Image Recognition
– for Search, Rescue & Emergency Response
Online Image Recognition – For your Toughest Missions
– Find a Target Object and its GPS Location
Rapidly scan thousands of images for…
People Aircraft Helicopters Vessels Vehicles Color General
SOS Ship-Wake Oil-Spill Clothing/Apparel Textures/Patterns
Travel Food Celebrity
– 19 Major Image Categories, 11,000 Keywords –
Image Recognition:
Overview:
SAR Technology’s Searchlight Image Locator – Scans thousands of images: Persons, Aircraft, Helicopters, Vehicles, Vessels, Color, General, SOS, Ship-Wake, Oil-Spill, Clothing/Apparel, Textures/Patterns, Travel, Food, Celebrity images – for the target objects.
Image Recognition missions consist of automatically taking hundreds of overlapping images while flying a route or specific Flight Plan pattern over a defined search area. These digital images are then rapidly computer-scanned by the Searchlight Image Locator to find any containing the target object.
Benefits:
– Image Recognition Flights
– deliver much larger area coverage (50-200X) than typical ground-search procedures.
– Image Recognition can significantly improve target object detection compared to visual searching.
– Image Recognition can detect specific-colored, rotated or only partially-visible target objects.
– Image Recognition is essentially time-independent (unlike FLIR).
– Missions may be started or continued for many days or even weeks after the initial event.
– Increases Responder Safety – by remote searching for objects in dangerous terrain.
– Permits time-constrained brief ‘Windows of Opportunity’ searches.
– Generates a permanent record of the aerial search coverage.
– Creates multiple opportunities to review stored search images.
– Provides the responder with faster, larger-area and higher-detection searches.
– Assists the responder with even the most challenging search missions.
Image Scans
– Users scan their own images into the online Searchlight Image Locator webpage
– to look for the target objects or color within their uploaded image collection.
– Still images (.jpg or .png) taken by drones, aircraft, helicopters, action cameras, digital cameras and cellphones, may all be scanned by the Searchlight Image Locator. See Image Guidelines
Location Geotag:
Images should be taken, if possible, with Location Enabled, to add the Latitude-Longitude geotag to each image in the scan report.
Tip: Confirm that location has been enabled before taking the full collection of images.
SAR Technology’s ‘Image Commander‘ software can also be used to geo-tag submitted images.
Target Image:
Wherever possible images of (or similar to) the target should also be scanned, to improve the image recognition and segmentation accuracy.
– Extensive library of crashed aircraft images.
– Includes commercial, military and light aircraft.
– Located in a wide variety of terrains.
Crashed Helicopter Recognition:
– Extensive library of crashed helicopter images.
– Includes commercial and military helicopters.
– Located in a wide variety of terrains.
Missing Vessel Recognition – Vessels, Ship-Wakes and Oil-Spills
– Extensive library of Vessels and Watercraft.
– Custom library of Ship-Wakes.
– Custom library of Oil-Spills.
– Vessels engaged in commercial and recreational activities.
– Located in a wide variety of marine environments.
People Recognition – In Back-Country, Urban Settings & Culture
– Extensive Libraries of people images.
– People in Backcountry Locations.
– People in Urban Settings.
– SAR/Police/Security/General Investigations.
People Classification:
– Identify People by Cultural Group.
– Identify People By Age Group.
– Identify People by Gender.
Vehicle Recognition – In Back-Country, Vehicles & People
– Extensive Libraries of Vehicle images.
– Wide range of Vehicle Types.
– Vehicles engaged in Off-Road/Back-Country travel.
– Vehicles and People in the same image.
– SAR/Police/Security/General Investigations.
– Vehicles in a wide variety of Terrains.
– Comprehensive General image library.
– Extensive collection of image types.
– High quality image collections.
– 19 Major Image Categories.
– Over 11,000 Image Keywords.
(displayed images are representative)