Keith Price Bibliography Multi-level Segmentation and Smoothing Methods

8.5.1 Multi-level Segmentation and Smoothing Methods

Chapter Contents (Back)
Smoothing. Multi-level Segmentation. Segmentation, Multi-Level. Pyramid Structure. See also Smoothing Techniques, Adaptive Smoothing.

Ranade, S., and Shneier, M.O.,
Using Quadtrees to Smooth Images,
SMC(11), No. 5, May 1981, pp. 373-376. BibRef 8105

Gross, A.D., and Rosenfeld, A.,
Multiresolution Object Detection and Delineation,
CVGIP( 39), No. 1, July 1987, pp. 102-115.
WWW Version. (Strange title) Segmentation using interest operator at different levels to determine when to segment an area. BibRef 8707

Bongiovanni, G., Cinque, L., Levialdi, S., and Rosenfeld, A.,
Image Segmentation by a Multiresolution Approach,
PR(26), No. 12, December 1993, pp. 1845-1854.
WWW Version. BibRef 9312

Cinque, L., Levialdi, S., Rosenfeld, A.,
Fast Pyramidal Algorithms for Image Thresholding,
PR(28), No. 6, June 1995, pp. 901-906.
WWW Version. Segmentation, Thresholds. BibRef 9506

Cantoni, V., Lombardi, L., Cinque, L., Levialdi, S., Guerra, C.,
Recognizing 2D Objects by a Multi-Resolution Approach,
ICPR94(C:310-316).
IEEE DOI Link BibRef 9400

Hong, T.H.[Tsai-Hong], Narayanan, K.A., Peleg, S.[Shmuel], Rosenfeld, A., Silberberg, T.M.[Teresa M.],
Image Smoothing and Segmentation by Multiresolution Pixel Linking: Further Experiments and Extensions,
SMC(12), No. 5, September/October 1982, pp. 611-622. This reports on more work on pixel linking (from BibRef 8209 UMD-TR 977, 989, 1025). The Eklundh and Rosenfeld paper is the primary reference. BibRef

Narayanan, K.A., Rosenfeld, A.,
Approximation of Waveforms and Contours by One-Dimensional Pyramid Linking,
PR(15), No. 5, 1982, pp. 389-396.
WWW Version. BibRef 8200

Silberberg, T.[Teresa], Peleg, S.[Shmuel], and Rosenfeld, A.,
Multi-Resolution Pixel Linking for Image Smoothing and Segmentation,
DARPA81(32-38). BibRef 8100

Eklundh, J.O., and Rosenfeld, A.,
Image Smoothing Based on Neighbor Linking,
PAMI(3), No. 6, November 1981 pp. 679-683. BibRef 8111

Eklundh, J.O., and Rosenfeld, A.,
Peak Detection Using Difference Operators,
PAMI(1), No. 3, July 1979, pp. 317-325. BibRef 7907

Pietikainen, M.[Matti], and Rosenfeld, A.[Azriel],
Image Segmentation by Texture Using Pyramid Node Linking,
SMC(11), No. 12, December 1981, pp. 822-825. This extends the pyramid linking method to using texture parameters for the linking decision. See also Texture discrimination with multidimensional distributions of signed gray-level differences. BibRef 8112

Pietikainen, M.[Matti], and Rosenfeld, A.[Azriel],
Gray Level Pyramid Linking as an Aid in Texture Analysis,
SMC(12), 1982, pp. 422-429. BibRef 8200

Pietikainen, M.[Matti], Rosenfeld, A.[Azriel], and Davis, L.S.,
Experiments with Texture Classification Using Averages of Local Pattern Matches,
SMC(13), 1983, pp. 421-426. BibRef 8300

Pietikainen, M.[Matti],
Image Texture Analysis and Segmentation,
Acta Universiartis Ouluensis, Series C, Technica No. 21, Electornica No. 1. Univ. of OuluMarch 1982. Oulu, Finland. Evaluation, Texture. A comparative study and a look at Pyramid node linking from Pietikainen's texture work while at UMd. BibRef 8203

