• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2011.11a
• /
• pp.321-324
• /
• 2011

본 논문에서는 contourlet 변환을 이용하여 잡음을 제거하는 방법을 제안한다. 영상 센서의 발전으로 이미지의 해상도가 좋아지는 반면 잡음에 민감해진다. 그러므로 이를 전처리 단계에서 처리해주는 것이 필요하다. 잡음은 주로 자연 영상의 윤곽선에서 민감하게 반응하기 때문에 고주파대의 잡음을 최대한 정확하게 제거하는 과정이 중요하다. Contourlet 변환은 기존의 wavelet 변환의 다중 스케일과 더불어 다양한 방향 필터뱅크를 이용하여 방향 성분에 대하여 풍부한 정보를 얻을 수 있는 변환이다. 영상의 화이트 가우시안 잡음을 제거하기 위해 contourlet 변환 영역에서의 계수를 이변수 가우스 확률 모델로 설정하고 Bayes 추정법을 사용한다. Bayes 추정법에 필요한 파라미터들은 근사적으로 추정한다. 제안한 방식을 통하여 잡음이 제거된 영상에 추가적으로 Wiener filter와 cycle-spinning을 적용하여 더 높은 PSNR (peak signal-to-noise ratio)값을 얻을 수 있다. 모의실험을 통해 제안한 방식의 PSNR 값과 결과영상으로 성능이 우수함을 확인하였다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2019.06a
• /
• pp.298-301
• /
• 2019

이미지 스티칭은 다수의 이미지를 합성하여 카메라의 좁은 시야각(Field of View) 문제를 해결하는 기술이다. 최근 동영상 기반 Panorama, Super Resolution, 360 VR (Virtual Reality) 등의 컨텐츠 사용이 증가함에 따라, 보다 빠르고 정확한 이미지 스티칭 기술의 필요성이 커지고 있다. 지금까지 필요 성능을 만족시키기 위해 많은 알고리즘이 제안되고 있지만, 정확성을 측정하는 객관적 평가 방법은 표준화되지 않고 있다. 최근에서야 PSNR (Peak Signal-to-Noise Ratio) 과 SSIM (Structural Similarity index method) 측정값을 제시하는 방법이 주를 이루고 있지만, 본 논문에서는 PSNR 과 SSIM 측정 방식의 문제점을 밝히고 지역 차분 픽셀 평가 방법을 제안한다. 기하적 유사성과 광도 측정 정보를 포괄하는 LDPM(Local Differential Pixel Mean) 평가 방식을 테스트 이미지를 통해 증명하고 SSIM 과 비교를 통해 해당 평가 방법의 이점을 밝힌다.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.3
• /
• pp.25-33
• /
• 2023

Noise generated during the acquisition and transmission of CT images acts as a factor that degrades image quality. Therefore, noise removal to solve this problem is an important preprocessing process in image processing. In this paper, we remove noise by using a deformable convolutional autoencoder (DeCAE) model in which deformable convolution operation is applied instead of the existing convolution operation in the convolutional autoencoder (CAE) model of deep learning. Here, the deformable convolution operation can extract features of an image in a more flexible area than the conventional convolution operation. The proposed DeCAE model has the same encoder-decoder structure as the existing CAE model, but the encoder is composed of deformable convolutional layers and the decoder is composed of conventional convolutional layers for efficient noise removal. To evaluate the performance of the DeCAE model proposed in this paper, experiments were conducted on CT images corrupted by various noises, that is, Gaussian noise, impulse noise, and Poisson noise. As a result of the performance experiment, the DeCAE model has more qualitative and quantitative measures than the traditional filters, that is, the Mean filter, Median filter, Bilateral filter and NL-means method, as well as the existing CAE models, that is, MAE (Mean Absolute Error), PSNR (Peak Signal-to-Noise Ratio) and SSIM. (Structural Similarity Index Measure) showed excellent results.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.51 no.3
• /
• pp.112-118
• /
• 2014

As higher quality of image is required for digital image scaling, longer processing time is required. Therefore the technology that can make higher quality image quickly is needed. We propose the double linear-cubic convolution interpolation which creates the high quality image with low complexity and hardware resources. The proposed interpolation methods which are made up of four one-dimensional linear interpolations and one one-dimensional cubic convolution perform linear-cubic convolution interpolation in horizontal and vertical direction. When compared in aspects of peak signal-to-noise ratio(PSNR), performance time and amount of hardware resources, the proposed interpolation provided better PSNR, low complexity and less hardware resources than bicubic convolution interpolation.

• Journal of information and communication convergence engineering
• /
• v.9 no.6
• /
• pp.763-768
• /
• 2011

Image denoising is a lively research field. Today the researches are focus on the wavelet domain especially using wavelet threshold method. We proposed an adaptive threshold method which considering the characteristic of different sub-band, the method is adaptive to each sub-band. Experiment results show that the proposed method extracts white Gaussian noise from original signals in each step scale and eliminates the noise effectively. In addition, the method also preserves the detail information of the original image, obtaining superior quality image with higher peak signal to noise ratio(PSNR).

