<![CDATA[Beef is one of the commodities that has contributed to the improvement of public nutrition, particularly the need for animal protein. Fresh beef is meat that is fresh red in color, starting from being cut up to 10 hours. So far, evaluation of freshness and identification of meat composition has been done manually by means of human visual observations. Due to human limitations, there are often different perceptions of each observer. On this basis, as an effort to obtain beef freshness accurately, this research has made a tool that can detect the freshness of beef with the help of digital image computing. By using the Raspberry Pi as a mini computer, a camera as a sensor and image processing which is then classified by Naive Bayes, this system can work properly, it can be proven by the output of the accurate classification of beef freshness. The choice of the naive Bayes method is based on the fact that this method is a very good classification method in which the class of freshness types is known from the start. This method can also work even though it only uses a little training data. When there is a slight change in training data, the naive Bayes method also adapts quite well. The results of the beef color conversion process are then classified at the color level based on SNI standards. From 40 training data and 20 tested data, an accuracy of 95% and an average computation rate of 0.009094 seconds.]]>
Copyrights © 2021
