Electro-optics and Infrared Image Registration
using Gaussian Pyramids
Hyeok-Gi Gwon, Ik-Hyun Lee, and Tae-Sun Choi
School of Mechatronics, Gwangju Institute of Science and Technology (GIST),
123 Cheomdan Gwagiro, Buk-Gu, Gwangju, 500-712, Republic of Korea
{regret66,ihlee,tschoi}@gist.ac.kr
http://sipl.gist.ac.kr
Abstract. Image registration is a spatial alignment of two or more im-
ages and essential technology in image fusion, surveillance. Unlike the
electro-optics (EO) sensor, infrared (IR) sensor absorbs the radiation
energy. The relation between IR and EO image of pixel intensity is
sometimes similar or often reverse. To overcome this complicated prob-
lem, we propose block-based processing incorporating corner detector
and descriptor. This method also uses the Gaussian pyramids to register
EO/IR image. The proposed method provides more accurate registration
results.
Keywords: Image registration, Gaussian pyramids, Feature extraction
1 Introduction
In particular, image registration is considerably important technique in remote
sensing, image fusion system, and surveillance. Image registration is a spatial
alignment of the same area acquired from same or different times, views and
sensors (reference and sensed image). Electro-optics (EO) image is acquired by
using reflection and radiation of visible rays, whereas infrared (IR) image is use
to absorb the radiation of energy and temperature information. Generally, single-
modal registration is commonly used for many applications with visible range
sensors during day and in indoor environments. In multi-modal case, infrared
sensors were used in special medical imaging or and night vision. In surveillance
system and remote sensing, infrared sensors provide more information than EO
sensors in condition of poor lighting, smoke, fog, and cloudy weather. According
to the different characteristics, using the both EO/IR sensors are more effective
and each compensative. EO sensors are better than IR sensor during the day,
while IR sensors provide more information in the dark. The drawback of using
two sensors at same time is hard to extract corresponding features due to the
complicated pixel intensity relation. It has sometimes similar intensity, while
often reverse. Moreover, the contrast of IR image is less than EO image.
In the literature [7], feature-based methods is recommended to use in remote
sensing which has lots of information, whereas area-based methods are employed
Advanced Science and Technology Letters
Vol.29 (SIP 2013), pp.55-59
http://dx.doi.org/10.14257/astl.2013.29.11
ISSN: 2287-1233 ASTL
Copyright © 2013 SERSC