Algoritmo Para Estimación De Tamaño De Fresas A Partir De Imágenes De Mapas De Profundidad En Toma Frontal

Abstract

This research delves into developing techniques for estimating strawberry measurements, focusing on an estimation algorithm that calculates dimensions based on proximity to the fruit. Additionally, the study highlights the importance of non-destructive evaluation methods in strawberry cultivation, emphasizing the need for systems like RealSense D435i cameras to capture fruit characteristics accurately and efficiently, reducing manual labor and potential fruit damage during assessment processes. Despite the algorithm's specific calculations, it encounters limitations, notably in program fluidity due to the high resource consumption of RealSense camera data processing. The study concludes that changing the image resolution affects data estimation, with higher resolutions leading to more pixels and a corresponding impact on depth data maps. Additionally, theoretical approaches face limitations, particularly in excluding volume values due to assuming ideal figures, overlooking deformities and variations in strawberry shapes. It's advised to employ methods like Archimedes' principle for accurate volume determination in non-homogeneous elements.