Click on “Download PDF” for the PDF version or on the title for the HTML version.

If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.

Abstract. Phenotyping in agriculture is a technique to discover genetic traits that are desired or expressed. A key phenotype is seed volume. Phenotypes are morphometric parameters of agricultural entities such as seeds, roots and leaves predict their behavior in different environmental conditions. While volume estimation is a well-studied problem for regular objects, the same cannot be said of irregular objects. Seeds are typically irregular-shaped, and the task of seed volume estimation while preserving the seed is a complex task due to several factors. To evaluate the accuracy of high-throughput volume estimation systems, we need to have the accurate size and volume measurements of test seeds. This paper proposes a novel, end-to-end automated cyber-physical low-throughput system that estimates the volume of a seed kernel using the technique of silhouette-based volume sculpting (3D carving). A high degree of similarity is observed between the estimates obtained by the proposed system and ellipsoid-like slicing techniques such as mirrored image slicing and two side-views image slicing demonstrating the feasibility of the proposed system. The proposed system is more efficient in terms of producing results with less variability in comparison to the existing techniques. An experiment conducted on a sample of five wheat seed kernels and five soy seed kernels independently for a total of twenty times yields less variability when using the proposed silhouette-based volume sculpting system as opposed to the ellipsoid slicing technique.