A new method for accurate, high-throughput volume estimation from three 2D projective images

摘要

In this article, a new method for estimating the volume of selected fruits and vegetables, using information obtained from three orthogonal images of the object is proposed. The method was based on a novel lower bound on this volume, combined with an application of a well-known upper bound. The lower bound was obtained by calculating the volume of an artificial object created from the intersection of the three object silhouettes. The method was applied to a significant range of food products and simple geometric objects, and it was found that the method estimated volume with approximately 11.9% error, generally, and 2%, for rounded objects, specifically.