Wheat class identification using computer vision system and artificial neural networks
A. Arefi 1,   A. Modarres 1,   R. Farrokhi 1
More details
Hide details
Department of Farm Machinery, Urmia University, Urmia, Iran
Int. Agrophys. 2011, 25(4): 319–323
The identification of four wheat varieties was performed by integrating machine vision and artificial neural network (ANN) using Matlab software. It was based on grain morphology and colour. In order to capture images from the samples, a chamber of imaging was developed and a program was coded in Matlab for segmentation of the samples. Area and 4 factors for describing shapes of grain were chosen as morphology features. For colour features, average, variance, skewness and kurtosis values of images in RGB and l*a*b* colour spaces were extracted. Eleven features of the 280 images were used in the training stage of ANN, 40 images for validation, and testing of the ANN was performed with 80 images. The overall success classification rate was 95.86%.