Intelligent agriculture field crop intelligent planting scientific research, learning the frontier knowledge of agriculture

In order to develop smart agriculture, the country has invested a lot of power in terms of funds, scientific research, and policies, especially in terms of technology and scientific research. China Agricultural University,

• Seeding the development of drones
• How does a seeding drone work
• Which country is the drone factory in
• How do I choose an agricultural drone factory
• the past month, there have been frequent flying plant protection

Shandong University of Science and Technology, Beijing Academy of Agriculture and Forestry Sciences, Wageningen University and other domestic and foreign universities, Interdisciplinary and interprofessional joint research has been carried out, and some progress has been made in the smart planting of field crops, including field unmanned farms, corn kernel detection and counting, corn-soybean strips, and machine learning of soil salinity, etc.

Participating institutions: China Agricultural University Key Laboratory of Intelligent Agricultural System Integration Research Ministry of Education, China Agricultural University Key Laboratory of Agricultural Information Acquisition Technology of the Ministry of Agriculture and Rural Affairs, North Dakota State University Department of Agriculture and Biological Engineering, Korea Kangwon University Department of Biosystems Engineering, Kangwon University

In order to quickly and accurately obtain information on the number of missing kernels during corn harvesting, and manage harvesting loss adjustments, the performance of different target detection networks on field corn kernel counting was compared. The team used RGB cameras to obtain data sets, constructed different target detection networks for grain recognition, conducted training, verification, and testing, and evaluated the performance of grain counting based on the recognition results of the test set images.

The results show that the YOLOv5-L network is superior to the other three networks in terms of average accuracy, model size, detection accuracy rate, missed detection rate, F1 value, processing speed, recognition and counting performance of the test set image detection, and It has strong robustness to images with different surface occlusion degrees and grain aggregation states, and can realize real-time monitoring of corn harvest loss in actual operations, with high precision and strong applicability.

Participating institutions: School of Land Science and Technology, China Agricultural University, Agricultural Equipment Research Institute of Zhejiang Academy of Agricultural Sciences, Crop System Analysis Center of Wageningen University, Marburg Research Center of New Zealand Plant and Food Research Institute

Aiming at the row-to-row differences in crop growth, phenotype and light interception in intercropping, the team analyzed the row-to-row differences in strip intercropping productivity. The team studied the construction of a plant function-structure model based on field observation data to quantify the row-to-row differences in light interception in intercropping systems. , and carried out intercropping field experiments in 2017-2018.

The results of field experiments showed that intercropping significantly increased the internode diameter of maize. Affected by corn shading, soybean internodes became longer and thinner, and the difference became more obvious as the soybean strips became narrower. Based on the three-dimensional FSPM, in the future, the layout optimization of intercropping and planting modes under different growth environments can be carried out to achieve the best advantage of system light interception.

Participating institutions: School of Surveying, Mapping and Spatial Information, Shandong University of Science and Technology, Key Laboratory of Ecological Environment of the Yellow River Delta, Shandong Province, Binzhou University, College of Agriculture, Qingdao Agricultural University

Aiming at the problem of soil salinization in the Yellow River Delta region, the team explored the inversion status of soil salinity content of UAV images under the condition of no vegetation coverage on the ground, and obtained two data sources and sample points of ground feature hyperspectral and UAV multispectral For soil salinity content, by optimizing sensitive spectral parameters, two machine learning algorithms, partial least squares regression and random forest, are used to establish a soil salinity content inversion model to realize the inversion of soil salinity content in the study area.

This study constructed and compared two different data source data retrieval models for soil salinity in the Yellow River Delta, optimized them based on the advantages of their respective data sources, and explored the retrieval method of soil salinity content in the case of no vegetation coverage on the surface. It provides a reference for deducing the degree of land salinization.

Participating institutions: Intelligent Equipment Technology Research Center of Beijing Academy of Agriculture and Forestry Sciences, National Key Laboratory of Intelligent Agricultural Power Equipment, National Agricultural Information Engineering Technology Research Center, etc.

As a form of intelligent agriculture, unmanned farms are an important exploration for building a strong agricultural country and a development direction for realizing agricultural modernization. With data, knowledge and intelligent equipment as the core elements, modern information technology and agriculture are deeply integrated to realize the integration of information perception, quantitative decision-making, intelligent control, precise input and personalized service required for the whole process of agricultural production.

Starting from the concept and structure of unmanned farms, the team discussed key technologies and equipment such as information perception and intelligent decision-making, precision operation systems and equipment, automatic driving, unmanned agricultural machinery equipment, and unmanned farm management and control platforms, and analyzed the development Key scientific and technical issues that need to be solved urgently in domestic unmanned farms. And taking the unmanned corn farm in Gongzhuling City, Jilin Province as an example, it demonstrates the specific application and effect of technologies such as the Internet of Things, big data, cloud computing, and artificial intelligence in the whole process of unmanned corn production.

According to the published smart agricultural research papers, ideally understand the current situation of domestic agricultural development and the existing opportunities and challenges, clarify the strategic goals and ideas for agricultural development, and master the frontier knowledge of agriculture.