Scholarlink

<p>There is a growing interest in the development of techniques that obtain automatic measurements of morphological characteristics of plant leaves, due to the diverse possibilities of its applications. Therefore the automatic separation between blade and petiole still consists in a challenge in this process, especially because leaves exhibit expressively intraespecific variability. Hence, the objective of the present study was to present two algorithms for the automatic separation between the blade and petiole based on the leaf contours. The two proposed algorithms were compared to other from the literature and modifications were presented, also it were implemented in the R software. For the performance evaluation of the algorithms, a set of leaves was used, in which manually marked insertion points were compared to the estimates of the algorithms. The performance of the algorithms depending on the species. However, three variations of the algorithms showed better performance on the analyzed leaves (two new ones and one modified from the literature). All algorithms were implemented in the R package <Emphasis FontCategory="NonProportional">petioleFborder</Emphasis>, which is presented and made available in this article. These contributions will collaborate to a low cost, open access and automatic acquisition of leaf measurements, which can be used in various applications.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

petioleFborder: a new R package to automatically detect leaf petioles by using edge-based methodology

  • 期刊论文

  • 发表日期：

    2026年6月18日

查看全文
  • G. F. Paludo,
  • A. M. Nhancololo,
  • J. S. S. Bueno-Filho

摘要

There is a growing interest in the development of techniques that obtain automatic measurements of morphological characteristics of plant leaves, due to the diverse possibilities of its applications. Therefore the automatic separation between blade and petiole still consists in a challenge in this process, especially because leaves exhibit expressively intraespecific variability. Hence, the objective of the present study was to present two algorithms for the automatic separation between the blade and petiole based on the leaf contours. The two proposed algorithms were compared to other from the literature and modifications were presented, also it were implemented in the R software. For the performance evaluation of the algorithms, a set of leaves was used, in which manually marked insertion points were compared to the estimates of the algorithms. The performance of the algorithms depending on the species. However, three variations of the algorithms showed better performance on the analyzed leaves (two new ones and one modified from the literature). All algorithms were implemented in the R package petioleFborder, which is presented and made available in this article. These contributions will collaborate to a low cost, open access and automatic acquisition of leaf measurements, which can be used in various applications.