A PRESENTATION ON
Detection and Classification of Apple Fruit Diseases using Complete Local Binary Pattern
Presented by --Shiv Ram Dubey
Table of Contents: Introduction ……………………………………………Slide 3 Previous Work ………………………………………..Slide 6
The Proposed Approach …………………………Slide 10 Experimental Result ……………………………….Slide 18
Conclusions …………………………………………….Slide 28 References ………………………………………………Slide 31
2
5/12/2014
Introduction
3
5/12/2014
Introduction
The classical approach for detection and identification of fruit diseases is based on the naked eye observation by the experts.
In some developing countries, consulting experts are expensive and time consuming due to the distant locations of their availability.
4
5/12/2014
Introduction (Cont …)
Automatic detection of fruit diseases is essential to automatically detect the symptoms of diseases as early as they appear on the growing fruits.
Some disease also infects other areas of the tree.
Some common diseases of apple fruits are apple scab, apple rot, and apple blotch [1].
5
5/12/2014
Previous Work
6
5/12/2014
Previous Work
Majority of the works performing defect segmentation of apples are done using simple threshold approach ([2], [3]).
A globally adaptive threshold method to segment fecal contamination defects on apples are presented in [4].
Bayesian classification is used by researchers ([5], [6]), where pixels are compared with a pre-calculated model and classified as defected or healthy.
7
5/12/2014
Previous Work (Cont …) Unsupervised classification does not benefit any guidance in the learning process due to lack of target values. Such an approach was used by [7] for defect segmentation. In [8], Ojala et al used the Local Binary Pattern histogram for rotation invariant texture classification and it has reported impressive classification outcomes on representative texture databases [9]. 8
5/12/2014
Previous Work (Cont …) Local Binary Pattern has also been adapted by other applications, such as face recognition [10] dynamic texture recognition [11] and shape localization [12]. A Complete Local Binary Pattern is presented in [13] as the completed modelling of Local Binary Pattern.
9
5/12/2014
The Proposed Approach
10
5/12/2014
The Proposed Approach
Framework of the proposed approach 11
5/12/2014
Image Segmentation K-Means clustering technique is used for the image segmentation. Images are partitioned into four clusters in which one cluster contains the majority of the diseased part of the image. K-Means clustering algorithm [14] was developed by J. MacQueen (1967). 12
5/12/2014
Image Segmentation (Cont …)
K-Means clustering for an apple fruit that is infected with apple scab disease (a) The infected fruit image, (b) first cluster, (c) second cluster, (d) third cluster, and (e) fourth cluster, respectively, and (f) single grayscale image colored based on their cluster index. 13
5/12/2014
Image Segmentation (Cont …)
Image segmentation results (a) Images before segmentation, (b) Images after segmentation.
14
5/12/2014
Feature Extraction The features used for the apple fruit disease classification problem are –
1. 2. 3. 4.
15
Global Color Histogram (GCH) [15], Color Coherence Vector (CCV) [16], Local Binary Pattern (LBP) [8], and Complete Local Binary Pattern (CLBP) [13].
5/12/2014
Training and Classification Recently, a unified approach is presented in [17] that can combine many features and classifiers. The author approaches the multi-class classification problem as a set of binary classification problem. For N-class problem N (N-1)/2 binary classifiers will be needed, where N is the number of different classes.
16
5/12/2014
Training and Classification (Cont …)
Multi-class Support Vector Machine (MSVM) as a set of binary Support Vector Machines (SVMs) is used for the training and classification.
17
5/12/2014
Experimental Result
18
5/12/2014
Data Set Preparation
1. 2. 3. 4.
19
To demonstrate the performance of the proposed approach, a data set of normal and diseased apples are used, which comprises four different categories: Apple Blotch (104), Apple rot (107), Apple scab (100), and Normal Apple (120): totalizing 431 apple fruit images. 5/12/2014
Data Set Preparation (Cont …)
Sample images from the data set 20
5/12/2014
Result Discussion If we use N images per class for training then remaining images are used for testing. The accuracy of the proposed approach is defined as,
Total number of images correctly classified Accuracy (%) = * 100 Total number of images used for testing
21
5/12/2014
Result Discussion (Cont …)
Accuracy (%) for the GCH, CCV, LBP, and CLBP features derived from RGB and HSV color images considering MSVM classifier
22
5/12/2014
Result Discussion (Cont …) For instance, in HSV color space with 50 training examples per class, the reported classification accuracy is – 80.94%
for 86.47% for 90.97% for 93.14% for
23
5/12/2014
GCH, CCV, LBP, and CLBP feature.
Result Discussion (Cont …)
Comparison of the accuracy achieved in RGB and HSV color space for the GCH, CCV, LBP, and CLBP features considering MSVM classifier
24
5/12/2014
Result Discussion (Cont …)
25
One important aspect when dealing with apple fruit disease classification is the accuracy per disease.
This information points out the diseases that need more attention when solving the confusions.
