Hieu M. Tran, Tuan M. Le, Hung Son Nguyen, Son V. T. Dao
The chapter explores a method of determining the mass and volume of the Vietnamese avocado 034 using the combinations of top-view images as well as the disc method of geometric slice modeling. The method studies the geometric characteristics, length, width, and height of the top-view image that is obtained by an image processing method. These characteristics are then incorporated in the disc method that cuts the fruit into 5–15 discs to estimate its mass. The strategy is intended to enhance agriculture and after-harvest processing by offering an alternative to the traditional weighing process, which is practical and scalable. The method was tested by using a dataset of 101 avocado samples under varying slicing configurations. The experimental outcomes prove that the proposed method has a correlation coefficient of about 97.72%, root mean square percentage error (RMSPE) of 4.26%, and mean absolute percentage error (MAPE) of about 3.48%, which shows high accuracy and reliability of the method in estimating mass. Also, the time per image is approximately 155 ms, which allows processing to be effective and in real time. This nondestructive, low cost, and scalable methodology would be suitable in agricultural and postharvest operations such as automated sorting and grading of fruits.