Case Study: Mapping Wadi Area to Analyze The Health of Plants Over a Period of Time
We worked on a small project to help analyze and identify the effects of high temperatures and drought on the health of plants. The mapped area is located in Al Ma’amour, a small village that separates Adam and Bahla. On both sides of the Wadi, plants and grass can be seen fighting their way through the high temperatures of the summer. The goal of this project is to monitor the health of plants throughout the year.
The size of the mapped area is more than 15 Kilometers squared. We collected more than 700 images using our latest drones in less than 2 hours. The drone was flying at an altitude of 50 meters in a not-very-windy day (8 m/s). The resolution of the Orthomosaic map is 1.2 cm/pixel, even though we could’ve delivered a much better resolution, we settled on the 1.2 resolution in order to speed up the delivery of the Orthomosaic map and the insights from analyzing the health of the plants. It took us less than 24 hours to have the map and analytics delivered from the moment the drone took off the ground.
Figure 1: 2D Orthomosaic map of a section of the mapped area.
Figure 2: 2D Vegetation map that represents the health of the plants.
Results and Insights
After collecting the images and the data associated with each image, it’s time to process and analyze the collected data. We first start by processing the images and develop a 2D Orthomosaic map of the area. The 2D map has a much higher resolution than any satellite image which gives us the advantage to have a better aerial view of the mapped region. After the 2D map is processed, we moved on to running a couple of algorithms to turn the aerial view into actionable insights.
We ran a computer vision algorithm on the model that has the ability to identify healthy plants from unhealthy plants. As shown on figure 2, the greener the area, the healthier the plants are. The algorithm detects the health of plants by classifying the wavelengths of lights that are collected via the drone RGB sensor.
Throughout the year, we will conduct many mapping sessions on the same region to monitor the changes of the health of the plants throughout the whole year. This is helpful in understanding hidden factors that play a vital role in how healthy the plants are.