Fruit and Vegetable Sorting Using Machine Learning and Parametric Sorter

The manual harvesting and sorting of fruits and vegetables in small quantities is a simple process, but when you want this same process in large quantities it is already a tedious and difficult process for humans to do, so an automation process is required, therefore an automation process is required. In such process it is required from starting with the recognition of the objects present in an image, the process of feature extraction, as well as the method of classification of the same objects and how to calculate the efficiency of the classification process and then culminate in a technological application. In this case of the automation and machine learning process, the following is proposed feature extraction based on color and texture statistics, using a parametric k-nearest neighbor (KNN) classifier. The three ways to evaluate the efficiency of this classifier will be by the method of: The re-substitution method, The Cross validation method, The Leave one out method. We will also rely apart from these three methods on the Confusion Matrix to measure the efficiency. This approach is based on Content-Based Image Retrieval (CBIR), which in this particular case will be applied to fruits and vegetables.