Project SkyHawk: Parking Assistant Drone System

Introduction: With the growth of the number of vehicles ownership worldwide, parking brings people a lot of trouble every day. Thus, we designed a system that could be deployed on any vehicles to help drivers find available parking spots in advance before entering a busy parking lot. Our system used the drone to acquire live video stream and applied computer vision algorithm to detect available parking spots. Navigation result would be displayed by a mobile app.

Approach: We use OpenCV to process photos shot by the drone. The complete system is composed of three main parts:

  1. A ground station that can create a local WiFi network, send control signal & receive the video stream from the drone, and provide photo input to the mobile app.
  2. A drone that can auto fly to the target parking lot, generate video stream and send it back images through WiFi broadcast channel.
  3. A mobile application that can display the processed result, let the user choose which empty spot to park, and provide live navigation to that specific parking spot.

Budget & Cost: The cost of the complete system is around$430. Our funding comes from a UROP grant for $400.

Market: Our system will be valuable to all drivers and parking lot management office.

Project status: 
Active
Department: 
EECS
Term: 
Fall
Academic year: 
2018-2019
Fall Poster: 
Winter Poster: 
Winter Video: