Abstract
Sorting apples by product category is a key stage of harvest preparation, especially for small and medium-sized farms, where traditional manual sorting leads to high labor intensity. This article presents the development and experimental study of a compact apple-sorting machine based on computer vision, color assessment, and indirect fruit-weight estimation. The designed machine was adapted for Kazakhstani apple varieties and integrates low-cost parts: a fruit feeding and positioning module, a computer vision system with an image processing unit, PLC-based control, and a sorting actuator. The decision-rule procedure uses color and visual geometric parameters of images, and classification by regression analysis. Statistical analysis identified projected fruit area as the primary predictor for indirect weight estimation using the proposed regression model. The model robustness and classification accuracy were assessed by confusion matrices and main validation metrics. The influence of conveyor speed on the stability of sorting was observed. The experimental results indicated that the optimal operating mode for the machine is an apple transport speed of 0.16 m/s, where sorting throughput is approximately 400 kg/hour, with an average accuracy of 92%. The proposed machine provides sufficient real-time sorting accuracy with a simple, cost-effective machine design and can be used as a useful sorting solution in small and medium-sized farms.