Comparison of Feature Extraction Techniques in Cotton Leaf Disease Classification using CBIR  
  Authors : Sushila Palwe; Chaitanya Budkule; Utkarsha Sonawane; 4 Vilas Giri

 

This paper presents a comparison of different Feature Extraction Techniques employed in Content Based Image Retrieval (CBIR), with its application in Cotton leaf Disease identification. Firstly, preliminary information about CBIR architecture is given. The accuracy of key points detection plays a vital role in the overall results in CBIR system. So, a comparative analysis of some well-known key point detectors are presented in this paper. After comparing them, it is finally concluded that SIFT is the overall best technique for key points detection in this particular CBIR system.

 

Published In : IJCAT Journal Volume 2, Issue 7

Date of Publication : July 2015

Pages : 259 - 268

Figures :09

Tables : --

Publication Link :Comparison of Feature Extraction Techniques in Cotton Leaf Disease Classification using CBIR

 

 

 

Sushila Palwe : Professor, Computer Department MAEER’s Maharashtra College of Engineering (MITCOE), Pune affiliated to Savitribai Phule Pune University.

Chaitanya Budkule : currently pursuing B.E Computer Engineering from MAEER’s Maharashtra College of Engineering (MITCOE), Pune affiliated to Savitribai Phule Pune University. (2014-15 Batch).

Utkarsha Sonawane : currently pursuing B.E Computer Engineering from MAEER’s Maharashtra College of Engineering (MITCOE), Pune affiliated to Savitribai Phule Pune University. (2014-15 Batch).

Vilas Giri : currently pursuing B.E Computer Engineering from MAEER’s Maharashtra College of Engineunering (MITCOE), P affiliated to Savitribai Phule Pune University. (2014-15 Batch).

 

 

 

 

 

 

 

CBIR

Feature Extraction

MSER

SURF

SIFT

ORB

For high Precision and Accuracy, which means lesser false positives in the output images, MSER is the best algorithm although its execution time becomes a bottleneck. SURF is also twice as faster as MSER and SIFT and thus imposes quite lesser execution overhead but with an ordinary Precision and Recall. So, if the goal is just high Accuracy and less execution time, SURF can be used as Feature Detection and Extraction technique.

 

 

 

 

 

 

 

 

 

[1] J. Matas, O. Chum, M. Urban, and T. Pajdla. "Robust wide baseline stereo from maximally stable extremal regions." Proc. of British Machine Vision Conference, pages 384-396, 2002. [2] Ryuji Funayama, Hiromichi Yanagihara, Luc Van Gool, Tinne Tuytelaars, Herbert Bay, "ROBUST INTEREST POINT DETECTOR AND DESCRIPTOR", published 2009-09-24 [3] Herbert Bay, Andreas Ess, Tinne Tuytelaars, Luc Van Gool "SURF: Speeded Up Robust Features", Computer Vision and Image Understanding (CVIU), Vol. 110, No. 3, pp. 346–359, 2008 [4] Ethan Rublee, Vincent Rabaud, Kurt Konolige, Gary Bradski "ORB: an efficient alternative to SIFT or SURF", Computer Vision (ICCV), 2011 IEEE International Conference on. IEEE, 2011 [5] http://en.wikipedia.org/wiki/ [6] www.fp.ucalgary.ca/mhallbey/tutorial.htm