Hayley Gilson, Antony Sanchez, Cris Koutsougeras ET 493 Fall 2018 Intersection Control Hayley Gilson, Antony Sanchez, Cris Koutsougeras ET 493 Fall 2018
Goal of Project: Develop Microcontroller-controlled intersection system that will use a camera and image processing software to determine traffic and use that to determine right-of-way.
Purpose of Traffic Lights Assign right of way to lanes in an intersection Assign Decrease wait times at intersection Decrease Reduce accidents that happen in accidents Reduce
Prior Designs for “Smart Intersections” Lasers Air-fused tubes Uses a inductor loop buried in the concrete and connected to the lights. Inductance and magnetic field increases in presence of iron Increase in Inductance sets off the system Induction loops
Why Use Camera-Controlled System? Cheaper to build/install Uses camera: won’t require the road to be torn up. Smarter: Won’t just detect the presence of car, but also detect amount of traffic in lanes
Semester Goals: 1 2 3 Set up system using microcontrollers and cameras Install OpenCV and Install onto system 2 Run program capable of basic image detection without input from outside party 3
Timeline September October November Meet with advisor have an understanding on what needs to be done for the project Conduct research on project September Research traffic lights and research OpenCV Acquire equipment required for the project Have OpenCV running on both microcontrollers October Set up the microcontrollers to perform simple image processing Start testing the camera analyzation and detection for objects. Continue research on failsafes and smart controls November
Parts needed: Arduino Uno controller (already acquired) Raspberry Pi controller Two cameras: One compatible with Arduino, the other with Raspberry Pi OpenCV software
Work Plan
Research Traffic Lights and the improvement of them Failsafes How they work How to improve them via camera usage Main focus of research for the semester Failsafes It’s foolish to believe that things won’t go wrong with a system What can be done to prevent catastrophe if the system malfunctions? Smart Detection The capability to anticipate what could happen Not just reacting to current situation, but predicting what can happen as a result Communication between intersections?
Construction Setting up system Using OpenCV Two systems will be built: One with Arduino, one with Raspberry Pi Functionally similar (Arduino will be somewhat simpler in terms of potential functions) Must be able to run and constantly analyze images from a camera Must function independently Using OpenCV Open-source image detection program Compatible with Arduino and Raspberry Pi Consists of multiple libraries
Construction Hayley Gilson: Railroad Intersection, Arduino Antony Sanchez: 4-way Intersection, Raspberry Pi Setup Arduino and camera and OpenCV Research train intersection functionality Use OpenCV on Arduino to do simple image processing Setup Raspberry Pi with Raspberry Pi camera and OpenCV Research traffic light algorithm Use OpenCV on Raspberry Pi to do simple image processing
Thank You for Your Time Feel Free to Ask Questions.