• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.10
• /
• pp.4968-4986
• /
• 2017

In this paper, we primarily address the difficulty of automatic generation of a plausible depth map from a single image in an unstructured environment. The aim is to extrapolate a depth map with a more correct, rich, and distinct depth order, which is both quantitatively accurate as well as visually pleasing. Our technique, which is fundamentally based on a preexisting DepthTransfer algorithm, transfers depth information at the level of superpixels. This occurs within a framework that replaces a pixel basis with one of instance-based learning. A vital superpixels feature enhancing matching precision is posterior incorporation of predictive semantic labels into the depth extraction procedure. Finally, a modified Cross Bilateral Filter is leveraged to augment the final depth field. For training and evaluation, experiments were conducted using the Make3D Range Image Dataset and vividly demonstrate that this depth estimation method outperforms state-of-the-art methods for the correlation coefficient metric, mean log10 error and root mean squared error, and achieves comparable performance for the average relative error metric in both efficacy and computational efficiency. This approach can be utilized to automatically convert 2D images into stereo for 3D visualization, producing anaglyph images that are visually superior in realism and simultaneously more immersive.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.40 no.3
• /
• pp.194-202
• /
• 2003

To convert printed images into electronic ones automatically, it requires region analysis of document images and character recognition. In these, regional analysis segments document image into detailed regions and classifies thee regions into the types of text, picture, table and so on. But it is difficult to classify the text and the picture exactly, because the size, density and complexity of pixel distribution of some of these are similar. Thu, misclassification in region analysis is the main reason that makes automatic conversion difficult. In this paper, we propose region analysis method that segments document image into text and picture regions. The proposed method solves the referred problems using one dimensional median filter based method in text and picture classification. And the misclassification problems of boldface texts and picture regions like graphs or tables, caused by using median filtering, are solved by using of skin peeling filter and maximal text length. The performance, therefore, is better than previous methods containing commercial softwares.

• Journal of Biomedical Engineering Research
• /
• v.28 no.6
• /
• pp.756-767
• /
• 2007

For the combination of phosphor screens having various thicknesses and a photodiode array manufactured by complementary metal-oxide-semiconductor (CMOS) process, we report the observation of image-quality degradation under the irradiation of 45-kVp spectrum x rays. The image quality was assessed in terms of dark pixel signal, dynamic range, modulation-transfer function (MTF), noise-power spectrum (NPS), and detective quantum efficiency (DQE). For the accumulation of the absorbed dose, the radiation-induced increase both in dark signal and noise resulted in the gradual reduction in dynamic range. While the MTF was only slightly affected by the total ionizing dose, the noise power in the case of $Min-R^{TM}$ screen, which is the thinnest one among the considered screens in this study, became larger as the total dose was increased. This is caused by incomplete correction of the dark current fixed-pattern noise. In addition, the increase tendency in NPS was independent of the spatial frequency. For the cascaded model analysis, the additional noise source is from direct absorption of x-ray photons. The change in NPS with respect to the total dose degrades the DQE. However, with carefully updated and applied correction, we can overcome the detrimental effects of increased dark current on NPS and DQE. This study gives an initial motivation that the periodic monitoring of the image-quality degradation is an important issue for the long-term and healthy use of digital x-ray imaging detectors.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.25 no.1
• /
• pp.78-84
• /
• 2015

Depth estimation is often required for robot vision, 3D modeling, and motion control. Previous method is based on the focus measures which are calculated for a series of image by a single camera at different distance between and object. This method, however, has disadvantage of taking a long time for calculating the focus measure since the mask operation is performed for every pixel in the image. In this paper, we estimates the depth by using the focus measure of the boundary pixels located between the objects in order to minimize the depth estimate time. To detect the boundary of an object consisting of a straight line and a circle, we use the Hough transform and estimate the depth by using the focus measure. We performed various experiments for PCB images and obtained more effective depth estimation results than previous ones.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.24 no.12
• /
• pp.1128-1139
• /
• 2013

KOMPSAT-5(KOrea Multi-Purpose SATellite-5) has a benefit of continuously conducting its mission in all weather and even night by loading SAR(Synthetic Aperture Radar) payload, which is different from optical sensor of KOMPSAT-2 satellite. During IOT(In-Orbit Test) periods, SAR image calibration should be conducted through ground target of which location and RCS is pre-determined. Differently from the conventional corner reflector, active transponder has a capability to change its internal transfer gain and delay, which allows active transponder to be shown in a pixel of SAR image with very high radiance and virtual location. In this paper, the development of active transponder is presented from design to I&T(Integration and Test).

