Drone aerial image processing

From DDIT EC Project

Area of Project[edit]

Image Processing[1]

Project Definition[edit]

Aerial Drone for parking lots using OpenCV[2] with Python[3]

Introduction[edit]

Car parking lots are an important object class in many traffic and civilian applications. With the problems of increasing urban traffic congestion and the ever increasing shortage of space, these parking lots are needed to be well equipped with automatic parking information and guidance systems. This can be achieved by an aerial drone. The goals of this project:

  • Counting no. of parked cars
  • Identifying available spots

Main Challenges[edit]

  1. How to know if it is a parking lot or not
  2. To detect and separate parking lot from obstacles
  3. Define border of parking lot
  4. Improve image quality and separate objects and shadows
  5. To detect entry and exit points of the parking lot
  6. Physical conditions affecting drone i.e. height etc
  7. Parking for physically handicapped people
  8. Sharpening the image
  9. Qualitative analysis of Thresholding

Current Goals[edit]

  1. Sherry: To try and detect whether the image contains a parking lot. If so, try to separate it from the image.Currently trying to extract color information which may be useful in detecting the presence of parking lot.
  2. Primeesha:To detect the number of cars and empty spaces considering the parking lot has been detected. Implemented LoG operator along with morphological operations to find vertical & horizontal lines. Now have to find their respective count.

Team[edit]

  1. Riya Bhangale: Vapi
  2. Neil Shah: Ahmedabad
  3. Sherry Gupta: Vapi
  4. Primeesha Nair: Nadiad

Knowledge Base[edit]

Literature Survey

Resources

Activity[edit]

Logbook.

Task Updates.

Test Bench

Meetings.

References[edit]