CLASSIFICATION OF BARLEY SEEDS BY USING IMAGE PROCESSING TECHNIQUES

Abstract

In this research work, a digital image analysis technique was used to classify malt barley and food barley based on their color, morphological and textural features. From each sample, 50 images were captured. Three hundred images were captured from samples collected from Srinka Agricultural Research Center (SARC). Artificial neural network was used to classify barley seeds and to cheek weather the samples were correctly classifies to their origin of collection or not. For the analysis, 12 color, 6morphological and 12 textural features, totally 30 features were extracted from images of barley samples. Three experimental classification setups were applied for classification designs by using morphology, color and texture. For all setups 300images were used as inputs of the machine. From these datasets, 240 images were used for training the machine while the remaining 60(20%) images were used for testing and validation. The accuracy of classification using morphological, texture and color feature from for all confusion matrixes was 83.3%. The seeds percentage of samples, which were not correctly, classified were 16.7% for color, texture and morphological features