Leaf Segmentation Using Modified YOLOv8-Seg Models

Peng Wang,Hong Deng,Jiaxu Guo,Siqi Ji,Dan Meng,Jun Bao,Peng Zuo

DOI: https://doi.org/10.3390/life14060780

2024-06-21

Life

Abstract:Computer-vision-based plant leaf segmentation technology is of great significance for plant classification, monitoring of plant growth, precision agriculture, and other scientific research. In this paper, the YOLOv8-seg model was used for the automated segmentation of individual leaves in images. In order to improve the segmentation performance, we further introduced a Ghost module and a Bidirectional Feature Pyramid Network (BiFPN) module into the standard Yolov8 model and proposed two modified versions. The Ghost module can generate several intrinsic feature maps with cheap transformation operations, and the BiFPN module can fuse multi-scale features to improve the segmentation performance of small leaves. The experiment results show that Yolov8 performs well in the leaf segmentation task, and the Ghost module and BiFPN module can further improve the performance. Our proposed approach achieves a 86.4% leaf segmentation score (best Dice) over all five test datasets of the Computer Vision Problems in Plant Phenotyping (CVPPP) Leaf Segmentation Challenge, outperforming other reported approaches.

biology

What problem does this paper attempt to address?

This paper focuses on the problem of plant leaf instance segmentation, which is an important technique in the field of plant phenotyping, growth monitoring, and precision agriculture. The authors used the YOLOv8-seg model and improved it by introducing the Ghost module and the BiFPN (Bidirectional Feature Pyramid Network) module to enhance the segmentation performance for small leaves. The experimental results show that the improved model achieved the best Dice score of 86.4% on all five testing datasets of the CVPPP Leaf Segmentation Challenge, outperforming other reported methods. In the study, the authors compared the differences between isolating leaf segmentation and identification, as well as live plant leaf segmentation, and they provided detailed descriptions of the datasets and methods used, including data preprocessing, model training, and validation procedures. By integrating data augmentation and specific modules, the segmentation performance of the model was improved.