shift

    v5 v6  
    9696
    9797== Mean shift ==
    98 Numerous nonparametric clustering methods can be separated into two parts:hierarchical clustering and density estimation. Hierarchical clustering composes either aggregation or division based on some proximate measure. The concept of the density estimation-based nonparametric clustering method is that the feature space can be considered as the experiential probability density function (p.d.f.) of the represented parameter. The mean shift algorithm can be classified as density estimation. It adequately analyzes feature space to cluster them and can provide reliable solutions for many vision tasks. The basics of mean shift are discussed as below.
     98Numerous nonparametric clustering methods can be separated into two parts:hierarchical clustering and density estimation. Hierarchical clustering composes either aggregation or division based on some proximate measure. The concept of the density estimation-based nonparametric clustering method is that the feature space can be considered as the experiential probability density function (p.d.f.) of the represented parameter. The mean shift algorithm can be classified as density estimation. It adequately analyzes feature space to cluster them and can provide reliable solutions for many vision tasks. The basics of mean shift are discussed as below [3].
     99
    99100'''Pros:'''
    100 * An extremely versatile tool for feature space analysis.
     101* An extremely versatile tool for feature space analysis. Mean-shift algorithm is applicable not only in image segmentation, but also in motion detector, image filtering and etc.
    101102* Suitable for arbitrary feature spaces.
    102103'''Cons:'''
    103 * The kernel bandwidth is the only factor can control the output.
    104 * The computation time is quite long.
     104* The range bandwidth (hr) and spatial bandwidth (hs) are the only two factors can control the outpu, especially the range bandwith results in an extrem influence to the result.
     105* The computation time is quite long, for the iterations of anisotrophic filtering, clustering and merging small objects.
    105106
    106107== Other segmentation methods ==
    … …  
    142143'''Cons:'''
    143144
    144 [1]
    145 [2]Adams, Rolf, and Leanne Bischof. "Seeded region growing." IEEE Transactions on pattern analysis and machine intelligence 16.6 (1994): 641-647
     145[1] http://stackoverflow.com/questions/4831813/image-segmentation-using-mean-shift-explained [[BR]]
     146[2] Adams, Rolf, and Leanne Bischof. "Seeded region growing." IEEE Transactions on pattern analysis and machine intelligence 16.6 (1994): 641-647[[BR]]
     147[3] Y. Cheng, “Mean shift, mode seeking, and clustering,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 17, no. 8, pp. 790-799, Aug. 1995[[BR]]
     148[4]Wang, Yu-Hsiang. "Tutorial: Image Segmentation." National Taiwan University, Taipei (2010): 1-36.[[BR]]
     149