• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2012.07a
• /
• pp.16-18
• /
• 2012

Haze is an extreme reason of the reduction of contrast when capturing image in the outdoor. Recently, there are several single image dehazing techniques, but they are not robust in dynamic variations of natural environment caused by the thickness, coverage of haze and appearance of sunlight. In this paper, we propose an effective and robust method to enhance luminance for image dehazing depending on histogram analysis. Compare with conventional methods, our proposal have better performance in term of contrast, and computation time.

• Journal of ILASS-Korea
• /
• v.1 no.3
• /
• pp.51-59
• /
• 1996

An experiment was carried out on the structure of twin spray from pressurize-swirl nozzles, in order to investigate the effect of different size of spray nozzles on the characteristics of the overlap of two single sprays, for example, mean diameter, number density, and spatial distribution of flow rate. Using image processing method, the distributions of size and velocity of droplets of a single spray and twin spray were measured and compared to investigate the overlapping effect of two identical sprays. Comparing experimental results from a twin-spray with those from two-single sprays shows that the flow rate distribution of the twin-spray was concentrated around the midst of the overlapping region of two sprays. In this region, Sauter mean diameter (SMD) did not change much in the twin spray from 6032 nozzles, but it was smaller by 10 micrometers in the twin-spray than two-single sprays from 60063 nozzles. In spite of large difference in Weber numbers of the colliding sprays between the 60063 and 6032 nozzles, the phenomena did not have a big change in the overlapping region of twin spray. This shows that in the collision between droplets from two single spray in the overlapping region to cause the disruption of droplets, the size distribution of spray droplets was also important as well as Weber number.

• Korean Journal of Optics and Photonics
• /
• v.13 no.3
• /
• pp.258-265
• /
• 2002

In this paper, we design single and symmetrical double lenses with DOE. The specifications are the following : Image area is 4.8 mm $\times$ 3.6 mm, F/# is 2.8 and the overall length (from first lens surface to image plane) is 6.8mm. After comparing the optical performance and characteristic values, we determine that symmetrical double lenses are superior to single lenses. Symmetrical double lenses have the merits of fewer zones, weaker flare, and smaller distortion than single lenses.

• The Research Journal of the Costume Culture
• /
• v.20 no.1
• /
• pp.62-71
• /
• 2012

The purpose of this research is to analyze the characteristics of the color combination in Eco fashion design. 532 pictures related to eco was collected through Eco brand web sites from 2008 through 2011. The characteristics for analysis include color combination and tone. The method of research is content and statistical analysis. The result of the research is summed up below. First, achromatic color was most frequently presented in Eco fashion. Also monotone color or single color combination were displayed often. Secondly, many dresses were of single color combination and T-shirt items displayed contrast color combination created by luminosity difference. Third, the outcome of comparing seasons was that both seasons presented single color and contrast color combination. The difference between countries proved slight as single color coloration dominated. This considered the process which was non-chemical manufacture for eco-friendly. For the combination of colors, the same coloration with single colors or contrast coloration due to the differences of brightness were presented quite often, generally, the modern image fashion without fancy factors have shown a lot. These results that minimalism trend was included in eco-friendly image.

• Journal of Multimedia Information System
• /
• v.8 no.4
• /
• pp.203-210
• /
• 2021

Deep leaning convolutional neural networks (CNN) have successfully been applied to image super-resolution (SR). Despite their great performances, SR techniques tend to focus on a certain upscale factor when training a particular model. Algorithms for single model multi-scale networks can easily be constructed if images are upscaled prior to input, but sub-pixel convolution upsampling works differently for each scale factor. Recent SR methods employ multi-scale and multi-path learning as a solution. However, this causes unshared parameters and unbalanced parameter distribution across various scale factors. We present a multi-scale single-path upsample module as a solution by exploiting the advantages of sub-pixel convolution and interpolation algorithms. The proposed model employs sub-pixel convolution for the highest scale factor among the learning upscale factors, and then utilize 1-dimension interpolation, compressing the learned features on the channel axis to match the desired output image size. Experiments are performed for the single-path upsample module, and compared to the multi-path upsample module. Based on the experimental results, the proposed algorithm reduces the upsample module's parameters by 24% and presents slightly to better performance compared to the previous algorithm.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.24 no.1
• /
• pp.57-64
• /
• 2006

