Marketplace

Aircraft Detection

Detects aircrafts on Pleiades imagery tiles of 1024 x 1024 pixels in GeoTIFF format.

Description

Orbital Insight's aircraft detection algorithm uses wide area object detection in Pleiades imagery to classify multiple types of commercial passenger and military aircraft at scale. Geospatial intelligence analysts can now conduct pattern of life analyses and activity based intelligence at global airfields of interest to conduct deeper analyses.

Technical Description

Orbital Insight's aircraft detector operates on high resolution imagery from Pléiades Streaming (One Atlas) with Raster Tiling (UP42). It produces bounding box detections corresponding to the location of planes within a specified input image (GeoTIFF). The bounding box is rectangular and vertically oriented.

Further, this algorithm classifies aircrafts into 5 classes:

  • carrier aircraft
  • fighter
  • helicopter
  • small aircraft
  • other

Note: This block runs on Pléiades Streaming (One Atlas) with Raster Tiling (UP42). Using this block with Pléiades Download (One Atlas) and Pan-Sharpening (UP42) is currently not supported.

General InformationDescription
Block TypeProcessing for Pléiades Aircraft Detection
Supported input dataPléiades Tile (1024x1024px)
Output data formatGeoJSON
PerformanceAlgorithm qualified on a large set of data encompassing different types of landscapes worldwide by a third party, with a high level of performances. The algorithm should be applied on airport areas. The algorithm was trained in over 50 countries spanning 6 continents. The training set spans many different conditions, for instance: different times of day, different terrains, different times of year, different configurations and types of planes, etc. Thus, the algorithm is location and direction invariant.
Algorithm Training Data DetailsOrbital Insight's Multi-class aircraft detection algorithm counted 54,000 aircraft from 2,500 satellite images across 350 air bases to create a historical baseline of activity and detect ongoing changes. A trained analyst saves significant time in data creation which can now be invested in data analysis.

Capabilities

Input Capabilities

raster
up42_standard
bands
[
  "red",
  "green",
  "blue"
]
dtype
uint8
format
GTiff
sensor
Pleiades
resolution
0.5
tile_width
1024
tile_height
1024
processing_level
l2

Output Capabilities

vector
up42_standard
format
GeoJSON
geometry_type
Polygon

End User License Agreement

View the End User License Agreement conditions in a new tab.