TEMPLATE DESIGN © 2008

www.PosterPresentations.com

Autonomous Landmine Detection Drone

IntroductionLandmines are a global issue. According to the

United Nations (UN), landmines kill more than

15,000 people each year, most of which are

children, and leave large fertile land uncultivated in

78 countries such as Angola, Egypt, and Eritrea.

The current method of minesweeping involves a

human being manually scanning an area using a

metal detector, ground penetrating Radar or

trained animals such as dogs and rats to sniff out

the mines. However, those methods are

expensive, inefficient, and risky. According to the

UN for every 5,000 landmines removed, one

worker is killed and two are injured.

It costs 300-1,000 dollars to remove a single mine.

It cost more than 5,280 dollars to train one rat

which can only live up to 6-8 years. This creates a

huge economic and humanitarian crisis in

communities in which they exist. However, we can

solve this problem using a swarm of autonomous

quad-copters equipped with metal detectors to

detect mines. The proposed idea is to make the

quad-copters autonomously scan a parameter

designated by the base station. The quad-copter

will virtually run non-stop with the exception of

battery swaps. The outcome of the proposed

swarm of quad-copter system will not only be

efficient it will also reduce the risk involved in

minesweeping. In addition, our proposed system is

also cost efficient as there is less overhead.

Abstract

Methodology

MaterialsSchematic of Drone Conclusion/Discussion

Graphics of Finished Drone

Today in the world we currently have an estimated

110 million anti-personnel Landmine in the world.

Sadly the majority of landmine victims are civilians

who step on a mine after armed conflicts have

ceased. The current methods of minesweeping

are expensive and inefficient. However, quad-

copters equipped with a metal detector can do the

tedious and dangerous task of minesweeping. In

this project we are proposing a quad-copter

platform that can be utilized for such projects. The

quad-copter is controlled from a computer. The

computer gets a top view of the quad-copter which

is then processed to localize the drone. The

computer then calculates the right value for pitch,

yaw and roll, then sends those value to the PCTX.

The PCTX generate PPM signal that correspond

to the value and transmits these values to the

radio transmitter through the trainer port and the

transmitter then sends the instruction to the

Pixhawk flight controller on the quad-copter.

Acknowledgements

Yikealo Abraha, Berhane Tesfamichael, Joshua Lee, Asadawut Promjak

Advisor: Andy Hulse

Andy Hulse- Santa Rosa

Junior College Faculty

Erick Shnell – Santa Rosa

Junior College Student

Camera

PC

PCTX

Radio

Transmitter

Flight

Controller

Radio

Receiver

Quantity Item Price Total

1 Frame $29.99 $29.99

4 Tiger Motors $40.00 $160.00

4 ESC $11.00 $44.00

1 Telemetry $25.00 $25.00

1 Pixhawk $200.00 $200.00

2 Standard Propellers $10.00 $20.00

2 Pusher Propellers $10.70 $21.40

1 Power board $5.00 $5.00

1 LiPo Batteries Free Free

1 LiPo Charger Free Free

1 Velcro $10.00 $10.00

1 Zip ties $5.00 $5.00

10 Bullet Connectors $10.00 $10.00

Totals: $530.39

Camera

• Takes Top view videos of the drone

• Sends RGB formatted images to PC

PC

• The PC receives RGB image frames and

process those to localize the quad-copter

• The PC then generates value for each channels

based on the needed yaw, pitch, and roll

• Which then sends floating point values for each

channel (9 channels) to PCTX

PCTX

• Receives floating point value for each of the 9

channels

• Generates PPM (Pulse Position-Modulation

signal based on the 9 values)

• The PPM signal is then sent to the trainer port of

the radio transmitter

Radio transmitter

• Radio transmitter relays the signal to the

receiver which is connected to the flight controller

• The flight controller receives the PPM signal and

translated to the corresponding pitch, yaw, and

roll

• RGB image format is converted to HSV (Hue

Saturation Value) format

• HSV image filtered to only show the spherical

indicators using inRange() function from the

Open image processing library

• The contour of the sphere are detected using

findContour function

• The center of the spheres is located by

averaging the location of the contours

• The frame distance of the two spheres is

calculated and used as to calculate K, where K

is the actual distance/frame distance.

• The distance of the two sphere is calculated

using K and frame distance from the center of

the frame.

References

Building upon Opencv (image processing

library), we created a library that localized the

quad-copter

Communication with the PCTX was not as

smooth as we expected. The interface needs a

further development

We also created a C++ Class for the quad-

copter with functions such as “get location” and

“move up”

Our greatest One of deficiency was calibrating

the drone’s sensors in order for complete

autonomous flight

Further development the Autonomous landmine

detection drone project will continue

In the future we plan to eliminate the PCTX and

radio transmitter; instead we would establish a

direct connection between the PC and the flight

controller

www.care.org/emergencies/facts-about-landmines

https://www.icrc.org/eng/resources/documents/misc/57j

mcy.htm

http://www.scielo.org.co/pdf/dyna/v81n183/v81n183a13.

pdf

Kyle Hounslow – Opencv tutorials

Object