TrichomeYOLO: A Neural Network for Automatic Maize Trichome Counting

Jie Xu; Jia Yao; Hang Zhai; Qimeng Li; Qi Xu; Ying Xiang; Yaxi Liu; Tianhong Liu; Huili Ma; Yan Mao; Fengkai Wu; Qingjun Wang; Xuanjun Feng; Jiong Mu; Yanli Lu

doi:10.34133/plantphenomics.0024

TrichomeYOLO: A Neural Network for Automatic Maize Trichome Counting

Plant Phenomics. 2023:5:0024. doi: 10.34133/plantphenomics.0024. Epub 2023 Feb 28.

Authors

Jie Xu^{1

2}, Jia Yao³, Hang Zhai^{1

2}, Qimeng Li^{1

2}, Qi Xu^{1

2}, Ying Xiang³, Yaxi Liu⁴, Tianhong Liu^{1

2}, Huili Ma^{1

2}, Yan Mao⁵, Fengkai Wu^{1

2}, Qingjun Wang^{1

2}, Xuanjun Feng^{1

2}, Jiong Mu³, Yanli Lu^{1

2}

Affiliations

¹ Maize Research Institute, Sichuan Agricultural University, Wenjiang 611130, Sichuan, China.
² State Key Laboratory of Crop Gene Exploration and Utilization in Southwest China, Sichuan Agricultural University, Wenjiang 611130, Sichuan, China.
³ College of Information Engineering, Sichuan Agricultural University, Yaan 625014, Sichuan, China.
⁴ Triticeae Research Institute, Sichuan Agricultural University, Wenjiang 611130, Sichuan, China.
⁵ College of Chemistry and Life Sciences, Chengdu Normal University, Wenjiang 611130, Sichuan, China.

Abstract

Plant trichomes are epidermal structures with a wide variety of functions in plant development and stress responses. Although the functional importance of trichomes has been realized, the tedious and time-consuming manual phenotyping process greatly limits the research progress of trichome gene cloning. Currently, there are no fully automated methods for identifying maize trichomes. We introduce TrichomeYOLO, an automated trichome counting and measuring method that uses a deep convolutional neural network, to identify the density and length of maize trichomes from scanning electron microscopy images. Our network achieved 92.1% identification accuracy on scanning electron microscopy micrographs of maize leaves, which is much better performed than the other 5 currently mainstream object detection models, Faster R-CNN, YOLOv3, YOLOv5, DETR, and Cascade R-CNN. We applied TrichomeYOLO to investigate trichome variations in a natural population of maize and achieved robust trichome identification. Our method and the pretrained model are open access in Github (https://github.com/yaober/trichomecounter). We believe TrichomeYOLO will help make efficient trichome identification and help facilitate researches on maize trichomes.