Burt, P.J.[Peter J.], Hong, T.H.[Tsai-Hong], and Rosenfeld, A.[Azriel],
Segmentation and Estimation of Image Region Properties Through Cooperative Hierarchical Computation,
SMC(11), No. 12, December 1981, pp. 802-809. This paper gives the primary definition of the linked pyramid method. Otherwise, little content. BibRef 8112

Pietikainen, M.[Matti], Rosenfeld, A.[Azriel], and Walter, I.[Ingrid],
Split-and-Link Algorithms for Image Segmentation,
PR(15), No. 4, 1982, pp. 287-298.
WWW Version. An alternative to split and merge. The major difference is that the levels are overlapped so that each pixel can be the descendant of several (4) others. The linking is done to the one that looks most like yourself. This eliminates some of the problems in the initial splitting. BibRef 8200

Cibulskis, J.M., Dyer, C.R.,
An Analysis of Node Linking in Overlapped Pyramids,
SMC(14), 1984, pp. 424-436. BibRef 8400

Baronti, S.[Stefano], Casini, A.[Andrea], Lotti, F.[Franco], Favaro, L., and Roberto, V.,
Variable Pyramid Structures for Image Segmentation,
CVGIP(49), No. 3, March 1990, pp. 346-356.
WWW Version. Various modifications to the earlier linked pyramid algorithms. BibRef 9003

Baronti, S.[Stefano], Casini, A.[Andrea], Lotti, F.[Franco], Porcinai, S.[Simone],
Segmentation of multispectral images of works of art through principal component analysis,
CIAP97(I: 14-21).
WWW Version. 9709
BibRef

Grosky, W.I., and Jain, R.C.,
A Pyramid-Based Approach to Segmentation Applied to Region Matching,
PAMI(8), No. 5, September 1986, pp. 639-650. BibRef 8609

Jain, R., Grosky, W.I.,
Hyper-Pyramids for Integration of Spatial Information,
SRMSF87(72-81). BibRef 8700

Antonisse, H.J.,
Image Segmentation in Pyramids,
CGIP(19), No. 4, August 1982, pp. 367-383.
WWW Version. Linked Pyramids: reduce with overlap, link to parent when average is closest, recompute the averages using only those linked, reset links, etc. until no change (at all levels). Limited to 4 classes (top of pyramid). Various "fixes" produce similar results (holes, etc. cause problems). Unforced linking - not forced to link to a parent (form new regions at higher resolutions). Link at low tolerance and cluster the small regions. BibRef 8208

Lifshitz, L.M., and Pizer, S.M.,
A Multiresolution Hierarchical Approach to Image Segmentation Based on Intensity Extrema,
PAMI(12), No. 6, June 1990, pp. 529-540.
IEEE Abstract. IEEE Top Reference.
WWW Version. Different smoothing levels, linked between levels, applied to medical image analysis. BibRef 9006

Morse, B.S., Pizer, S.M., Liu, A.,
Multiscale Medial Analysis Of Medical Images,
IVC(12), No. 6, July-August 1994, pp. 327-338.
WWW Version. BibRef 9407

Haring, S., Viergever, M.A., Kok, J.N.,
Kohonen Networks For Multiscale Image Segmentation,
IVC(12), No. 6, July-August 1994, pp. 339-344.
WWW Version. BibRef 9407

Liu, J.Q., Yang, Y.H.,
Multiresolution Color Image Segmentation,
PAMI(16), No. 7, July 1994, pp. 689-700.
IEEE Abstract. IEEE Top Reference.
WWW Version. Segmentation, Color. Evaluation, Segmentation. Segmentation, Evaluation. Introduced an evaluation function for segmentation that rewards fewer regions, larger (i.e. similar sizes) regions, and uniform regions. SQRT(Regions)*Sum Over Regions of (difference from mean squared divided by SQRT(region size). Use neighborhoods. BibRef 9407