Identificación de defectos de calidad en frutos de aguacate ‘Hass’ mediante procesamiento de imágenes
Year: 2025
Type: article
Abstract: Mexico is the main producer of 'Hass' avocado (Persea Americana Mill.) worldwide. However, small producers often present quality problems and do not meet the standards of the NMX-FF-016-2002 standard. Therefore, the objective of the research was to identify quality defects in fruits through image processing. To do this, in September 2023, fruits at physiological maturity were harvested in various orchards in the Valle de Bravo region, Mexico. The diameter, length and weight of each avocado were measured. The evaluation of quality defects was carried out by an expert and, simultaneously, photographs were taken to analyze the images using binarization and Canny algorithms to identify external physical damage and estimate the length, diameter and weight of each fruit. A t test was applied to compare the fruit size estimated by the algorithm and real data. As a result, significant differences (p < 0.01) were observed in length and diameter, therefore, an adjustment factor was applied to optimize the results of the algorithm. A multiple linear regression model was obtained (p < 0.01, r2 = 0.91) that estimates weight as a function of diameter and length. The binarization process helped identify damage from thrips, scrapes, and scab. While with the Canny method, damage from thrips and scab was identified. In conclusion, image analysis was efficient in identifying defects caused by scabs, thrips and scratches, as well as estimating fruit size.
Language: English
Topic: Smart Agriculture and AI
Subfield: Plant Science
Domain: Life Sciences
SDG: Zero hunger
Open Access status: diamond
OpenAlex ID: W4409323409