• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.57 no.2
• /
• pp.294-302
• /
• 2008

In this paper, we present a new image up-sampling method which registers low resolution images to the high resolution grid when Bayesian super-resolution image processing is performed. The proposed up-sampling method interpolates high-resolution pixels using high-frequency data lying in all the low resolution images, instead of up-sampling each low resolution image separately. The interpolation is based on B-spline non-uniform re-sampling, adjusted for the super-resolution image processing. The experimental results demonstrate the effects when different up-sampling methods generally used such as zero-padding or bilinear interpolation are applied to the super-resolution image reconstruction. Then, we show that the proposed hybird up-sampling method generates high-resolution images more accurately than conventional methods with quantitative and qualitative assess measures.

• Journal of Electrical Engineering and Technology
• /
• v.10 no.2
• /
• 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.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.33B no.8
• /
• pp.81-95
• /
• 1996

In this paper, we propose a motio estiamtion algorithm using low-resolution quantization to reduce the computation of the full search algorithm. The proposed algorithm consists of the low-resolution search which determins the candidate motion vectors by comparing the low-resolution image and the full-resolution search which determines the motion vector by comparing the full-resolution image on the positions of the candidate motion vectors. The low-resolution image is generated by subtracting each pixel value in the reference block or the search window by the mean of the reference block, and by quantizing it is 2-bit resolution. The candidate motion vectors are determined by counting the number of pixels in the reference block whose quantized codes are unmatched to those in the search window. Simulation results show that the required computational cost of the proposed algorithm is reduced to 1/12 of the full search algorithm while its performance degradation is 0.03~0.12 dB.

• Journal of the Korea Computer Graphics Society
• /
• v.25 no.3
• /
• pp.21-29
• /
• 2019

This paper proposes a novel framework for image fusion of satellite imagery to enhance spatial resolution of the image via structure-texture decomposition. The resolution of the satellite imagery depends on the sensors, for example, panchromatic images have high spatial resolution but only a single gray band whereas multi-spectral images have low spatial resolution but multiple bands. To enhance the spatial resolution of low-resolution images, such as multi-spectral or infrared images, the proposed framework combines the structures from the low-resolution image and the textures from the high-resolution image. To improve the spatial quality of structural edges, the structure image from the low-resolution image is guided filtered with the structure image from the high-resolution image as the guidance image. The combination step is performed by pixel-wise addition of the filtered structure image and the texture image. Quantitative and qualitative evaluation demonstrate the proposed method preserves spectral and spatial fidelity of input images.

• Proceedings of the Korean Information Science Society Conference
• /
• 2004.04b
• /
• pp.715-717
• /
• 2004

This paper proposes a new reconstruction method of high-resolution facial image from a low-resolution facial image based on a recursive error back-projection of top-down machine learning. A face is represented by a linear combination of prototypes of shape and texture. With the shape and texture information about the pixels in a given low-resolution facial image, we can estimate optimal coefficients for a linear combination of prototypes of shape and those of texture by solving least square minimization. Then high-resolution facial image can be obtained by using the optimal coefficients for linear combination of the high-resolution prototypes, In addition to, a recursive error back-projection is applied to improve the accuracy of synthesized high-resolution facial image. The encouraging results of the proposed method show that our method can be used to improve the performance of the face recognition by applying our method to reconstruct high-resolution facial images from low-resolution one captured at a distance.

• The Bulletin of The Korean Astronomical Society
• /
• v.44 no.2
• /
• pp.70.4-70.4
• /
• 2019

Image super-resolution (SR) is a technique that enhances the resolution of a low resolution image. In this study, we use three SR models (RCAN, ProSRGAN and Bicubic) for enhancing solar SDO/HMI magnetograms using deep learning. Each model generates a high resolution HMI image from a low resolution HMI image (4 by 4 binning). The pixel resolution of HMI is about 0.504 arcsec. Deep learning networks try to find the hidden equation between low resolution image and high resolution image from given input and the corresponding output image. In this study, we trained three models with HMI images in 2014 and test them with HMI images in 2015. We find that the RCAN model achieves higher quality results than the other two methods in view of both visual aspects and metrics: 31.40 peak signal-to-noise ratio(PSNR), Correlation Coefficient (0.96), Root mean square error (RMSE) is 0.004. This result is also much better than the conventional bi-cubic interpolation. We apply this model to a full-resolution SDO/HMI image and compare the generated image with the corresponding Hinode NFI magnetogram. As a result, we get a very high correlation (0.92) between the generated SR magnetogram and the Hinode one.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.43 no.2 s.308
• /
• pp.12-18
• /
• 2006