• Progress in Medical Physics
• /
• v.28 no.3
• /
• pp.83-91
• /
• 2017

In this work, we investigated the recently proposed phase-contrast x-ray imaging (PCXI) technique, the so-called single grid-based PCXI, which has great simplicity and minimal requirements on the setup alignment. It allows for imaging of smaller features and variations in the examined sample than conventional attenuation-based x-ray imaging with lower x-ray dose. We performed a systematic simulation using a simulation platform developed by us to investigate the image characteristics. We also performed a preliminary PCXI experiment using an established a table-top setup to demonstrate the performance of the simulation platform. The system consists of an x-ray tube ($50kV_p$, 5 mAs), a focused-linear grid (200-lines/inch), and a flat-panel detector ($48-{\mu}m$ pixel size). According to our results, the simulated contrast of phase images was much enhanced, compared to that of the absorption images. The scattering length scale estimated for a given simulation condition was about 117 nm. It was very similar, at least qualitatively, to the experimental contrast, which demonstrates the performance of the simulation platform. We also found that the level of the phase gradient of oriented structures strongly depended on the orientation of the structure relative to that of linear grids.

• Journal of Internet Computing and Services
• /
• v.8 no.5
• /
• pp.117-123
• /
• 2007

Catadioptric mirrors are widely used in automatic surveillance system. The major drawback of catadioptric mirror is its unequal image resolution. Equidistance catadioptric mirrir can be the solution to this problem. The exact axial alignment and the exact mount of mirror are the sources that can be avoided but the focal length variation is inevitable. In this paper, the effects of focal length variation on the computation of depth and height of object' point are explained and the effective and simple focal length finding algorithm is presented. First two object's points are selected, and the counterparts on the other stereo image are to be found using MSE criterion. Using four pixel distance from the image center, the assumed focal length is calculated. If the obtained focal length is different from the exact focal length, 24mm, the focal length value is modified by the proposed method. The method is very simple and gives the comparable results with the earlier sophisticated method.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.26 no.2
• /
• pp.103-110
• /
• 2008

General lens, widely using in Photogrammetry, has narrow view, and have to adjust "Image-Registration Method" after obtain images and it need cost; economic, period of time. Recent days, there is various study that use wide-angle lens, usually for robotics field, put to practical use in photogrammetry instead of general lens. In this studies, distortion tendency of wide-angle lens and utilize the correction techniques suitable to wide-angle lens by the existing photographic survey methods. After carrying out the calibration of the wide-angle lens, we calculated the correction parameters, and then developed the method that convert the original image-point to new image-point correcting distortion. For authorization the developed algorithm, we had inspection about shape and position; there are approximately 2D RMSE of 3 pixel, cx = 2, and cy = 3 different.

• Journal of Korea Multimedia Society
• /
• v.23 no.5
• /
• pp.650-657
• /
• 2020

The rapid growth of internet users and faster network speed are driving the new ICT services. ICT Technology has improved our way of thinking and style of life, but it has created security problems such as malware, ransomware, and so on. Therefore, we should research against the increase of malware and the emergence of malicious code. For this, it is necessary to accurately and quickly detect and classify malware family. In this paper, we analyzed and classified visualization technology, which is a preprocessing technology used for deep learning-based malware classification. The first method is to convert each byte into one pixel of the image to produce a grayscale image. The second method is to convert 2bytes of the binary to create a pair of coordinates. The third method is the method using LSH. We proposed improving the technique of using the entire existing malicious code file for visualization, extracting only the areas where important information is expected to exist and then visualizing it. As a result of experimenting in the method we proposed, it shows that selecting and visualizing important information and then classifying it, rather than containing all the information in malicious code, can produce better learning results.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.21 no.7
• /
• pp.1342-1348
• /
• 2017

In an era where digital data takes on great importance, images are essential to various media. Noise is generated during the acquisition and transmission of such images, caused by a number of external factors. The removal of noise is an essential step in image processing. There are various methods used to remove noise, in accordance with the cause or form of the noise. AWGN is one of the leading methods. As such, this paper applies the edge detection method using the mean of each pixel after categorizing in detail the partial masks into nine areas as part of the preliminary process, in order to minimize noise that had been added to the image. In addition, the paper suggests an algorithm that applies different filters to the partial masks by using the critical mass value of the transfigured edge detection. To verify the competence of the suggested algorithm, it was compared with existing methods by using magnified images and PSNR(peak signal to noise ratio).