• Annual Conference of KIPS
• /
• 2010.11a
• /
• pp.1118-1121
• /
• 2010

With the advent of cell phone digital cameras, and sophisticated photo editing software, digital images can be easily manipulated and altered. Although good forgeries may leave no visual clues of having been tampered with, they may, nevertheless, alter the underlying statistics of an image. Most digital camera equipped in cell phones employ a single image sensor in conjunction with a color filter array (CFA), and then interpolates the missing color samples to obtain a three channel color image. This interpolation introduces specific correlations which are likely to be destroyed when tampering with an image. We quantify the specific correlations introduced by CFA interpolation, and describe how these correlations, or lack thereof, can be automatically detected in any portion of an image. We show the efficacy of this approach in revealing traces of digital tampering in lossless and lossy compressed color images interpolated with several different CFA algorithms in test cell phones.

• Journal of electromagnetic engineering and science
• /
• v.14 no.4
• /
• pp.405-410
• /
• 2014

In this paper, we compare the quality of images reconstructed by a conventional delayed-sum (DS) algorithm and radiation pattern-based DS algorithm. In order to evaluate the quality of images, we apply the target-to-clutter ratio (TCR), which is commonly used in synthetic aperture radar (SAR) image assessment. The radiation pattern-based DS algorithm enhances the TCR of the image by focusing the target signals and preventing contamination of the radar scene. We first consider synthetic data obtained through GprMax2D/3D, a finite-difference time-domain (FDTD) forward solver. Experimental data of a 2-GHz bandwidth stepped-frequency signal are collected using a vector network analyzer (VNA) in an anechoic chamber setup. The radiation pattern-based DS algorithm shows a 6.7-dB higher TCR compared to the conventional DS algorithm.

• The Bulletin of The Korean Astronomical Society
• /
• v.35 no.1
• /
• pp.70.2-70.2
• /
• 2010

We present near- and mid-infrared images of the Crab Nebula, taken with the Infrared Camera (IRC) onboard the AKARI infrared space telescope. These images have a field-of-view of 10'*10' and show the full extent of the nebula at 3, 4, 7, 11, 15, and 24 um. The Crab nebula in near infrared is dominated by synchrotron emission while, in mid infrared, the ionic forbidden lines of Ar, Ne, S, and Fe makes significant contribution. We separate the line emission from synchrotron emission in 3-15 um AKARI bands using the ISOCAM CVF data, and present separate images for the line and synchrotron emissions in each band. We derive the total synchrotron fluxes of the Crab nebula in these bands, which are used to complete the synchrotron spectral energy distribution of the Crab nebula from radio to X-rays. We discuss the spectral variations of the Crab nebula.

• Journal of Fashion Business
• /
• v.16 no.6
• /
• pp.52-62
• /
• 2012

This study researched the jeans street fashion for young men looked like 20s and 30s in Seoul, and analyzed jeans by silhouette by fit and color by washing. The photos were of jeans from the waist to the bottom; and among a total of 1,159 images, 847 images that could be analyzed were selected and used for the analysis. From the images used in the analysis, the jeans fits were categorized according to silhouette and color that were used for the visual selection. As a result, young men preferred fashion jeans like slim or skinny jeans. Most of the tight fitting skinny and slim jeans were slim baggy style jeans, like low waist & long rise with a low back york, which was a key trendy pants style. The jeans color analysis by washing finish showed that dark jeans were most frequently, most of the dark jeans had the original denim color based on the non or normal wash techniques. It is hoped that the results of this study can be used as basic data for the development of designs for jeans brands.

• Journal of the Korean Institute of Telematics and Electronics
• /
• v.24 no.1
• /
• pp.35-45
• /
• 1987

For the registration problem in which the matching of two images is made, a new algorithm using an 1-D LPC model was proposed. The proposed algorithm employed LPC coefficients as feature vector of an image. The similarity of two images was measured using an LPC distance, proposed by Itakura, between each image's feature vector. The comparision of performance with normalized correlation method and template matching method was made by a computer simulation with several real images. The results of simulation showed that the proposed algorithm was more robust to image intensity variation and computationall efficient.

