Abstract:

Wheat is one of the major global food crops, and with the development of Internet of Things (IoT) technology, multispectral dynamic acquisition technology identifies substances and features that are difficult to distinguish in the visible range by capturing rich spectral information, thus providing more detailed data support for water and fertilizer deficiency diagnosis, pest and disease warning, etc. Currently, most studies use a drone remote sensing platform equipped with a multispectral camera to acquire multispectral images of the wheat canopy, however, the drone has high operation and maintenance costs and is unable to collect continuous growth information throughout the entire growth cycle of wheat in real time, in contrast to multispectral in-situ monitoring equipment that can collect real-time growth data throughout the entire growth cycle of a crop in a specific region on a day-by-day basis, thus realizing continuous crop growth dynamics monitoring. In this study, between April 9 and June 6, 2024, images of wheat in the test field set up in the National Precision Agriculture Research and Demonstration Base in Xiaotangshan, Beijing, were collected at the nodulation, earning, flowering, and grouting stages. The valid data after screening and organizing were multispectral images collected from 6:00 to 18:00 every day at a frequency of one hour, with a data volume of 1.42 GB. The image data were captured by the multispectral in situ monitoring equipment deployed in the natural field environment at regular intervals, and stored in the form of folders. The data are screened and organized by professional staff to ensure high quality and reliability. This dataset can be used to realize the tasks of water and fertilizer deficit diagnosis, pest and disease monitoring of wheat through the multispectral image data. The extracted information such as reflectance value, vegetation index, color characteristics, texture characteristics, vegetation coverage and other information can be brought into the prediction model for analysis and prediction. At the same time, the present dataset is also suitable for constructing the chlorophyll content of wheat, network model for biomass estimation and other studies.

Key words: In situ real-time monitoring, multispectral imagery, Beijing, wheat