A vision-based method to estimate volume and mass of fruit/vegetable: Case study of sweet potato

Among physical attributes, dimensions and weight of agricultural products are important parameters that useful for designing grading/packaging systems. Size measurements usually require extensive time labor. Moreover, it is often difficult to accurately do the having irregular shapes, such as sweet potatoes. Even though there many previous works on topic, they either lack accuracy or multiple captures computing power reconstruct three-dimensional representation products. This paper proposes measurement potato features, including body length, width, thickness, by using a simple vision system requires only single camera capture top-view image After background segmentation step, product virtually sliced into equal slices along longitudinal axis. The volume calculated summation volumes these individual slices. Since high correlation between products’ volume, can be calculated. experimental results obtained show proposed approach accomplished with highly competitive results, which have an up 96% (with R2 0.98) estimation task 95% 0.96) task. Due its simplicity, this model used design develop sizing/weighing/packaging