• Journal of Korea Multimedia Society
• /
• v.22 no.11
• /
• pp.1242-1250
• /
• 2019

Super resolution technique aims to convert a low-resolution image with coarse details to a corresponding high-resolution image with refined details. In the past decades, the performance is greatly improved due to progress of deep learning models. However, universal solution for various objects is a still challenging issue. We observe that learning super resolution with a general dataset has poor performance on faces. In this paper, we propose a super resolution fusion scheme that works well for both general- and face datasets to achieve more universal solution. In addition, object-specific feature extractor is employed for better reconstruction performance. In our experiments, we compare our fusion image and super-resolved images from one- of the state-of-the-art deep learning models trained with DIV2K and FFHQ datasets. Quantitative and qualitative evaluates show that our fusion scheme successfully works well for both datasets. We expect our fusion scheme to be effective on other objects with poor performance and this will lead to universal solutions.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.38A no.2
• /
• pp.174-182
• /
• 2013

Most methods for fusion-based finger vein recognition were to fuse different features or matching scores from more than one trait to improve performance. To overcome the shortcomings of "the curse of dimensionality" and additional running time in feature extraction, in this paper, we propose a finger vein recognition technology based on matching score-level fusion of a single trait. To enhance the quality of finger vein image, the contrast-limited adaptive histogram equalization (CLAHE) method is utilized and it improves the local contrast of normalized image after ROI detection. Gabor features are then extracted from eight channels based on a bank of Gabor filters. Instead of using the features for the recognition directly, we analyze the contributions of Gabor feature from each channel and apply a weighted matching score-level fusion rule to get the final matching score, which will be used for the last recognition. Experimental results demonstrate the CLAHE method is effective to enhance the finger vein image quality and the proposed matching score-level fusion shows better recognition performance.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.12 no.4
• /
• pp.125-135
• /
• 2009

Owing to its fast computing capability for fusing images, the FIHS(Fast Intensity Hue Saturation) fusion is widely used for fusion purposes. However, the FIHS fusion also distorts color in the same way such as the IHS(Intensity Hue Saturation) fusion technique. In this paper, a FIHS fusion technique(FIHS-BR) which reduces color distortion by using the ratio of each spectral band and an adaptive FIHS fusion(FIHS-SABR) using spatial information and the ratio of each spectral band are proposed. The proposed FIHS-BR fusion reduces color distortion by adding different spatial detail improvement values for each spectral band. The spatial detail improvement values are derived from the ratio of spectral band. And the proposed FIHS-SABR fusion reduces more color distortion by readjusting the spatial detail improvement values for each spectral band according to the ratio of the spectral bands. The spatial detail improvement values are derived adaptively from the characteristics of spatial information of the local image. To evaluate the performance of the proposed FIHS-BR fusion and FIHS-SABR fusion, a computer simulation is performed for IKONOS remote sensing image. Results from the experiments show that the proposed methods have less color distortion for the forest regions which reveal severe color distortion in the traditional FIHS fusion. From the evaluation results of the characteristics of spectral information for fused image, we show that the proposed methods have best results.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.22 no.4
• /
• pp.359-365
• /
• 2004

Data integration techniques are becoming increasing1y important for conquering a limitation with a single data. Image fusion which improves the spatial and spectral resolution from a set of images with difffrent spatial and spectral resolutions, and image registration which matches two images so that corresponding coordinate points in the two images correspond to the same physical region of the scene being imaged have been researched. In this paper, we compared with six image fusion methods(Brovey, IHS, PCA, HPF, CN, and MWD) with panchromatic and multispectral images of IKONOS and developed the registration method for applying to SPOT-5 satellite image and RADARSAT SAR satellite image. As the result of tests on image fusion and image registration, we could find that MWD and HPF methods showed the good result in term of visual comparison analysis and statistical analysis. And we could extract patches which depict detailed topographic information from SPOT-5 and RADARSAT and obtain encouraging results in image registration.

• Journal of Korean Society for Geospatial Information Science
• /
• v.18 no.2
• /
• pp.79-86
• /
• 2010

Generally, image fusion is defined as generating re-organized image by merging two or more data using special algorithms. In remote sensing, image fusion technique is called as Pan-sharpening algorithm because it aims to improve the spatial resolution of original multispectral image by using panchromatic image of high-spatial resolution. The pan-sharpened image has been an important task due to various applications such as change detection, digital map creation and urban analysis. However, most approaches have tended to distort the spectral information of the original multispectral data or decrease the spatial quality compared with the panchromatic image. In order to solve these problems, a novel pan-sharpening algorithm is proposed by considering the spectral and spatial characteristics of multispectral image. The algorithm is applied to the KOMPSAT-2 and QuickBird satellite image and the results showed that our method can improve the spectral/spatial quality compared with the existing fusion algorithms.

