Three cases where Satellite Imagery can tackle real-world problems
Computer vision is beginning to explode — its usefulness is increasing with new technology being developed every single day. Satellite Imagery (multispectral imagery taken by satellites orbiting the Earth) is a subset of computer vision, which can help humanity solve problems on a very large scale. Let’s look at some important use cases where satellite imagery is or can be applied to solve pressing problems.
In food-insecure regions of the world, maximum crop efficiency ensures maximum outputs. This is even more important in areas where the primary sector, agriculture, serves the majority of the economy — meaning that farming is many people’s livelihood.
Crop health can be monitored using multispectral imagery taken from an aerial source, like drones or satellites. Using time-series, patterns in crop health can be mapped and monitored.
Computer vision can therefore be applied in many aspects: To identify nutrient deficiencies caused due to mismanagement of fertilizer/natural erosion, identify diseases and pest infections in fields before they start to spread, and monitor crop health using existing tools to measure chlorophyll levels or plant canopy thickness, calculate the area of the field covered by weeds.
In food-insecure regions of the world, maximum crop efficiency ensures maximum outputs.
This is an example of an algorithm, the NDVI (Normalized Difference Vegetation Index) being used on multispectral satellite imagery taken by Landsat over the San Jose-Bay area. The crop
Algae are organisms that photosynthesize to make energy. They lack the reproductive system or complex structure of other plants. Algae are most commonly found in water bodies such as lakes, forming a colored covering on their surface.
Algae are a natural part of ecosystems. However, in large quantities, they cause lots of damage: these Algal blooms occur when the population of algae in water rapidly increases. These algal blooms
- Produce dangerous toxins that can sicken or kill people and animals
- Create large dead zones
- Raise drinking water and purification costs
- Hurt industries depending on clean water
In India, Rajasthan’s Udaisagar lake, J&K’s Dal Lake, Madhya Pradesh’s Upper Lake in Bhopal, and Odisha’s Chilika lake suffer from Algal blooms on a frequent basis. They are hard to predict and damage the locality.
Algal blooms aren’t just limited to smaller bodies of water like lakes — they can affect entire seas and oceans given the right conditions. This picture shows a visible algal bloom captured by aerial imagery off the southern coast of England. The length of it is about 120 kilometers. So the area of impact of these blooms is large.
Satellite imagery can help combat Algal blooms. In the United States, the CyAN (Cyanobacteria Assessment Network) is the product of coordination between the EPA, NASA, and USGS to “develop an early warning indicator system to detect algal blooms in U.S. freshwater systems”, using satellite imagery.
The UN’s Food and Agricultural Organization pegs total forest area as approximately 4.06 billion hectares or 31% of global land area — that’s a lot of land! Deforestation can be divided into two distinct causes: manmade (e.g. logging) and natural (e.g. forest fires due to lightning strikes). The annual rate of deforestation was estimated to be around 10 million hectares every year between 2015 and 2020. Our forests are thinning yearly, and it is currently difficult to prevent it. How come?
One serious issue behind tackling deforestation is the sheer size of forest cover to monitor. A high percentage of deforestation happens through illegal, unmonitored means or through natural causes — both of which are hard to track and combat.
Computer vision and satellite imagery can be used to combat both natural and manmade deforestation causes. Algorithms like the Burnt Area Index already exist (the BAI uses values in the Red and NIR bands of multispectral imagery to compute the total surface area burnt).
These are just three of the heaps of use cases satellite imagery has. Even within these three, there is a multitude of different unique problems which the imagery can distinctly combat.