Apple disease classification using color, texture and shape features from images. Issue 5 (July 2016)

• Record Type: Journal Article
• Authors: Dubey, Shiv; Jalal, Anand
• Is Part Of: Signal, image and video processing. Volume 10:Issue 5(2016)
• Page Start: 819
• Page End: 826
• Publication Date: 2016-07
• Subjects: K-Means clustering -- Color -- Texture -- Shape -- LBP -- SVM -- Feature fusion; Signal processing -- Digital techniques -- Periodicals; Image processing -- Digital techniques -- Periodicals; Digital video -- Periodicals; 621.3822
• DOI: 10.1007/s11760-015-0821-1
• Languages: English
• ISSNs: 1863-1703

Abstract

Abstract The presence of diseases in several kinds of fruits is the major factor of production and the economic degradation of the agricultural industry worldwide. An approach for the apple disease classification using color-, texture- and shape-based features is investigated and experimentally verified in this paper. The primary steps of the introduced image processing-based method are as follows: (1) infected fruit part detection is done with the help of K-means clustering method, (2) color-, texture- and shape-based features are computed over the segmented image and combined to form the single descriptor, and (3) multi-class support vector machine is used to classify the apples into one of the infected or healthy categories. Apple fruit is taken as the test case in this study with three categories of diseases, namely blotch, rot and scab as well as healthy apples. The experimentation points out that the introduced method is better as compared to the individual features. It also points out that shape feature is not better suited for this purpose.