Fruits and Vegetables Classification using Progressive Resizing and Transfer Learning


Kishore M,  Associate Professor, S. B. Kulkarni, K. Senthil Babu
Department of CSE, SDMCET, Dharwad, KS School of Engineering and Management, Karnataka, India.


The customer satisfaction is the key to succeeding business. In order to work on this, time consumption is the most important factor. Currently, the method of segregation of fruits and vegetables based on category and grade is done manually. Though implemented, this method is time consuming, costly, inefficient, and a considerably difficult task. In places where food is prepared in bulk, the quality of the fruits and vegetables is often overlooked, and the overall quality of the food consumed is reduced. In order to overcome this drawback, computer vision has been introduced and implemented to segregate the products based on features like color, size, quantity, and so on. We are proposing an automatic and an effective method of evaluation for fruits and vegetables using Machine Learning techniques. The algorithm used here does not require human intervention, and the system has higher accuracy compared to the human-involved systems because it uses an automated computer algorithm. The fruits and vegetables classifier is efficient, non-destructive, and accurate which reduces man power. The existing classifier can be used in various applications such as user friendly, a self-checkout system for the visually impaired customers in supermarkets, sorting of products before releasing it to the markets.