Comparison of the performance ACS with human expert is presented in Table 2 in the form of confusion matrix. As shown, the developed ACS had 89.2% accuracy in determining the desired working conditions for the whitening machine. The accuracy of the image processing unit in determination of DOM and PBK indices was equal to 92.6% and 97.7%, respectively. In recent years, several researchers have worked on grading of agricultural products based on their quality using intelligent systems. Alavi (2013) developed a fuzzy-logic-based model to grade dates using Mamdani fuzzy inference system. The input parameters for the system were the information obtained from image processing in the form of two indices: length and freshness. They reported the accuracy of the system to be 91%. A similar system was developed by Dehrouyeh et al. (2010) for egg classification based on damages. The results showed that the accuracy of the model in identification of clean, low dirt and high dirt eggs was equal to 86%, 83%, 88%, respectively.