This paper presents a method to extract a good high-resolution image from a low-resolution video in an automatic manner. Since a high-resolution image reconstruction method utilizing several low-resolution input images works better than a conventional interpolation method utilizing single low-resolution input image only if the input images are well registered onto a common high-resolution grid, low-resolution input images should be carefully chosen so that the registration errors can be carefully considered. In this paper, the statistics obtained from the motion-compensated low-resolution images are utilized to evaluate the feasibility of the input image candidates. Maximum motion-compensation error is estimated from the high-resolution image observation model. U the motion-compensation error of the input image candidate is greater than the estimated maximum motion-compensation error, the input image candidate is discarded. The number of good input image candidates and the statistics of the motion-compensation errors are used to choose final input images. The final input images chosen from the input image selection block are given to the following high-resolution image reconstruction block. It is expected that the proposed method is utilized to extract a good high-resolution image efficiently from a low-resolution video without any user intervention.

• The Bulletin of The Korean Astronomical Society
• /
• v.44 no.1
• /
• pp.52.1-52.1
• /
• 2019

The Helioseismic and Magnetic Imager (HMI) is the instrument of Solar Dynamics Observatory (SDO) to study the magnetic field and oscillation at the solar surface. The HMI image is not enough to analyze very small magnetic features on solar surface since it has a spatial resolution of one arcsec. Super resolution is a technique that enhances the resolution of a low resolution image. In this study, we use a method for enhancing the solar image resolution using a Deep-learning model which generates a high resolution HMI image from a low resolution HMI image (4 by 4 binning). Deep learning networks try to find the hidden equation between low resolution image and high resolution image from given input and the corresponding output image. In this study, we trained a model based on a very deep residual channel attention networks (RCAN) with HMI images in 2014 and test it with HMI images in 2015. We find that the model achieves high quality results in view of both visual and measures: 31.40 peak signal-to-noise ratio(PSNR), Correlation Coefficient (0.96), Root mean square error (RMSE) is 0.004. This result is much better than the conventional bi-cubic interpolation. We will apply this model to full-resolution SDO/HMI and GST magnetograms.

• Proceedings of the KSRS Conference
• /
• 2005.10a
• /
• pp.197-200
• /
• 2005

Super resolution image reconstruction method refers to image processing algorithms that produce a high resolution(HR) image from observed several low resolution(LR) images of the same scene. This method is proved to be useful in many practical cases where multiple frames of the same scene can be obtained, such as satellite imaging, video surveillance, video enhancement and restoration, digital mosaicking, and medical imaging. In this paper we applied super resolution reconstruction method in spatial domain to video sequences. Test images are adjacently sampled images from continuous video sequences and overlapped for high rate. We constructed the observation model between the HR images and LR images applied by the Maximum A Posteriori(MAP) reconstruction method that is one of the major methods in the super resolution grid construction. Based on this method, we reconstructed high resolution images from low resolution images and compared the results with those from other known interpolation methods.

• Proceedings of the KSRS Conference
• /
• 2004.10a
• /
• pp.115-118
• /
• 2004

Images with high resolution are desired and often required in many visual applications. When resolution can not be improved by replacing sensors, either because of cost or hardware physical limits, super resolution image reconstruction method is what can be resorted to. Super resolution image reconstruction method refers to image processing algorithms that produce high quality and high resolution images from a set of low quality and low resolution images. The method is proved to be useful in many practical cases where multiple frames of the same scene can be obtained, including satellite imaging, video surveillance, video enhancement and restoration, digital mosaicking, and medical imaging. The method can be either the frequency domain approach or the spatial domain approach. Much of the earlier works concentrated on the frequency domain formulation, but as more general degradation models were considered, later researches had been almost exclusively on spatial domain formulations. The method in spatial domains has three stages: i) motion estimate or image registration, ii) interpolation onto high resolution grid and iii) deblurring process. The super resolution grid construction in the second stage was discussed in this paper. We applied the Maximum A­Posteriori(MAP) reconstruction method that is one of the major methods in the super resolution grid construction. Based on this method, we reconstructed high resolution images from a set of low resolution images and compared the results with those from other known interpolation methods.