Recently the extraction of building height information has been investigated using remotely sensed image and digital maps. In this study, based on the digital photogrammetry principle and mono imagery method the building height information can be extracted by using relationship between ground coordinates and image coordinates. To evaluate the result the comparison was done with building height from 1:5000 aerial photo. The experiment shows that extraction of building height could be performed using IKONOS single imagery and digital map and it is proved that the building height could be reconstructed within some extent.

• Journal of Broadcast Engineering
• /
• v.23 no.3
• /
• pp.369-382
• /
• 2018

We propose a single-shot high dynamic range (HDR) imaging algorithm using a deep convolutional neural network (CNN) for row-wise varying exposures in a single image. The proposed algorithm restores missing information resulting from under- and/or over-exposed pixels in an input image and reconstructs the raw radiance map. The main contribution of this work is the development of a loss function for the CNN employing the human visual system (HVS) properties. Then, the HDR image is obtained by applying a demosaicing algorithm. Experimental results demonstrate that the proposed algorithm provides higher-quality HDR images than conventional algorithms.

• Journal of the Korean Geotechnical Society
• /
• v.16 no.1
• /
• pp.221-226
• /
• 2000

In this study, the characteristics of gas-water relative permeability curves in a single fractured-plate according to the various aperture size were analyzed by using the Hele-Shaw type glass plate model. The plate was made of glasses for the observation of the two-phase flow pattern, and seven cases were set up based on the aperture size in the range of field scale from 30 to $120\mum$. The experiment was conducted by steady-state method, and the water saturation was determined more accurately by the developed digital image process technique. The empirical equations of relative permeability to gas and water for single fractured-plate were correlated by using the aperture size which directly affects the two-phase flow pattern and critical saturation.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.5
• /
• pp.2539-2554
• /
• 2017

Regression-based image super resolution (SR) methods have shown great advantage in time consumption while maintaining similar or improved quality performance compared to other learning-based methods. In this paper, we propose a novel single image SR method based on hierarchical regression to further improve the quality performance. As an improvement to other regression-based methods, we introduce a hierarchical scheme into the process of learning multiple regressors. First, training samples are grouped into different clusters according to their geometry similarity, which generates the structure layer. Then in each cluster, a compact dictionary can be learned by Sparse Coding (SC) method and the training samples can be further grouped by dictionary atoms to form the detail layer. Last, a series of projection matrixes, which anchored to dictionary atoms, can be learned by linear regression. Experiment results show that hierarchical scheme can lead to regression that is more precise. Our method achieves superior high quality results compared with several state-of-the-art methods.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.66 no.7
• /
• pp.1117-1122
• /
• 2017

Pedestrian detection is a very important and valuable part of artificial intelligence and computer vision. It can be used in various areas for example automatic drive, video analysis and others. Many works have been done for the pedestrian detection. The accuracy of pedestrian detection on multiple pedestrian image has reached high level. It is not easily get more progress now. This paper proposes a new structure based on the idea of HOG and convolutional filters to do the pedestrian detection in single pedestrian image. It can be a method to increase the accuracy depend on the high accuracy in single pedestrian detection. In this paper, we use Multiblock HOG and magnitude of the pixel as the feature and use convolutional filter to do the to extract the feature. And then use NEWFM to be the classifier for training and testing. We use single pedestrian image of the INRIA data set as the data set. The result shows that the Convolutional Multiblock HOG we proposed get better performance which is 0.015 miss rate at 10-4 false positive than the other detection methods for example HOGLBP which is 0.03 miss rate and ChnFtrs which is 0.075 miss rate.