5/12/2014
Result Discussion (Cont …)
Accuracy per class for the CLBP feature in RGB and HSV color spaces using MSVM as a classifier
26
5/12/2014
Result Discussion (Cont …) For CLBP feature in HSV color space, for instance, reported classification accuracy are –
27
89.88%, 90.71%, 96.66%, and 99.33% for the Apple Blotch, Apple Rot, Apple Scab, and Normal Apple respectively,
resulting average accuracy 93.14% when training is done with 50 images per class. 5/12/2014
Conclusion
28
5/12/2014
Conclusion An image processing based solution is proposed for the detection and classification of apple fruit diseases. The proposed approach is composed of mainly three steps: image segmentation, feature extraction, and training and classification. Three types of apple diseases namely: Apple Blotch, Apple Rot, and Apple Scab are taken as a case study.
29
5/12/2014
Conclusion (Cont …) The proposed solution can significantly support automatic detection of fruit diseases. Normal apples are easily distinguishable with the diseased apples. CLBP feature shows more accurate result for the classification of apple fruit diseases. Further work includes consideration of fusion of more than one feature to improve the output of the proposed method.
30
5/12/2014
References
31
5/12/2014
References 1)
2)
3)
4)
5)
6)
J. Hartman, ―Apple Fruit Diseases Appearing at Harvest‖, Plant Pathology Fact Sheet, College of Agriculture, University of Kentucky, http://www.ca.uky.edu/agcollege/plantpathology/ext_files/PPFShtml/PPFS-FRT-2.pdf, viewed on December 2011. Q. Li, M. Wang, and W. Gu, ―Computer vision based system for apple surface defect detection,‖ Computers and Electronics in Agriculture, vol. 36, pp. 215– 223, Nov. 2002. P. M. Mehl, K. Chao, M. Kim, and Y. R. Chen, ―Detection of defects on selected apple cultivars using hyperspectral and multispectral image analysis,‖ Applied Engineering in Agriculture, vol. 18, pp. 219–226, 2002. M. S. Kim, A. M. Lefcourt, Y. R. Chen, and Y. Tao, ―Automated detection of fecal contamination of apples based on multispectral fluorescence image fusion,‖ Journal of food engineering, vol. 71, pp. 85–91, 2005. O. Kleynen, V. Leemans, and M. F. Destain, ―Development of a multi-spectral vision system for the detection of defects on apples,‖ Journal of Food Engineering, vol. 69, pp. 41–49, 2005. V. Leemans, H. Magein, and M. F. Destain, ―Defect segmentation on ‗jonagold‘ apples using colour vision and a bayesian classification method,‖ Computers and Electronics in Agriculture, vol. 23, pp. 43–53, June 1999. 32
5/12/2014
References (Cont …) 7) V. Leemans, H. Magein, and M. F. Destain, ―Defect segmentation on ‗golden
delicious‘ apples by using colour machine vision,‖ Computers and Electronics in Agriculture, vol. 20, pp. 117–130, July 1998. 8) T. Ojala, M. Pietikäinen, and T. T. Mäenpää, ―Multiresolution gray-scale and rotation invariant texture classification with Local Binary Pattern,‖ IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 24, no. 7, pp. 971-987, 2002. 9) T. Ojala, T. Mäenpää, M. Pietikäinen, J. Viertola, J. Kyllönen, and S. Huovinen, ―Outex – new framework for empirical evaluation of texture analysis algorithm,‖ in Proc. International Conference on Pattern Recognition, 2002, pp. 701-706. 10) T. Ahonen, A. Hadid, and M. Pietikäinen, ―Face recognition with Local Binary Patterns: application to face recognition,‖ IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 28, no. 12, pp. 2037-2041, 2006. 11) G. Zhao, and M. Pietikäinen, ―Dynamic texture recognition using Local Binary Patterns with an application to facial expressions,‖ IEEE Trans. On Pattern Analysis and Machine Intelligence, vol. 27, no. 6, pp. 915-928, 2007.
33
5/12/2014
References (Cont …) 12)
13)
14)
15)
16) 17)
X. Huang, S. Z. Li, and Y. Wang, ―Shape localization based on statistical method using extended local binary pattern,‖ in Proc. International Conference on Image and Graphics, 2004, pp.184-187. Z. Guo, L. Zhang, and D. Zhang, ―A completed modeling of local binary pattern operator for texture classification,‖ IEEE Trans. On Image Processing, vol. 19, no. 6, pp. 1657-1663, 2010. J. MacQueen, ―Some methods for classification and analysis of multivariate observations,‖ In L. M. LeCam and J. Neyman, editors, Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability, 1967, vol. 1, pp. 281—297, University of California Press. R. Gonzalez, R. Woods, Digital Image Processing, 3rd ed., Prentice-Hall, 2007. G. Pass, R. Zabih and J. Miller, ―Comparing images using color coherence vectors,‖ In ACM Multimedia, 1997, pp. 1–14. A. Rocha, C. Hauagge, J. Wainer, and D. Siome, ―Automatic fruit and vegetable classification from images,‖ Computers and Electronics in Agriculture, Elsevier; vol. 70, pp. 96-104, 2010.
34
5/12/2014
Thank You
35
5/12/2014