You are on page 1of 1

Traffic Sign Detector

-Shivangi Vyas, Sandhya Sharma, Ramesh Situla, Paras Pathak, Lewis Shemery, Yiya Nkhoma

Vision Requirements Current Status


• The system shall run on iOS phone. Our traffic sign detection very accurately detects North American and
​ riving, while it may seem simple, is a very
D
• The application shall start within 5 seconds. German traffic signs. We have been able to achieve and accuracy of % in
involved task that requires an extremely high level
• The application shall use iOS phone camera. detecting traffic sign in real time. We currently have an iOS application that
of attention to be paid to do it safely and effectively.
• The System shall capture the live video, detect traffic signals, and process allows users to detect traffic signs. Our model can me implemented in an
Nowadays with so many distractions around us, it is
extremely beneficial to have technology that can the image. environment for an autonomous vehicle.
• The application shall detect the traffic signals very precisely.
assist and alert drivers of upcoming signs they
• The application shall notify the users in a timely manner.
might have missed. The goal of this product is not
to encourage inattentiveness but to provide what
could be seen as another set of eyes able to see and
alert drivers of critical things they may be unaware
of.

Mission Detailed Design Phase Future Plans

• The Goal of this application is to help driver Object detector: We have used TensorFlow's Object Detection API to create Traffic Sign Detector is currently working for iOS application. We hope to
successfully detect traffic signs that are often an accurate machine learning model that is capable of localizing traffic signs further develop an application that is compatible with different devices. Our
overlooked. It is also intended to serve as a and identifying them in live video. It will take frames from video sent to it classification does a pretty good job at accurately detecting traffic signs, but
core model for autonomous vehicles and help from the video processor and put a bounding box with a classification we wish to improve our accuracy and get yet better result. We also intend on
the system accurately and precisely detect around any traffic signs detected. expanding our database to make our application available on a global level.
traffic signs in real-time.  
Video Processor: The video processor is a program that takes video input
from the camera and prepares it to be sent to the object detector. The video is
then processed frame by frame. It will remove the left half of the image
where there would be no signs to help achieve real-time performance. It will
then resize the image and apply filters to normalize the image
 
Mobile phone: The mobile phone is mounted in a car with the camera facing
the road. It runs an app which uses the camera to record video from the car
dash. It sends this video data to the video processor.

2020 UTA College of Engineering Innovation Day April 22, 2020

You might also like