Vehicle detection
Rapidly detect and classify vehicles from small, compressed or low-resolution images.
Rootworks vehicle intelligence
AutoRecon
Vehicle identification from the images other systems struggle with.
AI-powered visual analysis for image, video and investigation workflows, built to work with compressed, low-light, partial exterior and interior vehicle imagery.
UK fleet coverage
Built to recognise the cars on UK roads
Our analysis of official 2025 Department for Transport licensing data indicates that AutoRecon's current catalogue covers approximately three quarters of licensed UK cars manufactured from 2015 onward.
View the DfT data source
75%
working estimate of post-2015 UK car coverage
89.0%
model coverage
Approximately 19.16 million of 21.52 million licensed post-2015 UK cars align with AutoRecon's 302 supported models.
74.8%
generation-level coverage
Approximately 16.09 million cars align when the appropriate model generation year is also taken into account.
In practical terms, AutoRecon is estimated to cover approximately 70–80% of post-2015 UK cars, with 75% used as the working estimate.
Estimates are calculated by mapping the current AutoRecon reference catalogue to DfT licensing data. Recognition results vary with image quality, view, vehicle variant and available visual detail.
Capabilities
From pixels to usable vehicle intelligence
Designed for real-world media rather than controlled catalogue photography.
Partial-image matching
Analyse cropped views using lights, grilles, wheels, mirrors, body shape and rear details.
Low-light tolerance
Support recognition from night-time, CCTV, garage, shadowed and poor-exposure imagery.
Interior recognition
Identify vehicles from dashboard, steering wheel, infotainment, trim and seating cues.
Generation awareness
Distinguish model generations, facelifts and visual variants where reference coverage exists.
Visual matching
Find similar vehicles, angles, features or reference images within a managed archive.
About AutoRecon
Built for serious vehicle recognition use cases
AutoRecon turns imperfect vehicle media into searchable intelligence. It is designed for images where the vehicle is small, dark, compressed, partially visible or only seen from inside the cabin.
See the workflowImperfect images
Detect vehicles from small, compressed or low-resolution photographs.
Partial vehicle views
Work from a front corner, rear light, wheel, grille or cropped body section.
Low-light media
Support darker CCTV, night-time, garage and poor-exposure imagery.
Interior recognition
Use dashboard, steering wheel, infotainment, trim and seating cues.
Generation awareness
Distinguish model generations, facelifts and visual variants where coverage is available.
Integration potential
Support web applications, internal tooling, batch jobs and API-led workflows.
Current reference catalogue
Broad coverage of the modern vehicle market
AutoRecon's reference catalogue brings together manufacturer, model and generation-level vehicle intelligence, with coverage focused on vehicles manufactured from 2015 onward.
30
vehicle manufacturers represented
302
models covered
A broad catalogue of distinct manufacturer and model combinations.
722
model generations catalogued
Generation-level records provide more precise identification than model matching alone.
Workflow
From media ingest to ranked results
Use the product as a focused analysis tool or connect it to a wider investigation workflow.
-
01
Ingest
Submit a vehicle image, video frame, batch folder or API-delivered media item.
-
02
Analyse
Run recognition, view classification and similarity search across difficult imagery.
-
03
Review
Return ranked candidates, confidence indicators, references and investigation notes.
Request a demonstration
Bring us the imagery.
See what AutoRecon finds.
Speak to Rootworks about pilot access, API integrations or a custom vehicle recognition workflow.