IMAGE PROCESSING FOR MULTIPLE MICRO-RADIOGRAPHY IMAGES
Andreas C. Louk
1,2
, Gede B Suparta
3
and Nurul Hidayah
2
1
Dept. of Physics Nusa Cendana University, Kupang, East Nusa Tenggara, Indonesia
2
Graduate Program Dept. of Physics, Gadjah Mada University, Yogyakarta, Indonesia
3
Dept. of Physics Gadjah Mada University, Yogyakarta, Indonesia
Email: gbsuparta@ugm.ac.id
Key-words: micro-radiography, small specimen, noise ratio, image processing
Abstract. An image processing method has been developed for processing multiple images of x-ray
micro-radiography. An x-ray micro-radiography image reflects quantum mottle so that its
information content may tends to be corrupted. Therefore, a digital processing method has been
developed to reduce the effect of quantum mottle as well as reducing the noise level. A set of
radiographs are collected then summed. An image subtraction by a background image is carried out
prior to the summation process. The signal to noise ratio (SNR) and contrast to noise ratio (CNR)
after processing are compared with the SNR and CNR prior to the processing. As a result the final
image for small specimen under x-ray micro-radiography inspection is better than original image
without processing based on SNR and CNR assessments.
Introduction
In the process of acquiring digital images, noise always emerges as imperfection to the resulting
image. Noise causes degradation in image quality. Noise arises from the limitation of image
acquiring device during conversion from the light intensity or optical density into binary data[1].
Increasing noise tends to be more significant when the image is dark, in which light intensity or
optical density is low[2].
In x-ray micro-radiography system, the limitation of image acquiring device and the x-ray source
contribute noise in the form of quantum mottles [2]. Quantum mottles itself is stochastic in nature
as a result of randomly uneven photon emerged from the x-ray source. The effect of quantum
mottles is mostly inevitably. So, the common way to reduce that effect is suppressing it [2]. For this
reason, many image processing methods have been proposed to overcome quantum mottles. Most
of them are using filtering process and those apply for single image [3] and multiple images [4].
We have successfully developed an x-ray micro-radiography system at the Department of
Physics Gadjah Mada University Yogyakarta Indonesia. It comprises a laboratory x-ray generator
of a Molybdenum anode target as an x-ray radiation sources and a fluorescence screen that is
coupled with a CCD camera along with digitization apparatus as a detection unit. Some real time
digital radiographs have been obtained successfully for a number of study [5,6].
Recently, we have a study on earthenware material and ceramic material for pottery [4]. In
general, we study effects of particle size of material to the quality of pottery. One of our concerns is
whether there is effect on the particle size of powder clay to the noise level on the radiography
images. We presume that the smaller the particle size, the better the image we produced. We also
presume that the more coarse the particle the more random the diffusion interaction of the x-ray
radiation in the clay during radiography process. Such random interaction should decrease the
contrast of the radiography image.
This paper presents an attempt to reduce noise level on the x-ray micro-radiography images
resulted by our system that we have been developed previously. The noise may arise from x-ray
quantum mottle and material characteristics, e.g. particle size. The effect of noise in image quality
is evaluated using the Signal to Noise Ratio (SNR) and Contrast to Noise Ratio (CNR). Signal to
Noise Ratio (SNR) defines how well signals can be distinguished among measured noise [7], while
Advanced Materials Research Vol. 896 (2014) pp 676-680
© (2014) Trans Tech Publications, Switzerland
doi:10.4028/www.scientific.net/AMR.896.676
All rights reserved. No part of contents of this paper may be reproduced or transmitted in any form or by any means without the written permission of TTP,
www.ttp.net. (ID: 36.73.81.250-06/02/14,13:52:10)