A Deep Look might only be 1 image but that image represents over 20,000 separate and unique entities. The entities are stored as .csv files and a GeoJson vector overlay. Our analysts have a deep level of skill, patience and professionalism that allows them to carefully and methodologically extract every last ounce of data. This skill sets dates back to time spent in the military methodologically counting ‘beans’ and turning the data into likely enemy equipment tables. This skill is brought up to date via our own proprietary software workbench and IP protected processes.
For certain clients we count stocks, piles, entities that are simply logged away for future consideration and reverse analysis. Where possible we augment the analyst with Artificial Intelligence technologies, processes and outputs.
Data dissemination is often in the form of the .csv files generated from the GeoJson overlay or as a .shp files for a GIS system such as QGIS.
Combining the skills from the First Look report to intimate counting of individual entities allowed MIS to produce a Deep Look report within 24hrs of the Tianjin series of explosions. Near exact car counts of loss were quickly provided to both the car manufactures and insurance companies. The outputs were large .csv / Excel tables of loss coupled with a Long / Lat.
To date we have 10,000 hours of data scraping / counting and we have been proudly delivering this data extraction service for a tech major since 2014.