## THRESHOLD COMPARISON STUDY

 The following techniques are compared: 2 Color Adaptive Spatial Subdivision Color Reduction Otsu's Between Class Variance Isodata K-Means Fuzzy C-Means Kapur's Entropy Sahoo's Entropy (power=2) Sahoo's Entropy (power=0.1) Triangle Method Local Adaptive (Mean Plus Offset) Pictures Were Obtained from: blood.jpg from http://www.istanbul.edu.tr/eng/ee/jeee/main/pages/issues/is62/62008.pdf fingerprint.jpg from http://www.istanbul.edu.tr/eng/ee/jeee/main/pages/issues/is62/62008.pdf flower.jpg from http://www.jhlabs.com/ip/blurring.html house.jpg from http://en.wikipedia.org/wiki/Otsu's_method kanji.jpg from http://www.measurement.sk/2004/S1/Yong.pdf lena2g_edge1.jpg was created using the IM function -edge 1 parts.gif from http://www.ph.tn.tudelft.nl/Courses/FIP/noframes/fip-Segmenta.html rice.jpg from http://www.istanbul.edu.tr/eng/ee/jeee/main/pages/issues/is62/62008.pdf tank.jpg from http://stinet.dtic.mil/cgi-bin/GetTRDoc?AD=ADA464347&Location=U2&doc=GetTRDoc.pdf textsample.jpg from http://signal.ece.utexas.edu/seminars/dsp_seminars/01fall/211_seeger_mf.pdf My rankings are based solely on visual inspection and personal preferences.

 Original Image 2 Color Otsu BCV Isodata K-Means Fuzzy C-Means Kapur Entropy Sahoo Entropy P=2 Sahoo Entropy P=0.1 Triangle Method Local Adaptive RESULTS Not Processed(Not Applicable?) Best: Fuzzy C-Means Close Seconds: Triangle Sahoo P=2

 Original Image 2 Color Otsu BCV Isodata K-Means Fuzzy C-Means Kapur Entropy Sahoo Entropy P=2 Sahoo Entropy P=0.1 Triangle Method Local Adaptive RESULTS Not Processed(Not Applicable?) Close Tie: 2 Color Otsu BCV Isodata K-Means Fuzzy C-Means

 Original Image 2 Color Otsu BCV Isodata K-Means Fuzzy C-Means Kapur Entropy Sahoo Entropy P=2 Sahoo Entropy P=0.1 Triangle Method Local Adaptive RESULTS Not Processed(Not Applicable?) Close Tie: 2 Color Otsu BCV Isodata K-Means Fuzzy C-Means Kapur Entropy Sahoo Entropy P=0.1

 Original Image 2 Color Otsu BCV Isodata K-Means Fuzzy C-Means Kapur Entropy Sahoo Entropy P=2 Sahoo Entropy P=0.1 Triangle Method Local Adaptive RESULTS Not Processed(Not Applicable?) Best: 2 Color Close Seconds: Otsu BCV Isodata K-Means

 Original Image 2 Color Otsu BCV Isodata K-Means Fuzzy C-Means Kapur Entropy Sahoo Entropy P=2 Sahoo Entropy P=0.1 Triangle Method Local Adaptive RESULTS Not Processed(Not Applicable?) Best: Sahoo Entropy P=0.1 Close Seconds: Sahoo Entropy P=2 Kapur Entropy Fuzzy C-Means

 Original Image 2 Color Otsu BCV Isodata K-Means Fuzzy C-Means Radius=30 Kapur Entropy Sahoo Entropy P=2 Sahoo Entropy P=0.1 Triangle Method Local Adaptive RESULTS Not Processed(Not Applicable?) Close Tie: 2 Color Otsu BCV Isodata K-Means Sahoo Entropy P=0.1 Close Second: Fuzzy C-Means Radius=30

 Original Image 2 Color Otsu BCV Isodata K-Means Fuzzy C-Means Radius=3 Kapur Entropy Sahoo Entropy P=2 Sahoo Entropy P=0.1 Triangle Method Local Adaptive RESULTS Tie: Local Adaptive Fuzzy C-Means Radius=3 Close Seconds: 2 Color Otsu BCV Isodata K-Means Kapur Entropy

 Original Image 2 Color Otsu BCV Isodata K-Means Fuzzy C-Means Kapur Entropy Sahoo Entropy P=2 Sahoo Entropy P=0.1 Triangle Method Local Adaptive RESULTS Close Tie: 2 Color Otsu BCV Isodata K-Means Sahoo Entropy P=0.1 Close Seconds: Fuzzy C-Means Triangle Method

 Original Image 2 Color Otsu BCV Isodata K-Means Fuzzy C-Means Kapur Entropy Sahoo Entropy P=2 Sahoo Entropy P=0.1 Triangle Method Local Adaptive RESULTS Best: Local Adaptive Close Seconds: Triangle Method Sahoo Entropy P=2 Fuzzy C-Means Radius=10

 Original Image 2 Color Otsu BCV Isodata K-Means Fuzzy C-Means Kapur Entropy Sahoo Entropy P=2 Sahoo Entropy P=0.1 Triangle Method Local Adaptive RESULTS Not Processed(Not Applicable?) Close Tie: Otsu BCV Isodata K-Means Triangle Method

 Original Image 2 Color Otsu BCV Isodata K-Means Fuzzy C-Means Kapur Entropy Sahoo Entropy P=2 Sahoo Entropy P=0.1 Triangle Method Local Adaptive RESULTS Best: Local Adaptive

 Summary Of Rankings 11 Images 2Color OtsuBCV Isodata K-Means FuzzyC-Means KapurEntropy SahooEntropyP=2 SahooEntropyP=0.1 TriangleMethod LocalAdaptive Number OfFirst Places 5 5 5 5 4 1 0 4 1 3 Number OfSecond Places 1 2 2 2 4 2 3 0 2 1 Number OfEither Places 6 7 7 7 8 3 3 4 3 4

## Conclusion: No One Method Is Universally Best

(as if we already did not know that)