• 한국컴퓨터정보학회논문지
• /
• 제22권3호
• /
• pp.45-51
• /
• 2017

In this paper, we propose a consecutive-frame super-resolution method to tackle a moving object problem. The super-resolution is a method restoring a high resolution image from a low resolution image. The super-resolution is classified into two types, briefly, single-frame super-resolution and consecutive-frame super-resolution. Typically, the consecutive-frame super-resolution recovers a better than the single-frame super-resolution, because it use more information from consecutive frames. However, the consecutive-frame super-resolution failed to recover the moving object. Therefore, we proposed an improved method via moving object detection. Experimental results showed that the proposed method restored both the moving object and the background properly.

• Journal of Information Processing Systems
• /
• 제18권1호
• /
• pp.12-25
• /
• 2022

In recent years, deep convolutional neural networks have made significant progress in the research of single image super-resolution. However, it is difficult to be applied in practical computing terminals or embedded devices due to a large number of parameters and computational effort. To balance these problems, we propose CSRNet, a lightweight neural network based on channel split residual learning structure, to reconstruct highresolution images from low-resolution images. Lightweight refers to designing a neural network with fewer parameters and a simplified structure for lower memory consumption and faster inference speed. At the same time, it is ensured that the performance of recovering high-resolution images is not degraded. In CSRNet, we reduce the parameters and computation by channel split residual learning. Simultaneously, we propose a double-upsampling network structure to improve the performance of the lightweight super-resolution network and make it easy to train. Finally, we propose a new evaluation metric for the lightweight approaches named 100_FPS. Experiments show that our proposed CSRNet not only speeds up the inference of the neural network and reduces memory consumption, but also performs well on single image super-resolution.

• International Journal of Computer Science & Network Security
• /
• 제21권11호
• /
• pp.17-22
• /
• 2021

Recently, image super-resolution techniques used in convolutional neural networks (CNN) have led to remarkable performance in the research area of digital image processing applications and computer vision tasks. Convolutional layers stacked on top of each other can design a more complex network architecture, but they also use more memory in terms of the number of parameters and introduce the vanishing gradient problem during training. Furthermore, earlier approaches of single image super-resolution used interpolation technique as a pre-processing stage to upscale the low-resolution image into HR image. The design of these approaches is simple, but not effective and insert the newer unwanted pixels (noises) in the reconstructed HR image. In this paper, authors are propose a novel single image super-resolution architecture based on synchronized depthwise separable convolution with Dense Skip Connection Block (DSCB). In addition, unlike existing SR methods that only rely on single path, but our proposed method used the synchronizes path for generating the SISR image. Extensive quantitative and qualitative experiments show that our method (SDCN) achieves promising improvements than other state-of-the-art methods.

• 한국방송∙미디어공학회:학술대회논문집
• /
• 한국방송∙미디어공학회 2019년도 하계학술대회
• /
• pp.98-101
• /
• 2019

At present, deep convolutional neural networks have made a very important contribution in single-image super-resolution. Through the learning of the neural networks, the features of input images are transformed and combined to establish a nonlinear mapping of low-resolution images to high-resolution images. Some previous methods are difficult to train and take up a lot of memory. In this paper, we proposed a simple and compact deep recursive residual network learning the features for single image super resolution. Global residual learning and local residual learning are used to reduce the problems of training deep neural networks. And the recursive structure controls the number of parameters to save memory. Experimental results show that the proposed method improved image qualities that occur in previous methods.

• International Journal of Computer Science & Network Security
• /
• 제21권12spc호
• /
• pp.463-468
• /
• 2021

Single image super-resolution (SISR) is an image processing technique, and its main target is to reconstruct the high-quality or high-resolution (HR) image from the low-quality or low-resolution (LR) image. Currently, deep learning-based convolutional neural network (CNN) image super-resolution approaches achieved remarkable improvement over the previous approaches. Furthermore, earlier approaches used hand designed filter to upscale the LR image into HR image. The design architecture of such approaches is easy, but it introduces the extra unwanted pixels in the reconstructed image. To resolve these issues, we propose novel deep learning-based approach known as Lightweight deep CNN-based approach for Single Image Super-Resolution (LDCSIR). In this paper, we propose a new architecture which is inspired by ResNet with Inception blocks, which significantly drop the computational cost of the model and increase the processing time for reconstructing the HR image. Compared with the other state of the art methods, LDCSIR achieves better performance in terms of quantitively (PSNR/SSIM) and qualitatively.

