王春颖,泮玮婷,李祥,刘平.基于ST-LSTM的植物生长发育预测模型[J].农业机械学报,2022,53(6):250-258.
WANG Chunying,PAN Weiting,LI Xiang,LIU Ping.Plant Growth and Development Prediction Model Based on ST-LSTM[J].Transactions of the Chinese Society for Agricultural Machinery,2022,53(6):250-258.
摘要点击次数: 974
全文下载次数: 330
基于ST-LSTM的植物生长发育预测模型   [下载全文]
Plant Growth and Development Prediction Model Based on ST-LSTM   [Download Pdf][in English]
投稿时间:2022-01-09  
DOI:10.6041/j.issn.1000-1298.2022.06.026
中文关键词:  植物表型  深度学习  生长发育  预测模型  时空长短时记忆网络
基金项目:山东省自然科学基金重点项目(ZR2020KF002)、山东省重点研发计划(重大科技创新工程)项目(2021LZGC013)和国家自然科学基金项目(31871543)
作者单位
王春颖 山东农业大学 
泮玮婷 山东农业大学 
李祥 山东农业大学 
刘平 山东农业大学 
中文摘要:提早预知植物生长发育是智能育种过程的重要组成部分,针对植物表型难以精准预测和模拟的问题,利用植物生长发育的空间和时间依赖性,提出了一种基于时空长短时记忆网络(Spatiotemporal long short-term memory,ST-LSTM)的植物生长发育预测模型,实现植物生长发育的预测。首先,通过微调Mask R-CNN模型实现识别、提取植物掩模,预处理具有时空相关性的植物生长发育图像序列,构建植物生长发育预测数据集。然后,基于ST-LSTM建立植物生长发育预测模型,利用历史生长发育图像序列,融合时空深度特征,预测植物未来的生长发育图像序列。研究结果表明,所提出模型预测的图像序列与生长发育实际图像序列具有较高的一致性和相似性,首个预测时间节点的结构相似度为0.8741,均方误差为17.10,峰值信噪比为30.83,测试集的冠层叶面积、冠幅和叶片数预测R2分别为0.9619、0.9087和0.9158。该研究实现了基于植物生长发育图像序列的生长发育预测,有效减少了田间反复试验的时间、土地和人力成本,为提高智能育种效率提供了参考。
WANG Chunying  PAN Weiting  LI Xiang  LIU Ping
Shandong Agricultural University
Key Words:plant phenotype  deep learning  growth and development  prediction model  ST-LSTM
Abstract:Early prediction for the growth and development of plants was an important component of the intelligent breeding process. However, it is difficult to accurately predict and simulate plant phenotypes. A prediction model of plant growth and development was proposed based on spatiotemporal long short-term memory (ST-LSTM) to predict future growth and development of plant. Firstly, the plant masks were recognized and extracted by the pre-trained Mask R-CNN model and the background of the plant image was removed by morphological operations. Then, the plant growth and development prediction data set was constructed. After that, utilizing the spatial and temporal dependence of plant growth and development, the image sequence of plants future growth and development was predicted by the prediction model for plant growth and development using the spatial and temporal depth characteristics integrated from the image sequence of early plant growth and development. The results showed that the image sequence predicted by the proposed model had high consistency and similarity with the actual image sequence of growth and development. At the first prediction time node, the structural similarity index measure was 0.8741, the mean square error was 17.10, and the peak signal to noise ratio was 30.83. The prediction determination coefficient (R2) of canopy leaf area, crown width, and leaf number were 0.9619, 0.9087 and 0.9158, respectively. Finally, the research realized the prediction of growth and development based on the image sequence of plant growth and development, which would effectively reduce the time, land and labor cost of repeated experiments in the field, and provided a reference for improving breeding efficiency.

Transactions of the Chinese Society for Agriculture Machinery (CSAM), in charged of China Association for Science and Technology (CAST), sponsored by CSAM and Chinese Academy of Agricultural Mechanization Science(CAAMS), started publication in 1957. It is the earliest interdisciplinary journal in Chinese which combines agricultural and engineering. It always closely grasps the development direction of agriculture engineering disciplines and the published papers represent the highest academic level of agriculture engineering in China. Currently, nearly 8,000 papers have been already published. There are around 3,000 papers contributed to the journal each year, but only around 600 of them will be accepted. Transactions of CSAM focuses on a wide range of agricultural machinery, irrigation, electronics, robotics, agro-products engineering, biological energy, agricultural structures and environment and more. Subjects in Transactions of the CSAM have been embodied by many internationally well-known index systems, such as: EI Compendex, CA, CSA, etc.

   下载PDF阅读器