Xinjiang Agricultural Sciences ›› 2025, Vol. 62 ›› Issue (2): 261-269.DOI: 10.6048/j.issn.1001-4330.2025.02.001

• Crop Genetics and Breeding·Cultivation Physiology·Physiology and Biochemistry • Previous Articles     Next Articles

Salt tolerance in germination period of cotton seeds based on convolutional neural network and synthetic dataset

WANG Yongpan1,2(), MA Jun3, LI Chenyu1,2, YAO Mengyao1,2, WANG Zixuan1,2, HUANG Lingzhi1, ZHU Haiyan1, LIU Wanrong2, LI Bo2, YANG Yang2(), GAO Wenwei1()   

  1. 1. College of Agriculture, Xinjiang Agricultural University,Urumqi 830052, China
    2. Xinjiang Key Laboratory of Crop Biotechnology /Research Institute of Nuclear Technology and Biotechnology,Xinjiang Academy of Agricultural Sciences,Urumqi 830091,China
    3. Institute of Economic Crops, Xinjiang Academy of Agricultural Sciences, Urumqi 830091, China
  • Received:2024-08-28 Online:2025-02-20 Published:2025-04-17
  • Correspondence author: YANG Yang, GAO Wenwei
  • Supported by:
    Project of National Natural Science Foundation of China(32060497)

基于卷积神经网络和合成数据集训练鉴定棉花种子萌发期的耐盐性

王勇攀1,2(), 马君3, 李晨宇1,2, 姚梦瑶1,2, 王子轩1,2, 黄灵芝1, 朱海艳1, 刘皖蓉2, 李波2, 杨洋2(), 高文伟1()   

  1. 1.新疆农业大学农学院,乌鲁木齐 830091
    2.新疆农作物生物技术重点实验室/新疆农业科学院核技术生物技术研究所,乌鲁木齐 830091
    3.新疆农业科学院经济作物研究所,乌鲁木齐 830091
  • 通讯作者: 杨洋,高文伟
  • 作者简介:王勇攀(1998-),男,河南商丘人,硕士研究生,研究方向为棉花遗传育种,(E-mail)wyp620917@163.com
  • 基金资助:
    国家自然科学基金项目(32060497)

Abstract:

【Objective】To establish a convenient and accurate non-destructive detection method for cotton seed germination phenotypes, so as to characterize the salt tolerance of different cotton germplasm at the germination stage. 【Methods】A synthetic dataset was generated using 150 images of cotton seed germination at different stages and used to train the Mask R-CNN model. Using the trained model, we performed instance segmentation and feature extraction of seed shell and germ in real-world images of 60 cotton germplasm that germinated under 125 mmol/L NaCl treatment, and used them to infer the seed germination rate, germination potential, and germination length, so as to evaluate the salt tolerance of these 60 cotton germplasm in the germination stage. 【Results】The generated synthetic dataset contained 2,000 images and corresponding mask data. The accuracy of the Mask R-CNN model trained based on this dataset for the segmentation of seed shells and germs in real images was above 95 %, and the phenotypic values that inferred by model were highly consistent with them measured by manual operation (R2 > 0.98, P < 0.001), indicating that the phenotypes could be accurately obtained using the model. The cluster analysis of the salt tolerance index for each trait classified the 60 cotton materials into four levels; using the affiliation function method for a comprehensive evaluation of the salt tolerance of the cotton varieties. Kezimian 4 (0.95), MC-30 (0.88), and Lu8zao (0.81) had a larger D-value and indicated high salt tolerance. 【Conclusion】In this study, we have established a method for phenotyping cotton seed germination traits based on the convolutional neural network model that trained by using synthetic dataset. and using this method, we have identified the seed germination salt tolerance of 60 cotton germplasm in a non-destructive, rapid and accurate manner.

Key words: cotton; convolutional neural network; image segmentation; seed germination

摘要:

【目的】建立便捷且精准的棉花种子萌发表型的无损检测方法,鉴定不同棉花种质萌发期的耐盐性。【方法】利用150张不同阶段的棉花种子萌发图像生成合成数据集,并以此进行Mask R-CNN模型训练。利用训练好的模型,对60份棉花种子在125 mmol/L NaCl处理下萌发真实图像中的种壳和胚芽进行实例分割和表型提取,计算种子发芽率、发芽势和胚芽长度,评价60份棉花种子的萌发期耐盐性。【结果】生成的合成数据集包含2 000组合成图像及其相应掩模,利用该数据集训练的Mask R-CNN模型对真实图像中种壳和胚芽的分割准确度在95 %以上,基于模型提取数据获得的种子发芽率、发芽势、胚芽长度和真实测量值高度线性相关(R2 > 0.98,P< 0.001),利用模型能够准确的获取表型。各性状的耐盐指数的聚类分析将60份棉花材料分为4个水平;珂字棉4号(0.95)、MC-30(0.88)、陆8早(0.81)等材料的D值较大,其耐盐性较高。【结论】建立了基于卷积神经网络和合成数据集训练的棉花种子萌发期性状鉴定方法,并使用该方法,无损、快速且精准的鉴定了60份棉花种质种子萌发期的耐盐性。

关键词: 棉花, 卷积神经网络, 图像分割, 种子萌发