An Automatic Egg Quality Grading Using Nature-Inspired Algorithm Based Classification

Human error may happen at egg quality grading manually by farmer. This condition may lead a miss to differ an egg with good quality and an egg with bad quality. It also impact for human body health because protein from egg can not be absorbed optimally in human body. To tackle this problem, an automatic system for egg quality grading is performed. However, the development system may need a preliminary study which exploit a nature-isnpired algorithm such ant colony optimization (ACO) as logic design for the system. In this paper, ACO based classification is performed to make a group in features of testing data which is similar with training's data. Some methods such k-nearest neighbour, support vector machine, naïve bayes, and random forest are exploited to against ACO based classification. Although the ACO based classification result may not be the best, it is still a compromising algorithm for this system because the accuracy is above 90% although some modifications and studies must be performed in the next activity including to tackle the algorithm running-time which costs 40.03 seconds to finish the task and it is the longest time among other methods.