8.3.1 Complete Segmentation Systems Based on Ohlander Technique

Chapter Contents (Back)
Color. Color Segmentation. Segmentation, Color. Segmentation, Thresholds. Segmentation, Ohlander. Segmentation, Histogram.

Ohlander, R.[Ronald], Price, K.E.[Keith E.], Reddy, D.R.[D. Raj],
Picture Segmentation Using a Recursive Region Splitting Method,
CGIP(8), No. 3, December 1978, pp. 313-333.
Elsevier DOI
PDF File. Color. Segmentation, Systems. Segmentation, Histogram. Segmentation, Ohlander. Segmentation, Region Splitting. Select a region, generate histograms, choose the threshold based on the best separated peaks, repeat until uniform regions or too small. Various modifications are added to make it work for large and gray-level images. BibRef 7812

Price, K.E.,
Segmentation,
PRIP79(512-514). BibRef 7900 USC Computer Vision BibRef

Ohlander, R.[Ronald],
Analysis of Natural Scenes,
Ph.D.Thesis (CS), 1975, BibRef 7500 CMU-CS-TR-April, 1975. Original for the above paper. BibRef

Shafer, S.A.[Steven A.],
MOOSE Users' Manual Implementation Guide Evaluation,
TR IfI-HH-B-70/80, Bericht 70, Univ. HamburgApril 1980. Evaluation, Segmentation. Segmentation, Evaluation. An implementation of the Ohlander segmentation by Shafer when visiting U of Hamburg. BibRef 8004

Shafer, S.A., and Kanade, T.,
Recursive Region Segmentation by Analysis of Histograms,
ICASSP82(1166-1171). Segmentation, Systems. Phoenix. System: Phoenix.
HTML Version.
See also Phoenix Image Segmentation System: Description and Evaluation, The. After implementing a version of the Ohlander segmentation technique, Shafer proposed and implemented a variation that used the type of regions generated by the various possible threshold to determine the optimal threshold. This method applied all reasonable thresholds, as determined by analyzing the histograms, and chose the set of regions which were the most compact and had the clearest borders. This is based on the observation that, often, several histograms have peaks that correspond to the same regions, but one may give a more precise split than another even when its peak is not as clear according to the given criteria. BibRef 8200

Laws, K.I.,
The Phoenix Image Segmentation System: Description and Evaluation,
SRI AICenter-TN 289, December 1982. Evaluation, Segmentation. System: Phoenix. Phoenix. Segmentation, Evaluation. BibRef 8212

Laws, K.I.,
On the Evaluation of Scene Analysis Algorithms,
DARPA83(148-155). Segmentation, Evaluation. BibRef 8300

Bhanu, B.[Bir], and Lee, S.K.[Sung-Kee],
Genetic Learning for Adaptive Image Segmentation,
Hingham, MA: KluwerAcademic Press, 1994. ISBN 0-7923-9491-7, 296pp.
WWW Link. Genetic Algorithms. Segmentation, Learning. Complete description of segmentation system BibRef 9400

Bhanu, B.[Bir], Lee, S.K.[Sung-Kee], Ming, J.C.[John C.],
Adaptive image segmentation system,
US_Patent5,048,095, Sep 10, 1991
WWW Link. BibRef 9109

Bhanu, B., Lee, S., Das, S.,
Adaptive Image Segmentation Using Genetic And Hybrid Search Methods,
AeroSys(31), No. 4, October 1995, pp. 1268-1291. BibRef 9510

Bhanu, B., Lee, S., and Ming, J.,
Adaptive Image Segmentation Using a Genetic Algorithm,
SMC(25), No. 12, December 1995, pp. 1543-1567. BibRef 9512
Earlier: DARPA89(1043-1055). BibRef
And:
Self-Optimizing Control System for Adaptive Image Segmentation,
DARPA90(583-596). BibRef
Earlier: A1, A3 and A2:
Adaptive Image Segmentation,
Draftversion. BibRef
Earlier:
Closed-Loop Adaptive Image Segmentation,
CVPR91(734-735).
IEEE DOI Given a known (hand) segmentation, vary the parameters of the Phoenix segmentation system (from the IU Testbed (
See also Phoenix Image Segmentation System: Description and Evaluation, The. )) to get the best fit. The program learns the parameter values through a genetic algorithm (mutation or exchange) rather than a complete search. BibRef

Bhanu, B.[Bir], Bao, X.[Xin], Ping, J.[Jing],
Reinforcement learning Integrated Image Segmentation and Object Recognition,
DARPA97(1145-1154). BibRef 9700

Bhanu, B.[Bir], Lee, S.K.[Sung-Kee], and Das, S.[Subhodev],
Adaptive Image Segmentation Using Multi-Objective Evaluation and Hybrid Search Methods,
AAAI-MLCV93University of California, Riverside. BibRef 9300

Bhanu, B.[Bir], Fonder, S.[Stephanie],
Learning Based Interactive Image Segmentation,
ICPR00(Vol I: 299-302).
IEEE DOI 0009
BibRef

Bhanu, B., and Faugeras, O.D.,
Segmentation of Images Having Unimodal Distributions,
PAMI(4), No. 4, July 1982, pp. 408-419. BibRef 8207
And: Correction: PAMI(4), No. 6, November 1982, pp. 689. Segmentation, Unimodal. BibRef

Bhanu, B.[Bir], Parvin, B.A.[Bahram A.],
Segmentation of Natural Scenes,
PR(20), No. 5, 1987, pp. 487-496.
Elsevier DOI BibRef 8700
Earlier:
Segmentation of Images Using a Relaxation Technique,
CVPR83(151-153). Segmentation, Relaxation. BibRef

Ohta, Y.[Yu_Ichi], Kanade, T.[Takeo], and Sakai, T.[Toshiyuki],
Color Information for Region Segmentation,
CGIP(13), No. 3, July 1980, pp. 222-241.
Elsevier DOI BibRef 8007
And:
A Production System for Region Analysis,
IJCAI79(684-686). Segmentation, Color. Color, Transforms. Optimal transform generation, uses
See also Picture Segmentation Using a Recursive Region Splitting Method. for the basic method. Computed the eigenvectors for a number of sample images to find optimal weights for the transforms for segmentation. The first is close to just an average of the 3 ((R+G+B)/3). The others are: I2=(R-B)/2 (or (B-R)/2); and I3=(2G -R -B)/4. This linear transform generates results similar to dynamic Karhunev Loeve transformation of RGB. BibRef

Ohta, Y.[YuIchi],
Knowledge-Based Interpretation of Outdoor Natural Color Scenes,
Morgan Kaufmann1985. BibRef 8500 Bookversion of the thesis? BibRef

Ohta, Y.[YuIchi],
A Region-Oriented Image-Analysis System by Computer,
Ph.D.Thesis (Info. Sci.), March 1980. BibRef 8003 Kyoto Univ. Model Based Recognition. A more complete report than the segmentation paper above. BibRef

Chapter on 2-D Region Segmentation Techniques, Snakes, Active Contours continues in
Complete Systems Derived from the Univ. Massachusetts Work .


Last update:Mar 16, 2024 at 20:36:19