Home // ACHI 2023, The Sixteenth International Conference on Advances in Computer-Human Interactions // View article

Toward an Automated Pruning for Apple Trees Based on Computer Vision Techniques

Authors:
Keming Hu
Oky Dicky Ardiansyah Prima

Keywords: AI, semantic segmentation, thinning tree.

Abstract:
Effective pruning can contribute to the growth of plants. Similarly, pruning apple trees can help them absorb nutrients and grow stronger. However, Japan's apple farming industry is facing many challenges today, such as aging population, talent shortage, and reduced farmland area. Despite many studies attempting to solve the problems of aging population and talent shortage through automated pruning, but as a preliminary step for pruning, the process of identifying apple trees is too complex and difficult to achieve in real-world environments. In this paper, we propose a more straightforward apple trees recognition method based on computer vision to achieve pruning of apple trees in real environments. The method roughly consists of three steps: 1) segmenting apple trees through semantic segmentation, 2) skeletonizing the apple tree by segmentation image, 3) the representation of graph tree is done by applying breadth-first search. We tested 12 models for apple tree segmentation, and the Segfomer model achieved an accuracy of 76.72 and an intersection over union(IoU) of 64.29.

Pages: 50 to 54

Copyright: Copyright (c) IARIA, 2023

Publication date: April 24, 2023

Published in: conference

ISSN: 2308-4138

ISBN: 978-1-68558-078-0

Location: Venice, Italy

Dates: from April 24, 2023 to April 28, 2023