The main objective is to detect fruit's quality and grade them depending on the size & its quality. We have used Neural Networks Algos with K-Clustering technique to generate different cluster images.
We have real time images of different fruits(Spherical ones) like Apple, Orange, Lemon etc. We had trained some images as trained datasets i.e, "Good" and "Bad". Whenever a test image was fed into the program, it would notify us with Good or Bad fruit using GUI dialog box.
The System developed using MATLAB and some hardware Conveyor and achieved efficiency of over 88%. Application: Fruit and Vegetable related processing industries, Government Silos, Hypermarkets etc.