Liangquan Jia ^1* Tao Wang ¹ Xiangge Li ^1* Lu Gao ^1* Qiangguo Yu ^2* Xincheng Zhang ^3* Shanlin Ma ^3*

• ¹ School of Information Engineering, Huzhou University, Huzhou, China
• ² Huzhou University, Huzhou, China
• ³ Institute of Crop Science,Huzhou Academy of Agriculture Sciences, Huzhou 313000, China, Huzhou, China

With the rapid advancement of plant phenotyping research, understanding plant genetic information and growth trends has become crucial. Specifically, for the growth status assessment of plant seedlings, measuring seedling length has emerged as an indispensable tool and a core criterion for assessing seed viability. Traditional ruler-based and manual analysis methods are evidently timeconsuming and labor-intensive. Therefore, we propose an efficient and versatile deep learning approach aimed at enhancing the efficiency of plant seedling phenotyping analysis and addressing issues associated with traditional algorithms. We have improved the DeepLabv3+ model, naming it DFMA, and introduced a novel ASPP structure, called PSPA-ASPP. In our self-constructed rice seedling dataset, our segmentation results achieved a mean Intersection over Union (mIoU) of 81.72%. On publicly available datasets, the detection results for three different species were notably superior to other models, with scores of 87.69%, 91.07%, and 66.44%, respectively. Our segmentation model can generate plant segmentation masks, clearly depicting developmental details of various seed components, such as the embryonic shoot, embryonic axis, and embryonic root. Additionally, we have introduced a seedling length measurement algorithm, providing precise parameters for the development of different components, thereby enabling comprehensive plant phenotyping analysis.With the rapid advancement of plant phenotyping research, understanding plant genetic information and growth trends has become crucial. Measuring seedling length is a key criterion for assessing seed viability, but traditional ruler-based methods are time-consuming and labor-intensive. To address these limitations, we propose an efficient deep learning approach to enhance plant seedling phenotyping analysis. We improved the DeepLabv3+ model, naming it DFMA, and introduced a novel ASPP structure, PSPA-ASPP. On our self-constructed rice seedling dataset, the model achieved a mean Intersection over Union (mIoU) of 81.72%. On publicly available datasets, including Arabidopsis thaliana, Brachypodium distachyon, and Sinapis alba, detection scores reached 87.69%, 91.07%, and 66.44%, respectively, outperforming existing models. The model generates detailed segmentation masks, capturing structures such as the embryonic shoot, axis, and root, while a seedling length measurement algorithm provides precise parameters for component development. This approach offers a comprehensive, automated solution, improving phenotyping analysis efficiency and addressing the challenges of traditional methods.