MATLAB-Based Precision Image Analysis for Fruit Sorting

Fruits are extremely fundamental in our everyday diet. The palatable fruits are harvested, sorted, and packed for conveyance to the consumers. It needs a large number of expert resources and a long time to sort and grade the fruits from the agricultural field to the fruit markets. Automation in the agricultural field and fruit markets is a must to reduce the time as well as the dependency on the manual resource. Thus, the objective of this project is to fully automate the sorting process of handling fruits. The proposed method used an image acquisition system (camera), which acquires the images of the various selected fruits (Apple, Onion, Banana, Pepper and Tomato) used for the training data. The textural and colour features of the selected fruits were extracted and then, processed using the MATLAB software with Support vector machine (SVM) algorithm as the classifier. The fruit recognition system classified the input fruit sample by determining the similarities between the colour and gray level co-occurrence matrix values of the inputted fruits samples and the values obtained from the training datasets. The proposed method is accurate and flexible. Also, a graphical user interface was developed to be used independently of the software, the recognition rate of the system had an average accuracy of 97%.