• Journal of Korean Society for Geospatial Information Science
• /
• v.16 no.3
• /
• pp.15-20
• /
• 2008

High-spatial resolution remote sensing satellites (IKONOS-2, QuickBird and KOMPSAT-2) have provided low-spatial resolution multispectral images and high-spatial resolution panchromatic images. Image fusion or Pan-sharpening is a very important in that it aims at using a satellite image with various applications such as visualization and feature extraction through combining images that have a different spectral and spatial resolution. Many image fusion algorithms are proposed, most methods could not preserve the spectral information of original multispectral image after image fusion. In order to solve this problem, modified induction technique which reduce the spectral distortion of fused image is developed. The spectral distortion is adjusted by the comparison between the spatially degraded pan-sharpened image and original multispectral image and our algorithm is evaluated by QuickBird satellite imagery. In the experiment, pan-sharpened image by various methods can reduce spectral distortion when our algorithm is applied to the fused images.

• Journal of Korea Multimedia Society
• /
• v.19 no.12
• /
• pp.1951-1959
• /
• 2016

In this paper, we propose a two-scale fusion method for infrared and visible images using saliency and variance. The images are separated into two scales respectively: a base layer of low frequency component and a detailed layer of high frequency component. Then, these are synthesized using weight. The saliencies and the variances of the images are used as the fusion weights for the two-scale images. The proposed method is tested on several image pairs, and its performance is evaluated quantitatively by using objective fusion metrics.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.31 no.6_2
• /
• pp.593-599
• /
• 2013

China's Mapping Satellite-1, developed by the China Aerospace Science and Technology Corporation (CASC), was launched in three years ago. The data from Mapping Satellite-1 are able to use for efficient surveying and geometric mapping application field. In this paper, we fuse the panchromatic and multispectral images of Changchun area, which are obtained from the Mapping Satellite-1, the one that is the Chinese first transmission-type three-dimensional mapping satellite. The four traditional image fusion methods, which are HPF, Mod.IHS, Panshar and wavelet transform, were used to approach for effectively fusing Mapping Satellite-1 remote sensing data. Subsequently we assess the results with some commonly used methods, which are known a subjective qualitative evaluation and quantitative statistical analysis approach. Consequently, we found that the wavelet transform remote sensing image fusion is the optimal in the degree of distortion, the ability of performance of details and image information availability among four methods. To further understand the optimal methods to fuse Mapping Satellite-1 images, an additional study is necessary.

• Korean Journal of Remote Sensing
• /
• v.28 no.6
• /
• pp.611-622
• /
• 2012

This study proposed two multisensor fusion methods for segment-based image classification utilizing a region-growing segmentation. The proposed algorithms employ a Gaussian-PDF measure and an evidential measure respectively. In remote sensing application, segment-based approaches are used to extract more explicit information on spatial structure compared to pixel-based methods. Data from a single sensor may be insufficient to provide accurate description of a ground scene in image classification. Due to the redundant and complementary nature of multisensor data, a combination of information from multiple sensors can make reduce classification error rate. The Gaussian-PDF method defines a regional measure as the PDF average of pixels belonging to the region, and assigns a region into a class associated with the maximum of regional measure. The evidential fusion method uses two measures of plausibility and belief, which are derived from a mass function of the Beta distribution for the basic probability assignment of every hypothesis about region classes. The proposed methods were applied to the SPOT XS and ENVISAT data, which were acquired over Iksan area of of Korean peninsula. The experiment results showed that the segment-based method of evidential measure is greatly effective on improving the classification via multisensor fusion.

• Nuclear Medicine and Molecular Imaging
• /
• v.41 no.3
• /
• pp.252-254
• /
• 2007

Image fusion is fast catching attention as Wagner pointed out in his 2006 version of the recent progress and development presented at the annual meeting of Society of Nuclear Medicine. Prototypical fusion of bone scan and radiograph was already attempted at in 1961 when Fleming et al. published an article on strontium-85 bone scan. They simply superimposed dot scan on radiograph enabling simultaneous assessment of altered bone metabolism and local bone anatomy. Indeed the parallel reading of images of bone scan and radiography, CT, MRI or ultrasonography has been practiced in nuclear medicine long since. It is fortunate that recent development of computer science and technology along with the availability of refined CT and SPECT machines has permitted us to open a new avenue to digitally produce precise fusion image so that they can readily be read, exchanged and disseminated using internet. Ten years ago fusion was performed using Bresstrahlung SPECT/CT and it is now achievable by PET/CT and SPECT/CT software and SPECT/CT hardware. The merit of image fusion is its feasibility of reliable assessment of morphological and metabolic change. It is now applicable not only to stationary organs such as brain and skeleton but also to moving organs such as the heart, lung and stomach. Recently, we could create useful fusion image of cardiac SPECT and 64-channel CT angiograph. The former provided myocardial metabolic profile and the latter vascular narrowing in two patients with coronary artery stenosis and myocardial ischemia. Arterial stenosis was severe in Case 1 and mild in Case 2.