Considerations of using Aerial Data VS Satellite Data

Space Intelligence, captured by satellite may be considered an alternative or competitor to Aerial Intelligence as captured by manned or unmanned aircraft.


In practice, they can be complementary tools to get the same job done. In this post, we go over the comparative strengths and weaknesses of drone collected data VS satellite collected data.


Cadence

Satellites often have fixed, regular orbits, meaning that they will regularly pass over a location at a fixed, regular schedule.


Drones, with a few exceptions, still require an operator or pilot to go to the site with the drone and deploy it.


On the one hand, the satellite can be a little more hands off, but doesn't offer as flexible of an options when the data needs to be collected at a specific time. This can be critical when the intelligence to be captured is of a specific event, such as a suspected crime.


Resolution

Drones get 5X the resolution, or Ground Sample Distance, of satellite images for the same cost. This allows for:

more detailed qualitative analysis

more accurate object identification



Weather Blockage

Common natural effects such as…


Haze & dust

Fogs, Mists & Clouds

Smoke from wildfires

Night


…may prevent satellite imaging.

Consider wildfires in California, during which Californians cannot see the sky for weeks at a time. Likewise, the satellites cannot see the Earth either.




Similarly, but to a lesser extent, moisture in the air is more pronounced in more tropical regions and can block a large portion of satellite images for long durations.


Drone collected data is mostly unaffected by these factors, but can be more readily affected by strong winds or rain preventing or delaying safe deployment of the drone.


Human Facial Recognition

Facial data recognition can be an important goal for intelligence data, especially when the data is being used for law enforcement, or patient tracking for pandemic intelligence and prevention.


There are two common ways to recognize faces:

  • Analysts review the photos manually

  • Computer Vision Algorithm analyses photos and compares to a database

Where obviously the latter is an order of magnitude more effective, timely and scalable if such a facial database is available.


Studies such as Face Recognition on Drones: Issues and Limitations suggest that in order for a computer vision algorithm to effectively recognize a face the angle of depression of the image should be less than 60°. This is nearly impossible to achieve with today's satellite collected data.

Additionally, the image should have at least 50 pixels between the eyes. Also impossible to achieve with today's satellite collected data.

Conclusively, any intelligence data requiring facial recognition should be collected at lower altitudes by drones or manned helicopters.


Coverage Area

Where satellite data always trumps drone data is coverage area. If the need is for a very large area captured at a single instance, satellite, or high altitude aerostat such as Worldview are the only options available today to do this.


Even long endurance drones capable of capturing areas of 200km^2 and larger will still have to stitch together multiple images, which have a time interval between them.


Summary



  • Facebook - Black Circle
  • Twitter - Black Circle
  • YouTube - Black Circle