• Journal of Digital Policy
• /
• v.3 no.1
• /
• pp.27-32
• /
• 2024

In this paper, proposes a super resolution method that enhances the quality of results by refining texture features, contrasting each, and utilizing the results as weights. For the improvement of quality, a precise and clear restoration result in details such as boundary areas is crucial in super resolution, along with minimizing unnecessary artifacts like noise. The proposed method constructs a residual block structure with multiple paths and skip-connections for feature estimation in conventional Convolutional Neural Network (CNN)-based super resolution methods to enhance quality. Additional learning is performed for sharpened and blurred image results for further texture analysis. By contrasting each super resolution result and allocating weights through this process, the proposed method achieves improved quality in detailed and smoothed areas of the image. The experimental results of the proposed method, evaluated using the PSNR and SSIM values as quality metrics, show higher results compared to existing algorithms, confirming the enhancement in quality.

• KIPS Transactions on Software and Data Engineering
• /
• v.9 no.10
• /
• pp.303-308
• /
• 2020

The ReLU(Rectified Linear Unit) function has been dominantly used as a standard activation function in most deep artificial neural network models since it was proposed. Later, Leaky ReLU, Swish, and Mish activation functions were presented to replace ReLU, which showed improved performance over existing ReLU function in image classification task. Therefore, we recognized the need to experiment with whether performance improvements could be achieved by replacing the RELU with other activation functions in the super resolution task. In this paper, the performance was compared by changing the activation functions in EDSR model, which showed stable performance in the super resolution task. As a result, in experiments conducted with changing the activation function of EDSR, when the resolution was converted to double, the existing activation function, ReLU, showed similar or higher performance than the other activation functions used in the experiment. When the resolution was converted to four times, Leaky ReLU and Swish function showed slightly improved performance over ReLU. PSNR and SSIM, which can quantitatively evaluate the quality of images, were able to identify average performance improvements of 0.06%, 0.05% when using Leaky ReLU, and average performance improvements of 0.06% and 0.03% when using Swish. When the resolution is converted to eight times, the Mish function shows a slight average performance improvement over the ReLU. Using Mish, PSNR and SSIM were able to identify an average of 0.06% and 0.02% performance improvement over the RELU. In conclusion, Leaky ReLU and Swish showed improved performance compared to ReLU for super resolution that converts resolution four times and Mish showed improved performance compared to ReLU for super resolution that converts resolution eight times. In future study, we should conduct comparative experiments to replace activation functions with Leaky ReLU, Swish and Mish to improve performance in other super resolution models.

• JSTS:Journal of Semiconductor Technology and Science
• /
• v.17 no.3
• /
• pp.465-472
• /
• 2017

Security has become one of the most important requirements for various devices for multi-sensor based embedded systems. The AES (Advanced Encryption Standard) algorithm is widely used for security, however, it requires high computing power. In order to reduce the CPU power for the data encryption of images, we propose a new image encryption module using hardware memoization, which can reuse previously generated data. However, as image pixel data are slightly different each other, the reuse rate of the simple memoization system is low. Therefore, we further apply an approximate concept to the memoization system to have a higher reuse rate by sacrificing quality. With the novel technique, the throughput can be highly improved by 23.98% with 14.88% energy savings with image quality loss minimization.

• Journal of Korea Multimedia Society
• /
• v.21 no.2
• /
• pp.108-116
• /
• 2018

In this paper, we propose an image enhancement method based on Multi-Scale Retinex theory that designs Unsharp Masking Filter (UMF) and emphasizes the contrast ratio adaptively. Unsharp Masking (UM) technique emphasizes image sharpness and improves contrast ratio by adding high frequency component to the original image. The high frequency component is obtained by differentiating between original image and low frequency image. In this paper, we present how to design an UMF kernel and to adaptively apply it to increase the contrast ratio according to multi-scale retinex theory which resembles human visual system. Experimental results show that the proposed method has better quantitative performance indexes such as PSNR, ambe & SSIM and better qualitative feature like halo artifact suppression.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2019.11a
• /
• pp.228-229
• /
• 2019

Digital refocusing and perspective change are the most expected applications of light field (LF) images. As LF image has a large amount of data, its compression is very essential. The fidelity of LF image after compression needs to be evaluated differently depending on a specific application such as perspective change or refocusing. In this paper, we investigate the fidelity of images after perspective change and refocusing. Several state-of-the-art objective quality metrics are compared. Our experiment shows that IWPSNR is the most reliable metric for both perspective and focus changes, but it does not outperform the popular metrics such as PSNR and SSIM.

