Highly efficient fruit mass and size estimation using only top view images

This paper presents a new methodology for the estimation of mass and size of high-export-volume Vietnamese fruits (such as cavendish-type bananas and cucumbers), characterized by only its top-view images using image analysis techniques. Most previous works focused on volume estimation using a plurality of cameras in order to infer the three-dimensional information. In this work, we only use a single camera mounted on top of the fruit. A machine learning technique is also used to find the most essential features. We have found that our proposal lead to a relatively small estimation error compared to the results obtained from the measurements using a water-replacement method and a static digital scale. The results indicate that our system shows a great potential to be used in a real industrial setting. Future work will aim to investigate other features such as ripeness and bruises to increase the effectiveness and practicality of the system.