The use of remote sensing data for precision agriculture started in early 1980s. The data were used to study variations for crop and soil conditions.With the successful launch of high resolution multispectral satellites, the use of satellite data in agriculture sector has increased tremendously.
Although imagery is available from satellite systems, there are some distinct disadvantages associated with their use, such as higher cost for smaller spatial extent, as well as lower spatial and temporal resolution. Advantages of data from UAVs over satellite images include flexibility, lower cost among several others.
A farmer or agronomist can program a UAV to fly a directed path whenever they want. This allows a crop to be monitored for things that might be of interest in the growth of the plant. UAVs are capable of providing ultra-high resolution images, video capturing and NIR photography. The potential application of UAVs in agriculture is limitless. Some of them are as under:
- Identifying and monitoring the spread of crop destroying weeds/pests
- Monitoring the crop health
- Nitrogen content mapping, soil brightness mapping
- Crop cover, Biomass estimation, yield prediction
UAV technology allows a farmer to check to conditions in the field and get a better overall picture without spending time and money to travel himself over the entire field. The UAV images could be utilized to track irrigation, pests, and crop health. UAV services go well beyond traditional photography and offer crop imaging technology that would allow a farmer to spot diseases early and stop the damage. UAV can also carry sensors that pick up information invisible to the naked eye. NIR reflectance cameras can be used to measure response to vegetation stress. Some of the sensors used are
- RGB – visual inspection, elevation modeling, plant counting
- NIR – soil property and moisture analysis, crop health/stress analysis, water management, plant counting
- RE (Red Edge) – crop health analysis, plant counting, water management
- Thermal Infrared – plant physiology analysis, maturity evaluation, yield forecasting
We are investigating maize health through NDVI camera mounted on a multi-rotor UAV at different crop growth stages. Along with health monitoring, we are planning to:
- Compute the vegetation indices and find out the relationship between vegetation indices and soil nutrients.
- Identify the dragging factors like pest, weeds, fertilizers deficiency or excess and prepare the prescription map to improve the crop health.
We are using spreading wings UAV (Hexa-copter) with multispectral sensor (Sony Alpha A6000 VIZ-NIR camera) subjected to monitor the health of maize plant. The study area includes 18 plots located at Bansghari covering an area of about 4000 sq. m. near to Kathmandu University.
Total station survey was carried out to locate individual plot boundaries and evenly distribute the GCPs for post processing images acquired using the hexa-copter. Geo-located soil samples was taken from every plot and tested for various soil nutrients. Soil Maps showing the distribution of nutrients – nitrogen, potassium, phosphorous have been prepared based on the test result. We conducted flight recently on 27th May 2015 using RGB camera for instance. DJI Ground Station is used for flight planning. Circular red colored hard paper was placed on the top of GCPs to make it distinctly visible in the image. Later on, these GCP markers were used for geo-referencing images followed by mosaicing. Plant counting and DSM generation is yet to be performed.
On our next flight, we will be using VIZ-NIR camera. The images captured in NIR, Red and Green will be extracted and post processing operation (geo-referencing and mosaicing) will be carried out. After mosaicing images, Normalized Difference Vegetation Index (NDVI) map will be prepared which will provide preliminary insight of the plant condition. NDVI is the indicator of greenness/health/status of the plant whose value ranges from -1 to +1 evaluated using NIR and Red band. The value near to 1 represents good health and greenness of the plant whereas the value near to 0 or negative indicates poor health or harvesting stage. The greenness is directly linked with the chlorophyll content in the plant leaves.
NDVI computed and the soil sample results will be correlated to develop regression equation which then helps to determine soil nutrition at particular location based on the NDVI value. Thus nutrition prescription map can be produced which can be prescribed to farmers guiding to apply fertilizers at right time, right place and at right way. This helps increased profitability and sustainability, improved product quantity/quality, effective and efficient pest management, water and soil conservation as well as reduced adverse effect on farmers and consumers health.
Uma Shankar Panday