• Proceedings of the KAIS Fall Conference
• /
• 2011.12b
• /
• pp.441-444
• /
• 2011

본 논문은 2D/3D 영상 처리에서 깊이지도 생성을 위한 Optical flow에서 레벨을 간소화하여 연산량을 감소시키고 객체의 고유벡터를 이용하여 영상의 잡음을 제거하는 연구이다. Optical flow는 움직임추정 알고리즘의 하나로 두 프레임간의 픽셀의 변화 벡터 값을 나타내며 블록 매칭과 같은 알고리즘에 비해 정확도가 높다. 그러나 기존의 Optical flow는 긴 연산 시간과 카메라의 이동이나 조명의 변화에 민감한 문제가 있다. 이를 해결하기 위해 연산 시간의 단축을 위한 레벨 간소화 과정을 거치고 영상에서 고유벡터를 갖는 영역에 한해 Optical flow를 적용하여 잡음을 제거하는 방법을 제안하였다. 제안한 방법으로 2차원 영상을 3차원 입체 영상으로 변환하였고 SSIM(Structural SIMilarity Index)으로 최종 생성된 영상의 오차율을 분석하였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.9 no.10
• /
• pp.4160-4176
• /
• 2015

Data compression like image and video compression has come a long way since the introduction of Compressive Sensing (CS) which compresses sparse signals such as images, videos etc. to very few samples i.e. M < N measurements. At the receiver end, a robust and efficient recovery algorithm estimates the original image or video. Many prominent algorithms solve least squares problem (LSP) iteratively in order to reconstruct the signal hence consuming more processing time. In this paper non-iterative threshold based recovery algorithm (NITRA) is proposed for the recovery of images and videos without solving LSP, claiming reduced complexity and better reconstruction quality. The elapsed time for images and videos using NITRA is in ㎲ range which is 100 times less than other existing algorithms. The peak signal to noise ratio (PSNR) is above 30 dB, structural similarity (SSIM) and structural content (SC) are of 99%.

• Journal of Korea Society of Digital Industry and Information Management
• /
• v.19 no.2
• /
• pp.57-68
• /
• 2023

Because salt and pepper noise is a type of impulse, even a small amount of noise could cause a large image degradation. In this paper, we proposed a salt-and-pepper noise removal method using the convolutional neural network. It consists of four phases. In the first step, the proposed method reconstructs noisy image using a traditional salt-and-pepper noise reduction method, and in the second step, the result image of previous step is filtered with Gaussian low pass filter. After that, we reconstruct the filtered image using convolution neural network. In the last step, the pixels with salt-and-pepper noise are replaced with the result of previous phase. Simulation results show that the proposed method yields not only objective image qualities(PSNR, SSIM) but also subjective image qualities for all SAP noise ratios.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2021.06a
• /
• pp.212-215
• /
• 2021

본 논문에서는 6 Fisheye lens 원본 영상에 대하여 Insta360 stitcher, AutoStitch[4], As-Projective-AsPossible(APAP)[5] 스티칭 방법으로 360 도 파노라마 영상을 생성하고 기하학적 왜곡과 컬러 왜곡을 비교 평가한다. 360 도 파노라마 Image Quality Assessment(IQA) 메트릭으로 Natural Image Quality Evaluator(NIQE)[6], Blind/Referenceless Image Spatial Quality Evaluator (BRISQUE)[7], Perception based Image Quality Evaluator(PIQE)[8], Feature Similarity(FSIM)[9] 그리고 high frequency feature 에 대한 Structural Similarity(SSIM)[10]을 측정하여 정량적 평가를 하며 정성적인 비교를 통하여 파노라마 영상의 품질과 평가 메트릭에 대한 벤치마크를 제공한다.

• Journal of Information Processing Systems
• /
• v.19 no.3
• /
• pp.275-288
• /
• 2023

Many image edge details are easily lost in the image denoising process, and the smooth image regions are prone to produce jagged. In this paper, we propose a wavelet-based image global k- singular value decomposition variational method to remove image noise. A layer of wavelet decomposition is applied to the noisy image first. Then, the image global k-singular value decomposition (IGK-SVD) method is used to remove the random noise of low-frequency components. Furthermore, a constructed variational denoising method (VDM) removes the random noise in the high-frequency component. Finally, the denoised image is obtained by wavelet reconstruction. The experimental results show that the proposed method's peak signal-to-noise ratio (PSNR) value is higher than other methods, and its structural similarity (SSIM) value is closer to one, indicating that the proposed method can effectively suppress image noise while retaining more image edge details. The denoised image has better denoising effects.