REPUBLIC OF SERBIA

MINISTRY OF DEFENCE

www.mod.gov.rs

MINISTRY OF DEFENCE

Material Resources Sector

Defensive Technologies Department

Military Technical Institute

www.vti.mod.gov.rs

THERMAL IMAGE NOISE REDUCTION POST-PROCESSING TECHNIQUES

NATAŠA VLAHOVIĆ

Vlatacom Institute, Belgrade, natasa.vlahovic@vlatacom.com

milan stojanović

Vlatacom Institute, Belgrade, milan.stojanović@vlatacom.com

ljubiša tomić

Military Technical Institute, Belgrade, ljubisa.tomic@gmail.com

katarina mišković

Military Technical Institute, Belgrade, msc.katarina.miskovic@gmail.com

branka stojanović

Vlatacom Institute, Belgrade, branka.stojanovic@vlatacom.com

aleksandar kovačević

Technical Test Center, Belgrade, aleksandarkovacevic1962@yahoo.com

Abstract: Thermal image has become more popular nowadays with an increased need for systems that enable night vision, severe fog vision, but also less privacy intrusion. Also, the decreased price and better image quality enabled usage of thermal cameras not only in military, but also in civil applications. In spite of the advantages of using thermal imaging, it should be emphasized that thermal tracking has several challenges, since the thermal videos, compared to the visual camera videos, possess several undesirable properties, such as low resolution, large percentage of dead pixels, lack of color, motion blur, changing appearance, occlusion, etc. The inevitable part of every system is noise that compromises original signal from sensor. That is also the case with thermal image, which can be compromised with different types of noise. In this paper, several different noise reduction techniques are presented for different types of noise in thermal image. All the techniques are tested on experimental data.

Keywords: thermal image, noise analysis, noise reduction, median filter.