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)