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2015.10a
• /
• pp.86-89
• /
• 2015

Quantitative Evaluation of Image Quality using Automatic Exposure Control This it leads, the relationship of control and image between of photographing condition, DR proper use method proposing. Chest phantom acquires Chest PA images which it follows in conditional change, Image evaluation factor (SNR (Signal to Noise Ratio), CNR (Contrast to Noise Ratio), PSNR (Peak Signal to Ratio), RMS (Root Mean Square)). It excepted RMS price, SNR, CNR, PSNR the case which uses AEC it came out being high The fact that the quality of image is better there was a case which does not use AEC. but Price it was slight. Image qualitative it was deficient in AEC use presence. Through this, the overall situation and most efficient use of radiation workers is better suited. But Passive AEC use would be appropriate for use rather than proactive AEC.

• Journal of Digital Contents Society
• /
• v.5 no.1
• /
• pp.87-94
• /
• 2004

In this paper, the digital watermarking algorithm is proposed using CDMA technique for protection of the mobile contents in the mobile environment. The digital watermarking was designed to robust the errors in the mobile environment where pathloss, multipath fading, interference, and noise exist. In case of the multimedia content service in the mobile environment, the construction method of the watermark, the algorithm of insertion and detection are also proposed. The watermark consists of the information of the mobile user. Invisibility and robustness required in watermarking are etimated. It is observed that PSNR of the mobile content inserted the watermark is 90.31 dB, and the signal processing and noise attack are also robust. Especially, because random noise occurs in wireless transmission can overcome, the proposed watermarking algorithm is adequate for protection of the multimedia contents in the mobile environment.

• Journal of the Korean Magnetics Society
• /
• v.26 no.3
• /
• pp.92-98
• /
• 2016

To aim of this study was to assess the full scan and half scan of imaging with half scan factor. Patients without a cerebral vascular disease (n = 30) and were subject to the full scan half scan, and set a region of interest in the cerebral artery from the three regions (C1, C2, C3) in the range of 7 to 8 mm. MIP (maximum intensity projection) to reconstruct the images in signal strength SNR (signal to noise ration), PSNR (peak signal noise to ratio), RMSE (root mean square error), MAE (mean absolute error) and calculated by paired t-test for use by statistics were analyzed. Scan time was half scan (4 minutes 53 seconds), the full scan (6 minutes 04 seconds). The mean measurement range (7.21 mm) of all the ROI in the brain blood vessel, was the SNR of the first C1 is completely scanned (58.66 dB), half-scan (62.10 dB), a positive correlation ($r^2=0.503$), for the second C2 SNR is completely scanned (70.30 dB), half-scan (74.67 dB) the amount of correlation ($r^2=0.575$), third C3 of a complete scan SNR (70.33 dB), half scan SNR (74.64 dB) in the amount of correlation between the It was analyzed with ($r^2=0.523$). Comparative full scan with half of SNR ($4.75{\pm}0.26dB$), PSNR ($21.87{\pm}0.28dB$), RMSE ($48.88{\pm}1.61$), was calculated as MAE ($25.56{\pm}2.2$). SNR is also applied to examine the half-scans are not many differences in the quality of the two scan methods were not statistically significant in the scan (p-value > .05) image takes less time than a full scan was used.

• Proceedings of the Korea Contents Association Conference
• /
• 2003.11a
• /
• pp.120-123
• /
• 2003

In this paper, a digital watermarking scheme whichis effective in protecting a copyright of video image under an image transformation and impulse noise is proposed. The proposing scheme is to use a coded watermark that insert the personal ID of copyrighter. The recovery ability is improved by the coded watermark. Also the coded watermark is abel to trace the illegal distributors. Binary image is used as watermark image, the value of PSNR and recovered rates of watermark are obtained in order to confirm the required invisibility and robustness in watermark system. The experimental results show that image quality is less degraded as the PSNR of 98.21 ㏈. It is also observed that excellent watermark recovery is achieved under the image transformation and impulse noise.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2021.06a
• /
• pp.176-179
• /
• 2021

최근 딥 러닝 기법의 하나인 합성곱 신경망(Convolutional Neural Network, CNN)은 영상 잡음(Noise) 제거 분야에서 전통적인 기법보다 좋은 성능을 나타내고 있지만 학습하는 과정에서 영상 내 디테일한 부분이 손실될 수 있다. 본 논문에서는 웨이블릿 변환(Wavelet Transform)을 기반으로 영상 내 디테일 정보도 같이 학습하여 영상 디테일을 향상하는 잡음 제거 합성곱 신경망 네트워크를 제안한다. 제안하는 네트워크는 디테일 향상 서브 네트워크(Detail Enhancement Subnetwork)와 영상 잡음 추출 서브 네트워크(Noise Extraction Subnetwork)를 이용하게 된다. 실험을 통해 제안하는 방법은 기존 알고리듬보다 디테일 손실 문제를 효과적으로 해결할 수 있었고 객관적 품질 평가인 PSNR(Peak Signal-to-Noise Ratio)와 주관적 품질 비교에서 모두 우수한 결과가 나온 것을 확인하였다.