India's first AI-powered
Global Shutter Camera
Captures every detail at 230 km/h, with advanced AI running on the camera.
Recon Global Shutter is now STQC-certified.
Policy Compliant
(Profile S, T, M, G)
See. Reason. Act.
Where do you work?
Highway
See
Three cameras capture every lane at 60 FPS.
Reason
Vision Language Model on the camera
“White car in lane 1. 112 km/h where the limit is 100.”
Over speed
Toll plaza
See
One camera for each lane reads vehicles at road speed.
Reason
Read on the camera
“State bus. Commercial plate. Lane 2.”
Class: Bus
City
See
Cameras watch each lane at the stop line.
Reason
Vision Language Model on the camera
“Scooter with three riders. None wears a helmet.”
Triple riding
Gated premises
See
GateGuard reads the plate before the barrier.
Reason
Checked on the camera
“This car is on the whitelist.”
Allow entry
Advanced AI cameras, built on our own research and deployed across India
Stellarview is a research-led edge AI company. Our cameras run Vision Language Models on the device, so they understand a scene and answer questions about it in real time.
We develop the core technology in-house, from the AI models to the camera hardware. Our research ranked World #1 at the NVIDIA AI City Challenge 2026.
The same technology runs at scale on highways, toll plazas, city junctions and gated premises across India.
Certified management systems




Built for the corridor, the city, the plaza, and the gate
From highways and tolling to smart cities, gated communities, and commercial complexes - each path maps to the products operators deploy.
The Stellarview ecosystem

Recon Global Shutter Camera
Global shutter AI camera with zero motion blur at 230 km/h.

EdgeTrack
One control centre for every camera, event and device on the network.

EdgeTrack VMS
Live video, playback, signed evidence and semantic search for control rooms.

GateGuard
ANPR access control that opens the barrier for authorised vehicles.

GateGuard Connect
Cloud dashboard to manage every gate, vehicle list and entry log.
AI Search
Ask a full sentence. Get the evidence. AI Search uses a Vision Language Model on the Recon camera. Describe the vehicle, the rider, the clothing, the violation and the time in one sentence. It returns matching frames with timestamp and plate.
Searches run on a live camera
138 matchesResults from one Recon camera on a city road. Faces are blurred.
1,000+ deployments across highways, tolling, enterprise and residential
NHAI, Lucknow Ring Road: 104 km under continuous enforcement
57 AI-Powered Global Shutter cameras across 22 gantries, built to NHAI's 2023 VIDES policy.
(NATRAX)
Figures above are across deployments. Site results for Lucknow Ring Road are in the case study.
Read case study →Bosch, Bangalore: automated access across 3 gates
−40% entry/exit time, −90% unauthorised entries.
Millennium International School
Full vehicle logs and tighter campus access control.
Ahuja Towers, Mumbai: faster peak-hour entry
ANPR access for a 53-storey high-rise.
All case studies →Stellarview in the news

Stellarview at Bengaluru Tech Summit 2026
We showed Recon in the AI & Robotics Pavilion at BIEC, Bengaluru, from 17 to 19 November.

Recon at the NVIDIA Inception Startup Pavilion, Huddle Global 2026
A live demonstration at the NVIDIA Inception Startup Pavilion in Kovalam, and a talk on imaging for edge AI.

Stellarview Vision-Language Model ranked world #1 in traffic video AI
Track 3 win at the NVIDIA AI City Challenge 2026, with research accepted at ECCV 2026 in Malmö, Sweden.
Research, events and announcements
Engineering
- Why global shutter for highway ANPR
- What did not work: VLM negative results
- Vision language models on the edge
- Reason or Recite paper explained
- Reading Indian number plates



































