Weight and volume estimation of single and occluded tomatoes using machine vision

The fundamental characteristics of agricultural products are appearance, size, and weight, which affect their market value, consumer preference, choice. Thus, food industries seek rapid, simple, nondestructive approaches to assess real-time measurements at the post-harvest stage before packaging for market. While sorting grading may be performed by humans, it is unreliable, time-consuming, complicated, subjective, onerous, expensive, easily influenced surroundings. Therefore, an astute method tomato fruit required. We evaluated two configurations on a conveyor belt: single tomatoes (no occlusion) multi-tomatoes (partially occluded). used polygon approximation concave convex point extraction algorithms segment occluded tomatoes. developed seven models regression using single-tomato image features. Bayesian regularization artificial neural network outranked all trained in weight estimation with root-mean-square error (RMSE) 1.468 g R2 0.971. For volume estimation, RBF SVM had best performance 0.982 RMSE 1.2683 cm3. It feasible implement proposed system as noninvasive in-line technique