AI-powered traffic analytics that detects and classifies 13 vehicle types on Indian roads in real-time. Deploy edge devices, get instant dashboards.

Not a generic counter. Drishti understands the chaos of Indian roads — auto-rickshaws, two-wheelers weaving through traffic, overloaded tempos, and everything in between.
Custom-trained on Indian vehicle dataset. Detects Hatchbacks, Sedans, SUVs, MUVs, Buses, Trucks, Three-wheelers, Two-wheelers, LCVs, Mini-buses, Tempo-travellers, Bicycles, and Vans.
Runs on Intel N95 mini-PCs with OpenVINO optimization. No cloud dependency — all inference happens locally at 30 FPS with <25ms latency.
Real-time SSE streaming to a web dashboard. Hourly breakdowns, vehicle class distribution, trend analysis, and Excel exports for reporting.
Network goes down? Detections buffer locally in SQLite with automatic retry. Zero data loss even during extended outages.
Registration key activation model. Encrypted settings. Role-based access control. Admin can revoke any inventory remotely in real-time.
One-click export of hourly vehicle counts by class. Perfect for outdoor media planners, traffic engineers, and municipal reports.
Mount an Intel mini-PC with a camera at your monitoring point. Connect to RTSP stream or USB webcam. Power on.
Enter the one-time registration key in the Drishti app. Set the counting line position. Detection starts immediately.
View real-time counts on the web dashboard. Get hourly breakdowns by vehicle class. Export Excel reports for clients.

Access the analytics dashboard to view real-time vehicle counts, hourly trends, and export reports.