• Journal of the Korea Society of Computer and Information
• /
• v.27 no.5
• /
• pp.85-92
• /
• 2022

In this paper, we propose a technique that can express low-resolution hair and fur simulations in high-resolution without noise using ConvNet and geometric images of strands in the form of lines. Pairs between low-resolution and high-resolution data can be obtained through physics-based simulation, and a low-resolution-high-resolution data pair is established using the obtained data. The data used for training is used by converting the position of the hair strands into a geometric image. The hair and fur network proposed in this paper is used for an image synthesizer that upscales a low-resolution image to a high-resolution image. If the high-resolution geometry image obtained as a result of the test is converted back to high-resolution hair, it is possible to express the elastic movement of hair, which is difficult to express with a single mapping function. As for the performance of the synthesis result, it showed faster performance than the traditional physics-based simulation, and it can be easily executed without knowing complex numerical analysis.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.11
• /
• pp.1477-1485
• /
• 2021

Recently, super-resolution has been intensively studied only on upscaling models with integer magnification. However, the need to expand arbitrary magnification is emerging in representative application fields of actual super-resolution, such as object recognition and display image quality improvement. In this paper, we propose a model that can support arbitrary magnification by using the weights of the existing integer magnification model. This model converts super-resolution results into the DCT spectral domain to expand the space for arbitrary magnification. To reduce the loss of high-frequency information in the image caused by the expansion by the DCT spectral domain, we propose a high-frequency attention network for arbitrary magnification so that this model can properly restore high-frequency spectral information. To recover high-frequency information properly, the proposed network utilizes channel attention layers. This layer can learn correlations between RGB channels, and it can deepen the model through residual structures.

• Proceedings of the Korea Water Resources Association Conference
• /
• 1998.05b
• /
• pp.3-3
• /
• 1998

In this pressentation it is argued that the heterogeneity of a hydrologic attribute which may seem to be nonstationary at one scale, may become stationary at a larger scale. The fundamental reason for transformation from nonstationarity to stationarity whith the increase in scale is the phenomenon of coarse-graining of the hydrologic processes with increasing scale. Due to the phenomenon of aliasing, a particular scale hydrologic process heterogeneity which is observed as a nonstationary process at that scale, may be observed as a stationary process at a higher(larger) scale whose size is bigger than the stationary extent of the lower scale heterogeneity. As one goes through a hierarchical sequence of larger and larger scales for observations, one would eliminate nonstationarities which emerge at some lower scales at the expense of losing information on the high frequency fluctuations of the lower scale heterogeneities which will no longer be observed at the larger sampling scales. We call this phenimenon as the "coarse-graining in hydrologic observations". In this presentation, it is also argued that by the coarse-graining of hydrologic processes due to the averaging and aliasing operations at increasing scales, the conservation laws corresponging to these scales may still be quite parsimonious, and need not be more complicated as the scales get larger. It is shown that shen a higher(larger) scale process is formed by averaging a lower(smaller) scale process in time or space, the high frequency components of the lower scale process will be eliminated by the averaging operation. Thereby, the resuliiting average hydrologic dynamics, free from the effects of the high frequency components of the lower scale process, can still be quite simple in form. This is demonstrated by means of some recent upscaling work on the solute teansport conservation equation for hetergeneous aquifers. By means of this solute transport example, it is also shown that for the ensemble average form of a hydrologic conservation equation to be equivalent to its volume-average form at any scale, the parameter functions of that conservation equation at the immediately lower scale must be ergodic.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.17 no.4
• /
• pp.69-85
• /
• 2014

Carbon stocks of NFI plots can be accurately estimated using field survey information. However, an accurate estimation of carbon stocks in other unsurveyed sites is very difficult. In order to fill this gap, various spatial information can be used as an ancillary data. In South Korea, there is the 1:5,000 forest type map that was produced by digital air-photo interpretation and field survey. Because this map contains very detailed forest information, it can be used as the high-quality spatial data for estimating carbon stocks. In this study, we compared three upscaling methods based on the 1:5,000 forest type map and 5th national forest inventory data. Map algebra(method 1), RK(Regression Kriging)(method 2), and GWR(Geographically Weighted Regression)(method 3) were applied to estimate forest carbon stock in Chungcheong-nam Do and Daejeon metropolitan city. The range of carbon stocks from method 2(1.39~138.80 tonC/ha) and method 3(1.28~149.98 tonC/ha) were more similar to that of previous method(1.56~156.40 tonC/ha) than that of method 1(0.00~93.37 tonC/ha). This result shows that RK and GWR considering spatial autocorrelation can show spatial heterogeneity of carbon stocks. We carried out paired t-test for carbon stock data using 186 sample points to assess estimation accuracy. As a result, the average carbon stocks of method 2 and field survey method were not significantly different at p=0.05 using paired t-test. And the result of method 2 showed the lowest RMSE. Therefore regression kriging method is useful to consider spatial variations of carbon stocks distribution in rugged terrain and complex forest stand.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.53 no.11
• /
• pp.73-81
• /
• 2016

In general, infrared images have less sharpness and image details than visible images. So, the prior image upscaling methods are not effective in the infrared images. In order to solve this problem, this paper proposes an algorithm which initially up-scales an input infrared (IR) image by using adaptive dynamic range encoding (ADRC)-based super-resolution (SR) method, and then fuses the result with the corresponding visible images. The proposed algorithm consists of a up-scaling phase and a fusion phase. First, an input IR image is up-scaled by the proposed ADRC-based SR algorithm. In the dictionary learning stage of this up-scaling phase, so-called 'pre-emphasis' processing is applied to training-purpose high-resolution images, hence better sharpness is achieved. In the following fusion phase, high-frequency information is extracted from the visible image corresponding to the IR image, and it is adaptively weighted according to the complexity of the IR image. Finally, a up-scaled IR image is obtained by adding the processed high-frequency information to the up-scaled IR image. The experimental results show than the proposed algorithm provides better results than the state-of-the-art SR, i.e., anchored neighborhood regression (A+) algorithm. For example, in terms of just noticeable blur (JNB), the proposed algorithm shows higher value by 0.2184 than the A+. Also, the proposed algorithm outperforms the previous works even in terms of subjective visual quality.

• Journal of Broadcast Engineering
• /
• v.24 no.1
• /
• pp.132-141
• /
• 2019

The necessity of transmitting video data over a narrow-bandwidth exists steadily despite that video service over broadband is common. In this paper, we propose a scalable video coding framework for low-resolution video transmission over a very narrow-bandwidth network by super-resolution of decoded frames of a base layer using a convolutional neural network based super resolution technique to improve the coding efficiency by using it as a prediction for the enhancement layer. In contrast to the conventional scalable high efficiency video coding (SHVC) standard, in which upscaling is performed with a fixed filter, we propose a scalable video coding framework that replaces the existing fixed up-scaling filter by using the trained convolutional neural network for super-resolution. For this, we proposed a neural network structure with skip connection and residual learning technique and trained it according to the application scenario of the video coding framework. For the application scenario where a video whose resolution is $352{\times}288$ and frame rate is 8fps is encoded at 110kbps, the quality of the proposed scalable video coding framework is higher than that of the SHVC framework.