Iranian Traffic Sign Recognition
← Back to archive

ADAS / IRAN ROAD SIGNS

Iranian Traffic Sign Recognition

A TSR case for SAIRAN: Iranian road-sign recognition under scarce data, low light, damage, and noisy conditions

NIGHTRAINDUSTATTENTIONCNN
Skip the 3D journey
PERCEPTION / LIVE

When the sign is small, the cost of error is large

The TSR case is designed around real road frames, occluded signs, damaged panels, and low-light conditions. The page mirrors that mission with road rhythm, scan lines, and the actual project video

01 Iran

Data domain

Local signs and real road conditions

02 ADAS

Use case

Driver assistance and smart mobility

03 Robust

Focus

Night, rain, snow, dust, and damaged signs

Project story

From limited data to a dependable perception layer for real roads

This case targets one of the hardest perception layers in intelligent driving: a small sign, sometimes damaged or partially hidden, inside frames disturbed by light, rain, dust, and vehicle motion. The technical path covered problem framing, architecture, training, evaluation, and deployment readiness. Augmentation strategies addressed scarce labelled data, while CNN and attention mechanisms improved robustness for Iranian road signs

Tools & technologies

Computer VisionCNNAttentionData AugmentationADASPython

TSR / SYSTEM MAP

Road perception dashboard

01

Recognition architecture

A TSR pipeline combining CNN features and attention mechanisms to catch small, noisy road signals

02

Data strategy

Augmentation plans for scarce labelled data and harsh environmental scenarios

03

Deployment readiness

Technical consulting from problem framing to training, evaluation, and ADAS readiness

PLAY PROJECT FILM
01Sample video from the road-sign recognition flow
Next mission · 02 BazarFootball - footballtalent