Student Projects

cancel
Showing results for 
Search instead for 
Did you mean: 

On Parking and Computer Vision

Contact Information:
University: University of Florida

Team Members: Jonathan Cruz (2017), Alexander Lilley (2017), Stephan Strassle (2017), Jayson Salkey (2017)

Faculty Advisers: Dr. Daniel Dickrell

Email Addresses: jonc205@ufl.edu, alexlille@ufl.edu

Submission Language: English

Project Information:

Title:           On parking and computer vision

Description:      Use real time image acquisition and computer image processing to accurately quantify cars in a parking lot.

Products:      NI myRIO (Academic Model), LabVIEW 2013, NI Vision Acquisition Software, Logitech Webcam

The Challenge:

Parking spots at UF are hard to come by. The goal of the project was to ease the trouble by providing the location and availability of parking spots throughout campus.

The Solution:

Our project uses the NI myRIO with a Logitech camera to acquire 5 images within 10 seconds and convert these images to grayscale.  The average pixel value of these 5 images are then averaged again to create a template to compare to every other spot.  This process is repeated every 30 minutes to account for the different lighting throughout the day.   New images are acquired every 10 seconds, and converted to grayscale. The  average pixel value of select regions of interest in the image(parking spots) are compared to the template pixel value with a window of ±18%. The parking spot's score of pass or fail is then passed to a host computer which publishes the information online.

Using the myRIO allowed for real time processing at remote locations which helps in  acquiring a good view of the parking lot and provide real time updates to the public.  NI Vision Acquisition Software helped in acquiring and processing images. 

This project is currently in beta with more full scale testing required. Only small scale testing has been performed with relative success. Two full sized parking spots were correctly monitored. 

The project took approximately 6 months to complete. 

Additional revisions include, but are not limited to:  noise filters,  and measuring light intensity. 

Attached:

MyRIO movie(1).mp4 - a video documenting our live test on two full sized parking spots at O'Connell center parking lot.

Poster_.pdf - Poster describing our project in more detail.

front_panel.png - image of front panel of labview code used in the project.  The front panel displays the template image and its pixel value vs the current image and the pixel value of each spot under evaluation.

Image Analysis.png - image of labview vi that converts an image to grayscale, calculates the average pixel value of regions of interest(parking spots), and compares them to the template pixel value.

Threshold.png - image of the labview vi that creates a template pixel value every 30 minutes of 5 images to compare to the parking spots in subsequent images.

MyRIO holder model.jpg - image of beta model that houses both the myRIO and Logitech Webcam.

Contributors