• 방송공학회논문지
• /
• 제24권5호
• /
• pp.703-712
• /
• 2019

단일 영상 초해상도 (Single Image Super-Resolution - SISR)기법은 카메라로 획득된 저해상도 영상에 필터 기반의 연산을 적용하여 좋은 화질의 고해상도 영상을 복원하는 과정이다. 최근에 심층 합성곱 신경망 학습의 발전에 따라 단일 영상 초해상도에 적용되는 심층 학습 기법들은 좋은 성과를 보여 주고 있다. 그 대표적인 방법으로 영상의 특징 맵 기반 웨이블릿 계수 학습을 통해 고해상도 영상을 복원하는 WaveletSRNet이 있다. 하지만 복잡한 알고리즘으로 인해 계산량이 증대되어 처리 속도가 늦고 특징 추출할 때 특징 맵을 효율적으로 활용하지 못 한다는 단점을 가지고 있다. 이를 개선하기 위해 본 논문에서는 단일 영상 초해상도 RDB-WaveletSRNet 기법을 제안한다. 제안된 기법은 잔여밀집블록(Residual Dense Block)을 사용하여 저해상도의 특징 맵을 효과적으로 추출하여 초해상도의 성능을 향상시키고 적절한 성장률을 설정하여 복잡한 계산량 문제까지 해결하였다. 또한 웨이블릿 패킷 분해를 사용하여 확대율에 맞게 웨이블릿 계수를 획득하므로 높은 확대율의 단일 영상 초해상도를 얻게 하였다. 다양한 영상에 대한 실험을 통하여, 제안하는 기법이 기존 기법보다 수행시간이 빠르며 영상 품질도 우수함을 입증하였다. 제안하는 방법은 기존 방법보다 화질은 PSNR 0.1813dB만큼 우수하며 속도는 1.17배 빠른 것을 실험을 통해 확인하였다.

• 대한임베디드공학회논문지
• /
• 제15권2호
• /
• pp.61-70
• /
• 2020

Convolutional Neural Networks (CNN) have been used extensively in recent times to solve image classification and segmentation problems. However, the use of CNNs in image super-resolution problems remains largely unexploited. Filter interpolation and prediction model methods are the most commonly used algorithms in super-resolution algorithm implementations. The major limitation in the above named methods is that images become totally blurred and a lot of the edge information are lost. In this paper, we analyze super resolution based on CNN and the wavelet transform super resolution method. We compare and analyze the performance according to the number of layers and the training data of the CNN.

• 디지털정책학회지
• /
• 제1권2호
• /
• pp.53-59
• /
• 2022

본 논문에서는 단일 영상을 기반으로 초해상도를 생성하는 과정에서 계층 구조를 거쳐 추출된 보조 경계 특징을 이용한 방법을 제안하였다. 초해상도의 품질을 향상시키기 위해서는 영상 내 경계 영역을 선명하게 표현하면서도 각 영역의 형태를 명확하게 구분하여야 한다. 제안하는 방법은 초해상도 과정에서 품질을 결정하는 중요한 요인인 경계 영역을 입력 영상의 구조적 형태를 유지하면서 개선된 초해상도 결과를 생성하기 위해 딥러닝 기반의 초해상도 방법에서 영상의 경계 영역 정보를 보조적으로 활용하는 구조를 사용하였다. 딥러닝 기반의 초해상도를 수행하기 위한 그룹 컨볼루션 구조에 더해 보조 경계 추출을 위한 고주파 대역의 정보를 기반으로 별도의 계층적 구조의 경계 누적 추출 과정을 수행하여 이를 보조 특징으로써 활용하는 방법을 제안하였다. 실험 결과 기존 초해상도 대비 PSNR과 SSIM에서 약 1%의 성능 향상을 보였다.

• 한국융합학회논문지
• /
• 제9권6호
• /
• pp.45-52
• /
• 2018

본 논문은 단일 이미지에서 초해상도 영상 생성을 위해 저해상도 이미지에서 생성한 패치정보를 기반으로 선형보간하는 방법을 제안하였다. 기존의 초해상도 생성 방법인 전역 공간의 회귀 모델을 사용하면 특정 영역에 대해 참조할 정보가 부족하여 일반적으로 품질이 떨어지는 결과가 나타난다. 이러한 결과를 보완하기 위해 제안하는 방법은 초해상도 이미지 생성 과정에서 영상을 패치 단위로 지역을 분할하여 의미있는 정보를 수집하고, 수집된 정보를 기반으로 초해상도 이미지 생성을 위해 확장시킨 이미지 매트릭스 영역의 구성정보를 분석하여 선형 보간 과정을 거치고 패치정보를 대응시켜 탐색한 최적의 패치 정보를 기준으로 선형 보간하는 방법을 제안하였다. 실험을 위해 원본 이미지를 복원된 영상과 PSNR, SSIM으로 비교 평가하였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제15권4호
• /
• pp.1246-1262
• /
• 2021

Recent studies have demonstrated the strong ability of deep convolutional neural networks (CNNs) to significantly boost the performance in single image super-resolution (SISR). The key concern is how to efficiently recover and utilize diverse information frequencies across multiple network layers, which is crucial to satisfying super-resolution image reconstructions. Hence, previous work made great efforts to potently incorporate hierarchical frequencies through various sophisticated architectures. Nevertheless, economical SISR also requires a capable structure design to balance between restoration accuracy and computational complexity, which is still a challenge for existing techniques. In this paper, we tackle this problem by proposing a competent architecture called Enhanced U-Net Network (EUN), which can yield ready-to-use features in miscellaneous frequencies and combine them comprehensively. In particular, the proposed building block for EUN is enhanced from U-Net, which can extract abundant information via multiple skip concatenations. The network configuration allows the pipeline to propagate information from lower layers to higher ones. Meanwhile, the block itself is committed to growing quite deep in layers, which empowers different types of information to spring from a single block. Furthermore, due to its strong advantage in distilling effective information, promising results are guaranteed with comparatively fewer filters. Comprehensive experiments manifest our model can achieve favorable performance over that of state-of-the-art methods, especially in terms of computational efficiency.