Evaluation of different models for mechanical properties calculation of sour orange fruit under compression test using image processing

Several methods have been proposed to determine the engineering properties of agricultural products based on empirical knowledge or statistical relationships between properties. Therefore, the purpose of this study was to construct a system to measure stress and forces applied to sour orange as well as to design a graphical interface to quickly measure mechanical properties (Elasticity module, stress and Poisson's ratio in x, y and z directions) during loading using image processing. The effective contact area of the compressed fruit was determined by image analysis and compared in x, y and z directions using three webcams with other methods (capsule, fix volume, ASABE and bounding-box). According to the results, the contact area in ASABE method at 6% strain with 7% error had no significant difference with the image processing method. According to the results, the contact area in ASABE method at 6% strain with 7% error had no significant difference with the image processing method. The Poisson's ratio and the strength of the samples increased and decreased in the strain steps, respectively. The Poisson's ratio and the strength of the samples increased and decreased in the strain steps, respectively. There was no significant difference between ASABE and image processing methods in Ex and Ey at strain of 3% and 6%, respectively. There was no significant difference between ASABE and image processing methods in Ex and Ey at strain of 3% and 6%, respectively. In addition, there was no significant difference in Ez at 3% strain between capsule, image processing, and bounding-box methods. Moreover, there was no significant difference in the amount of stress at 3% strain between capsule, image processing, ASABE and bounding-box methods. Moreover, there was no significant difference in the amount of stress at 3% strain between capsule, image processing, ASABE and bounding-box methods. Among the mentioned methods, the ASABE method with the lowest error at 6% strain was the most appropriate method in comparison with image processing.