Document Type : Research Paper
A new method based upon Gabor filter and machine vision for the recognition of pistachio varieties is proposed throughout the current study. In the suggested method, the image of a set number of pistachios is considered to represent a texture; instead of one by one pistachio processing (which is actually done in the current methods) the new method is expected to represent a higher rating as well as a higher performance. To evaluate the proposed method, it was applied on an image of a set of pistachios' containing 1000 sub-images of 5 varieties of the fruit and using a K-means clustering to classify the product. The experimental results confirm the efficiency of the method by a classification rating of 94.8%.