Authors : Archana Chaugule and S.N. Mali
Abstract: This research proposes an algorithm to implement feature extraction technique using discrete wavelet-GLCM-Tamura and angle and use the extracted features to represent the image for classification of seeds. A total of 69 discrete wavelet-GLCM-Tamura and 12 angle features were extracted from the high-resolution images of paddy seeds. These features were employed along with ANN to identify paddy varieties. These researches is aimed at comparing discrete wavelet-GLCM-Tamura and angle features using ANN for discriminating Indian paddy varieties and also evaluate variety-wise classification of individual grains. The classification of four paddy (rice) grains, viz. Karjat-6 (K6) and Ratnagiri-2 (R2), Ratnagiri-4 (R4) and Ratnagiri-24 (R24) was done and the features were evaluated in terms of accuracy. From the entire feature models, most suitable feature was identified for accurate classification. Angle features gave the best classification using ANN among both the feature sets.
Archana Chaugule and S.N. Mali, 2016. Analysis of DWT-GLCM-Tamura and Angle Features for Variety Identification of Seeds. Asian Journal of Information Technology, 15: 3238-3246.