OBJECT DETECTION THROUGH IMAGE PROCESSING FOR UNMANNED AERIAL VEHICLES

Screen Shot 2017-11-02 at 4.15.21 PM

System requirement: Adroid/ DJI SDK/ OpenCV.

 

Abstract

An Unmanned Aerial Vehicles (UAV) is an aircraft operated without a human pilot aboard and comes with the ability to fly to the desired locations using GPS coordinates, not limited to flying manually per human operator’s instructions using remote control equipment. Some UAV product vendors also saw commercial and research opportunities to publish their APIs in order to allow extending the capabilities of their UAV products. This thesis presents a method that utilizes a UAV, commonly called a “drone”, with a camera to fly over a small area of a playground in order to detect a red circular object using image processing techniques. The solution consists of a custom Android mobile application which sets the waypoints for the drone while capturing video and GPS coordinates. The system utilizes OpenCV library to process the images from the drone in order to alert human operator with the relevant image along with the drone’s approximate location when the image was captured. A possible future project extending the system could be to fly multiple UAVs with each having its own flight plans to cover a larger area, such as a disaster area, in order to find human beings or certain objects and alert rescue teams with captured images along with timestamp and location for them to investigate further.

Screen Shot 2017-03-31 at 5.21.31 PM

Detected App

Screen Shot 2017-03-31 at 5.20.48 PM

Screen Shot 2017-11-02 at 4.15.38 PM

 

  • A. Data captured and streamedScreen Shot 2017-11-02 at 4.16.06 PM

(*The right-hand side picture is from DJI website )

Screen Shot 2017-11-02 at 4.16.13 PM

 

  • B. RED circular object detected

Screen Shot 2017-11-02 at 4.16.35 PM

Screen Shot 2017-11-02 at 4.16.51 PM

  • Target detected

Screen Shot 2017-11-02 at 4.17.20 PMScreen Shot 2017-03-31 at 5.22.15 PM

Test Case

Screen Shot 2017-03-31 at 5.22.51 PM.png

Screen Shot 2017-03-31 at 5.23.08 PM.pngScreen Shot 2017-03-31 at 5.23.19 PM.png

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s