<![CDATA[The automatic identification of fruit dimensions and morphology is a major requirement in agricultural and food technology. This research formulates a technique for fruit separation by employing mathematical morphology on binary images to identify the attributes of fruit shape and size. The dataset has 100 fruit photos featuring diverse geometric designs. The image processing procedure commences with grayscale conversion, followed by threshold adjustment, and culminates in the use of morphological operations to refine the image. The area and perimeter characteristics of the objects are extracted and utilized to categorize the fruit forms into circles or ellipses, triangles, and squares or rectangles. Segmentation accuracy was evaluated using the Intersection over Union (IoU) metric, yielding an average result of 85.67%, which signifies a high degree of segmentation precision. This study's findings affirm that mathematical morphology techniques are exceptionally effective in the automated detection of fruit size and form, with the potential to enhance efficiency in agricultural crop sorting processes.]]>
Copyrights © 2025
