• 한국멀티미디어학회논문지
• /
• 제23권3호
• /
• pp.412-420
• /
• 2020

This paper proposes a performance-improving face recognition system based on a super resolution method for low-resolution images. The conventional face recognition algorithm has a rapidly decreased accuracy rate due to small image resolution by a distance. To solve the previously mentioned problem, this paper generates a super resolution images based o deep learning method. The proposed method improved feature information from low-resolution images using a super resolution method and also applied face recognition using a feature extraction and an classifier. In experiments, the proposed method improves the face recognition rate when compared to conventional methods.

• 한국측량학회지
• /
• 제30권6_2호
• /
• pp.653-663
• /
• 2012

Using soft classification, it is possible to obtain the land cover proportions from the remotely sensed image. These land cover proportions are then used as input data for a procedure called "super-resolution mapping" to produce the predicted hard land cover layers at higher resolution than the original remotely sensed image. Superresolution mapping can be implemented using a number of algorithms in which the Hopfield Neural Network (HNN) has showed some advantages. The HNN has improved the land cover classification through superresolution mapping greatly with the high resolution data. However, the super-resolution mapping is based on the spatial dependence assumption, therefore it is predicted that the accuracy of resulted land cover classes depends on the relative size of spatial features and the spatial resolution of the remotely sensed image. This research is to evaluate the capability of HNN to implement the super-resolution mapping for SPOT image to create higher resolution land cover classes with different zoom factor.

• 한국컴퓨터그래픽스학회논문지
• /
• 제25권3호
• /
• pp.105-113
• /
• 2019

물리 기반 유체 시뮬레이션은 고해상도 연산을 위해 많은 시간이 필요하다. 이 문제를 해결하기 위해 저해상도 유체 시뮬레이션의 한계를 딥 러닝으로 보완하는 연구들이 있으며, 그중에서는 저해상도의 시뮬레이션 데이터를 고해상도로 변환해주는 Super-resolution 분야가 있다. 하지만 기존 기법들은 전체 데이터 공간에서 밀도 데이터가 없는 부분까지 연산하므로 전체 시뮬레이션 속도 면에서 효율성이 떨어지며, 입력 해상도가 큰 경우에는 GPU 메모리가 부족해 연산할 수 없는 경우가 발생할 수 있다. 본 연구에서는 공간 분할 법 중 하나인 쿼드 트리를 활용하여 시뮬레이션 공간을 분할 및 분류하여 Super-resolution 하는 기법을 제안한다. 본 기법은 필요 공간만 Super-resolution 하므로 전체 시뮬레이션 가속화가 가능하고, 입력 데이터를 분할 연산하므로 GPU 메모리 문제를 해결할 수 있게 된다.

• 반도체디스플레이기술학회지
• /
• 제19권4호
• /
• pp.71-76
• /
• 2020

In this paper, we propose a super-resolution method that reconstructs compressed low-resolution images into high-resolution images. We propose a CNN model with a small number of parameters, and even if quantization is applied to the proposed model, super-resolution can be implemented without deteriorating the image quality. To further improve the quality of the compressed low-resolution image, a new degradation model was proposed instead of the existing bicubic degradation model. The proposed degradation model is used only in the training process and can be applied by changing only the parameter values to the original CNN model. In the super-resolution image applying the proposed degradation model, visual artifacts caused by image compression were effectively removed. As a result, our proposed method generates higher PSNR values at compressed images and shows better visual quality, compared to conventional CNN-based SR methods.

