Drone-based Multispectral Imaging for Agricultural Development

Detection of surface reflection and vegetation indices for multi-purpose evaluation

Towards a Solution

To solve some of these challenges, a wider use of technology is needed to improve agricultural outcomes. For instance, by the time human observation notices changes or damage to crops, it may be too late to take precautions. Using technology based on visible light and near infrared (NIR) spectra, natural formations can be identified and changes recognized long before symptoms develop. The images for this are taken with cameras mounted in unmanned aerial vehicles (UAVs, or drones). Due to the numerous benefits provided by UAVs, their use in agriculture has been increasing rapidly. Remote sensing systems can be used to monitor plants and detect disease agents, pests, weed flora and water stress. Damage rates and harvest time can be determined, along with yield estimations. Vegetation can be evaluated in terms of quality and quantity using related indices.

To address the challenges in Azerbaijan, the Drone-based Multispectral Imaging for Agricultural Development project, supported by the Turkish Cooperation and Coordination Agency (TİKA), makes use of unmanned aerial vehicles (UAVs), especially in regions that are risky in terms of weather and soil conditions, pesticides, faulty spraying, etc., for crop status examination, disease and pest detection, analysis of plant stress levels, yield estimation and weed detection. The UAVs capture higher resolution images and are faster and more cost-effective than conventional systems.

By using drones, together with complementary software and hardware, the project allows the:

• creation of indices for plant identification;
• creation of indices for disease detection and agricultural spraying with drones;
• detection of harmful pests and weeds;
• analyses of soil structure and characteristics; and
• detection of abiotic and biotic stress-induced symptoms.

To implement the project, TİKA donated two drones (a multispectral drone and an agricultural spraying drone) and mapping and image analysis software to the Agricultural Innovation Center of Azerbaijan. This equipment is being used in disease agent detection, pest detection, determination of damage rates, weed flora detection, water stress detection, harvest time detection and yield estimation through images taken with cameras on the UAVs. The drones take high-resolution photos of hundreds of hectares of land in a single flight. Imaging through the UAVs can be completed in a shorter time than traditional work, using less human and labour resources. UAVs are capable of photographing even in cloudy weather conditions, when satellite photographs are not available. Aerial activities are supported by ground studies. Although TİKA’s part of donating the drones and training the relevant personnel has been completed, the use of gadgets and know-how in other research activities and studies by the Agricultural Innovation Center of Azerbaijan will continue.

In addition to determining surface reflections and vegetation indices for multi-purpose evaluation, training on agricultural spraying with drones is carried out, including training on the use of both the hardware and the software. The Agricultural Innovation Center provides physical facilities, has obtained legal permissions for drone flights and designates the experts to participate in the training.

So far, three Azerbaijani experts were trained on UAVs over five days in December 2021, held in Ankara, Turkey, and 22 experts received UAV A1 category pilot training in Azerbaijan in April 2022.

For the project’s sustainability, Azerbaijan Ministry of Agriculture Science and Innovation Center and Azerbaijani farmers will be able to continue their training and research with the drones provided. Other training programmes are planned for different instructors and in different areas of UAVs used in agriculture, both in Azerbaijan and Turkey.

The results from the analysis of the drone data obtained are expected to be collected in scientific articles, making them available to countries around the world. Upon assessing the need and availability of a recipient country or institution, this project may be replicated. Research and preparations have already begun for replicating the project in more countries.