A few weeks ago, as soon as favorable weather conditions made the research possible, we began intensive work on the experimental fields of the Podkarpackie Agricultural Advisory Center (PODR) in Boguchwała. About 250 different varieties of plants are grown on arable land of 17.94 hectares, located on high-quality soils (classes I, II and IIIa). Each field with diverse soil composition and variable vegetation condition provides us with rich material for research. According to PODR, the subject matter of the experiments meets the current needs of agricultural practice in the Subcarpathian region and the needs of research institutions throughout the country. The diverse condition of the Center's agricultural fields makes it an ideal area for testing multispectral photogrammetry and advanced remote sensing analysis.
Analysis from the air - photogrammetric raid
The first photogrammetric raid conducted over the PODR area showed interesting correlations and differences in the condition of plants, which initiated further raids to verify and analyze the results in more detail. In the research we used a proprietary multi-rotor drone - the Hexacopter X-01 - equipped with one of the best multispectral sensors on the market - the Altum camera from Micasense. This device allows recording data in different light ranges, such as visible light, near infrared, red edge channel and thermal imaging, making it an ideal tool for analyzing plant health at a level of detail inaccessible to the naked eye.
Radiometric calibration and data processing
During the raid, we also collected the data needed for radiometric calibration, which is a key component of multispectral analysis. To make the data reliable and enable accurate comparisons, we used a light intensity sensor and a special reflective panel. Radiometric calibration is particularly important in remote sensing measurements, as it allows us to adjust the collected information to actual atmospheric and lighting conditions.
After the raids were completed, we proceeded to process the photogrammetric material, which includes calibration, correction and creation of an orthophoto. The resulting orthophotomap consists of visible light bands, near infrared, edge red channel and thermal imaging. Such an elaborate map makes it possible to accurately identify areas of concern. Based on the orthophoto, we calculated a number of indicator maps, such as NDRE (Normalized Difference Red Edge), MCARI (Modified Chlorophyll Absorption Ratio Index) and OSAVI (Optimized Soil-Adjusted Vegetation Index). Each of these maps provides detailed information on plant health and crop health to enable decision-making to support farm management.
Point cloud and numerical land cover model
To draw final conclusions, we also used a point cloud and a numerical land cover model (NMPT), which allow us to accurately map topography and vegetation structure. The point cloud, created from the collected data, provides a spatial visualization of the area, allowing us to identify the height of plants and even their density. The numerical land cover model provides additional information on land relief and land use, which is important in analyzing cultivated areas with varying terrain. For example, an area with a steeper slope may have problems with soil erosion or uneven water distribution, which can affect plant growth.
Verification of results in the field - field observations
Although the results obtained through photogrammetry and remote sensing provide a lot of valuable information, in most cases it is necessary to verify them in the field. Based on drone data, we can identify areas of better or worse plant condition, but we can rarely immediately pinpoint the cause of the problem. Therefore, in order to get a complete picture, it is necessary to conduct field observations and combine remote sensing data with other sources of information, such as soil maps and soil moisture data. By going to places that stand out with unusual values on the indicator maps, we can confirm the drone observations and identify specific causes of problems, such as plant diseases, nutrient deficiencies or inadequate irrigation.
With the information gathered, it is possible to quickly and precisely carry out agrotechnical treatments that improve the condition of plants and increase yields. Such individualized treatments are particularly important in precision agriculture, which seeks to minimize losses and optimize the use of chemicals and fertilizers.
Future prospects - further stages of research
The results of our research conducted so far on PODR fields are opening up new possibilities in the field of precision agriculture and remote sensing technologies. In the next part of the article, we will provide more detailed analysis on the results, as well as examples of practical applications of the knowledge gathered during the research. We will also consult with experienced PODR staff who will share their knowledge on practical aspects of crop management and interpretation of results from an agricultural perspective.
Stay tuned for further reading now, where you will find a detailed discussion of the results and learn how combining multispectral photogrammetry with agricultural experience and knowledge can support farmers in their quest for better and more sustainable crops.