IRIS: Cheolha Pedro Lee

Cheolha Pedro Lee
Electrical Engineering
M.S. 2001

UXO (UneXploded Ordnance) Detection

Lamids Durinal by Amber Galileo LWR (IR, 8-9 Micron, 12bit)
(avi)

24 hours with 15 intervals since 03:00PM

Type	Location	Condition
M15	67,50	Buried
M19	139,56	Buried
PGMDM	191,42	Surface
RAAM	222,42	Surface
FFV028	75,109	Buried
TM62	131,109	Buried
VS1.6	204,112	Buried

Seven buried or surface UXO are placed in an area of 8'x 8'. The sensors were affixed on tripods.

Lamids Durinal Cincinnati Electronics IRRIS-256ST (IR, 3-5 Micron, 12bit)
(avi)

24 hours with 15 intervals since 03:00PM

Type	Location	Condition
M15	96,105	Buried
M19	130,106	Buried
PGMDM	156,110	Surface
RAAM	171,111	Surface
FFV028	91,136	Buried
TM62	118,138	Buried
VS1.6	160,143	Buried

Seven buried or surface UXO are placed in an area of 8'x 8'. The sensors were affixed on tripods.

UXOA Diurnal by Amber Radiance 1 (IR, 3-5 Micron, 12bit)
(avi)

10:00AM-08:30PM usually 30 minutes intervals

Type	Location	Condition
PMD6	208,27	Buried
M15	100,99	Buried
TS50	183,184	Buried

Three buried UXO are placed in the area of 4'x8'.

This data came from the signature database of UXOCOE.
These images were histogram-equalized from the original one since their original intensity is too concentrated in specific levels.

Research Objectives:
There are millions of mines threatening people's lives in the world. Most of them are no longer necessary, even for military purposes. My research intends to find landmines by analyzing image data such as infrared images captured by sensors.

Methodology:
I converted some raw sensor data into general image data. I am looking for some unique character of mines while analyzing the sequence image.

Sample Data (from UXOCOE)