• Proceedings of the KOSOMBE Conference
• /
• v.1997 no.05
• /
• pp.79-83
• /
• 1997

A new computation algorithm to detect the intimal and adventitial contours from the intravascular images was developed. An image processing on gray level image was used. It uses arrays of pixels in each radial lines on the images. A Low-Pass filters was adopted at first step for one dimensional image processing. Some other calculations techniques were developed to increase the accuracy of automatically detected contours. The program implemented in C++ as a window95-base application.

• Proceedings of the KSRS Conference
• /
• 2004.10a
• /
• pp.689-692
• /
• 2004

Hydrographic data obtained from several ship experiments have shown that internal waves are frequently observed in the East (Japan) Sea, mostly in the continental slopes located along the western side of the East Sea. It is well known that oceanic internal waves can be detected well in synthetic aperture radar (SAR) images. Interactions between surface capillary-gravity waves and horizontally varying surface currents induced by internal waves produce variations in sea surface roughness which can be detected by SAR. C-band SAR images from ERS, ENVISAT ASAR and RADARSAT have been used to study the characteristics of internal waves in the East Sea. The observed properties of internal waves from many SAR images were compared and verified from in-situ measurements and theories.

• Journal of The Korean Astronomical Society
• /
• v.27 no.2
• /
• pp.191-201
• /
• 1994

We propose to use the entropy of power spectra defined in the frequency domain for the deconvolution of extended images. Spatial correlations requisite for extended sources may be insured by increasing the role of power entropy because the power is just a representation of spatial correlations in the frequency domain. We have derived a semi-analytical solution which is found to severely reduce computing time compared with other iteration schemes. Even though the solution is very similar to the well-known Wiener filter, the regularizingng term in the new expression is so insensitive to the noise characteristics as to assure a stable solution. Applications have been made to the IRAS $60{\mu}m\;and\;100{\mu}m$ images of the dark cloud B34 and the optical CCD image of a solar active region containing a circular sunspot and a small pore.

• Imaging Science in Dentistry
• /
• v.35 no.3
• /
• pp.157-165
• /
• 2005

Purpose : To find out the proper compression ratios without degrading image quality and affecting lesion detectability on diagnostic images used in dentistry compressed with JPEG 2000 algorithm. Materials and Methods : Sixty Digora periapical images, sixty panoramic computed radiographic (CR) images, sixty computed tomographic (CT) images, and sixty magnetic resonance (MR) images were compressed into JPEG 2000 with ratios of 10 levels from 5:1 to 50:1. To evaluate the lesion detectability, the images were graded with 5 levels (1 : definitely absent; 2: probably absent; 3: equivocal; 4: probably present; 5: definitely present), and then receiver operating characteristic analysis was performed using the original image as a gold standard. Also to evaluate subjectively the image quality, the images were graded with 5 levels (1 definitely unacceptable; 2: probably unacceptable; 3: equivocal, 4: probably acceptable; 5· definitely acceptable), and then paired t-test was performed. Results : In Digora, CR panoramic and CT images, compressed images up to ratios of 15 : 1 showed nearly the same lesion detectability as original images, and in MR images, compressed images did up to ratios of 25 : 1. In Digora and CR panoramic Images, compressed images up to ratios of 5 : 1 showed little difference between the original and reconstructed images in subjective assessment of image quality In CT images, compressed images did up to ratios of 10: 1 and in MR images up to ratios of 15 : 1 Conclusion : We considered compression ratios up to 5 : 1 in Digora and CR panoramic images, up to 10 : 1 in CT images, up to 15 : 1 in MR images as clinically applicable compression ratios.

• IEIE Transactions on Smart Processing and Computing
• /
• v.3 no.4
• /
• pp.191-194
• /
• 2014

This paper proposes a fast and efficient method to extract the skin and bone automatically in CT images. First, the images were smoothed by applying an anisotropic diffusion filter to remove noise. The whole body was then detected by thresholding, which was set automatically. In addition, the contour of the skin was segmented using morphological operators and connected component labeling (CCL). Finally, the bone was extracted by iterative thresholding.