• 정보저장시스템학회논문집
• /
• 제4권1호
• /
• pp.1-5
• /
• 2008

Super-resolution near-field structure (super-RENS) technology and solid immersion lens (SIL) based near-field (NF) technology have been expected as promising approaches to increase data capacity or areal density of optical disc. Super-RENS technology has been studied until now using mainly numerical aperture (NA) of 0.85 far-field optical system and possibility of tangential data density increment have been presented. NF technology has been studied with NA over 1 and presented demonstration of removable performance. To achieve much higher density, approach to increase NA of super-RENS by NF technology (Near-Field Super-Resolution, NFSR) can be a candidate and we think this technology would be advantageous compared to wavelength reduction or much higher NA increment of NF technology or much smaller effective optical spot size reduction of far-field super-resolution technology. In this paper we present readout result of ROM media having monotone pits using NF optical system with wavelength of 405nm and NA of 1.84 surface type SIL. GeSbTe material was used for super resolution active layer and pit length is 37.5nm which is shorter than resolution limit 55nm. We present the feasibility of NFSR technology by confirming the CNR threshold according to readout power (Pr) and CNR 33dB over threshold Pr.

• Nuclear Engineering and Technology
• /
• 제54권6호
• /
• pp.2204-2212
• /
• 2022

Recently, the demand for alpha imaging detectors for quantifying the distributions of alpha particles has increased in various fields. This study aims to reconstruct a high-resolution image from an alpha imaging detector by applying a super-spatial resolution method combined with the maximum-likelihood expectation maximization (MLEM) algorithm. To perform the super-spatial resolution method, several images are acquired while slightly moving the detector to predefined positions. Then, a forward model for imaging is established by the system matrix containing the mechanical shifts, subsampling, and measured point-spread function of the imaging system. Using the measured images and system matrix, the MLEM algorithm is implemented, which converges towards a high-resolution image. We evaluated the performance of the proposed method through the Monte Carlo simulations and phantom experiments. The results showed that the super-spatial resolution method was successfully applied to the alpha imaging detector. The spatial resolution of the resultant image was improved by approximately 12% using four images. Overall, the study's outcomes demonstrate the feasibility of the super-spatial resolution method for the alpha imaging detector. Possible applications of the proposed method include high-resolution imaging for alpha particles of in vitro sliced tissue and pre-clinical biologic assessments for targeted alpha therapy.

• Journal of Electrical Engineering and Technology
• /
• 제10권2호
• /
• pp.659-669
• /
• 2015

Face Recognition assumes much significance in the context of security based application. Normally, high resolution images offer more details about the image and recognizing a face from a reasonably high resolution image would be easier when compared to recognizing images from very low resolution images. This paper addresses the problem of recognizing faces from a very low resolution image whose size is as low as $8{\times}8$. With the use of CCTV(Closed Circuit Television) and with other surveillance camera-based application for security purposes, the need to overcome the shortcomings with very low resolution images has been on the rise. The present day face recognition algorithms could not provide adequate performance when employed to recognize images from VLR images. Existing methods use super-resolution (SR) methods and Relation Based Super Resolution methods to construct from very low resolution images. This paper uses a learning based super resolution method to extract and construct images from very low resolution images. Experimental results show that the proposed SR algorithm based on relationship learning outperforms the existing algorithms in public face databases.

• 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.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제13권8호
• /
• pp.3942-3961
• /
• 2019

Recently, the models of deep super-resolution networks can successfully learn the non-linear mapping from the low-resolution inputs to high-resolution outputs. However, for large scaling factors, this approach has difficulties in learning the relation of low-resolution to high-resolution images, which lead to the poor restoration. In this paper, we propose Stage Generative Adversarial Networks (Stage-GAN) with semantic maps for image super-resolution (SR) in large scaling factors. We decompose the task of image super-resolution into a novel semantic map based reconstruction and refinement process. In the initial stage, the semantic maps based on the given low-resolution images can be generated by Stage-0 GAN. In the next stage, the generated semantic maps from Stage-0 and corresponding low-resolution images can be used to yield high-resolution images by Stage-1 GAN. In order to remove the reconstruction artifacts and blurs for high-resolution images, Stage-2 GAN based post-processing module is proposed in the last stage, which can reconstruct high-resolution images with photo-realistic details. Extensive experiments and comparisons with other SR methods demonstrate that our proposed method can restore photo-realistic images with visual improvements. For scale factor ${\times}8$, our method performs favorably against other methods in terms of gradients similarity.

• 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.