Detection and classification of hazelnut fruit by using image processing techniques and clustering methods

In this study, theobjects found in the environment are detected and classified in real time, theresults obtained are presented. Hazelnut fruit is used in the experimentalstudies of the proposed method. The image belongs to hazelnut that is in a workenvironment is taken with the camera, it is processed by using image processingtechniques. The size and area data of hazelnut on the image plane iscalculated. By evaluating the obtained data, the hazelnut is divided into threeclasses as small (K1), medium (K2) and big (K3) in real time application. Thisprocess is performed using mean-based classification and K-means clusteringmethods. Detection and classification of cluster centers is provided by usingthe information database obtained from the data of hazelnut fruit. Hazelnutfruits found in the experimental environment are determined with 100% accuracyusing image processing techniques. The classification of hazelnut fruits usingthe mean-based and K-means clustering methods has been compared. As a result ofthe comparison, it is observed that the two methods realized are similar ratioof 90% to 100%.