TEKNOFEST Finalist Team

Autonomous Traffic Sign Recognition

Empowering autonomous vehicles with real-time, SSD-based computer vision for active control maneuvers and safe navigation.

Vision System Highlights

Enabling Autonomous Decision Making

Our computer vision modules utilize integrated camera feeds to accurately detect traffic signs and analyze their meanings in milliseconds, directly influencing vehicle telemetry.

01

Speed Adjustment

Instantly reads speed limit signs and autonomously throttles or brakes the vehicle to maintain strict legal compliance.

02

Intersection Logic

Accurate detection of STOP and Yield signs, triggering safe deceleration and complete stops at required junctions.

03

Maneuver Execution

Analyzes directional signs to execute correct turning maneuvers, ensuring the vehicle moves safely on its designated route.

04

SSD Architecture

Powered by the Single Shot MultiBox Detection (SSD) algorithm for high-speed, multi-scale object classification.

Enterprise Integration Ready

Our SSD-based computer vision architecture is designed for modular integration. It can be adapted to various autonomous platforms, from logistics robots to full-scale passenger vehicles, ensuring safe and reliable operation.

Competition-Proven Vision System

WIGETEC provides enterprise-grade computer vision APIs and edge-AI integration for robotaxis, smart mobility startups, and automotive manufacturers.

By following the rules determined by real-time traffic signs, our integrated system adjusts the vehicle's speed, performs turns, stops when required, and executes complex traffic maneuvers seamlessly.

TEKNOFEST Tested Robotaxi Compatible High Precision
Autonomous Vision Platform
Deep Learning Pipeline

Implementation in 3 Stages

Stage 1: Model Development & Capture

We build object detection models utilizing optimized, pre-trained weights specifically tuned for traffic signs. By capturing surrounding environmental images through integrated vehicle cameras, we construct a comprehensive, real-world dataset for robust training.

Model Development
Stage 2: Image Analysis with SSD

The live camera feed is passed through our SSD (Single Shot MultiBox Detection) algorithm. Employing advanced sliding window techniques and bounding boxes, the system detects objects at different scales and positions, extracting crucial visual features instantly.

Image Analysis SSD
Stage 3: Classification & Processing

Detected objects are separated into distinct classes to determine the exact type of traffic sign. This classified data is immediately passed to the vehicle's decision engine, ensuring the robotaxi adheres strictly to traffic regulations for safe navigation.

Classification Processing
Our Vision

WIGETEC 's Mission in Smart Mobility

Why Autonomous Vision?

Safe autonomous driving relies entirely on flawless perception. Recognizing a stop sign or a speed limit instantly is the difference between safe transport and catastrophic failure. We focus on advanced computer vision because the future of smart mobility requires systems that can see, interpret, and react to the environment faster and more accurately than a human driver.

Empowering Edge AI

Our core mission is bringing massive computational power to the "edge" – directly onto the vehicle. By utilizing optimized algorithms like Single Shot MultiBox Detection (SSD), we eliminate the latency of cloud processing. The vehicle processes its camera feeds locally, ensuring immediate response times even in areas with zero internet connectivity.

Future Horizons

Our traffic sign recognition module is just one building block. We are actively developing holistic perception pipelines that include pedestrian prediction, dynamic lane detection, and multi-vehicle communication (V2V). We envision a future where WIGETEC 's software backbone drives industrial logistics, public transport, and personal mobility with absolute safety.

Technology Core

The Tools Powering Our AI

We utilize industry-leading frameworks to deliver robust, real-time autonomous systems.

OpenCV
Image Processing
SSD Algorithm
Detection Logic
Deep Learning
Neural Networks
Sensor Data
Camera Feeds
Python/C++
Core Backend