Year: 2025
Type: article
Abstract: For small to medium-sized fruits, such as cherries and blueberries, phenotypic data such as fruit size and various diameters are key indicators for fruit breeding, providing essential references for the breeding process. Currently, manual measurements that are widely used are inefficient and introduce significant uncertainties. To address this, this study designed breeding equipment that leveraged machine vision technology to capture fruit images and perform size measurements and evaluations using LabVIEW programming. Using cherries as an example, the RGB images were converted to grayscale using a weighted average method, followed by flat-field correction to remove artifacts and noise. The Canny algorithm was then applied to extract the fruit contour, and the Otsu algorithm was employed for automatic thresholding in the image binarization. Morphological processing separated the fruit from the stem, and template-matching technology was used to define the region of interest (ROI) for measurement. The system achieved an average relative error of only 0.35%, with a standard deviation of 0.0828 and a measurement time of 2.547 s per cherry. This method significantly improved the efficiency and saved time compared to traditional manual measurements. The device successfully measured fruits such as cherries and blueberries, providing a new tool for fruit size measurement in the breeding process and offering valuable insights for the design of future equipment.
Source: Horticulturae
Authors: Runqiang Xu, Yuhua Li, Zhilong Zhang, Jiawei Wang, Qingzhong Liu, Dongzi Zhu
Institution: Shandong Agricultural University
Language: English
Topic: Smart Agriculture and AI
Subfield: Plant Science
Domain: Life Sciences
SDG: Zero hunger
Open Access status: gold
Estimated APC paid: $2,654
Award: G2782015960
OpenAlex ID: W4415991691