• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2022.06a
• /
• pp.1333-1335
• /
• 2022

최근 딥러닝은 다양한 컴퓨터 비전에 적용되어 높은 성능을 제공하고 있고 이에 따라 중간 프레임을 생성하는 비디오 프레임 보간 기법에도 딥러닝이 적용되고 있다. 많은 딥러닝 기반의 비디오 프레임 보간 기법은 크게 옵티컬 플로우를 추정하는 플로우 추정 네트워크와 합성 네트워크로 구성되며 본 논문에서는 합성 네트워크 부분의 성능향상을 위한 네트워크에 대하여 다룬다. 합성 네트워크에 주로 사용되는 UNet 구조와 GridNet 구조의 장단점과 네트워크에 따른 보간 결과의 차이에 대해서 알아보고 영상 복원에서 제안된 NAFNet 을 비디오 보간 기법에 맞게 변형시켜 합성 네트워크에 적용한 보간 결과의 차이를 보였다. 실험결과는 기존 네트워크 대비 Vimeo90K 데이터셋에 대하여 PSNR 값이 0.63dB 개선됨을 보여준다.

• Journal of Broadcast Engineering
• /
• v.19 no.6
• /
• pp.789-797
• /
• 2014

In general, the image of interlaced side-by-side (SBS) streams is not good because it goes through two reconstruction steps-de-interlacing and horizontal interpolation. Especially the accuracy of horizontal resolution is greatly deteriorated as processed de-interlacing value is used for reference of interpolated pixel at time of horizontal interpolation. To solve this problem, we propose an improved method considering interlaced SBS stream's characteristics. In this paper, we adopted two separate methods to transmitted line and blank line for using reliable information. The experimental result of the proposed method is better than the conventional algorithm in terms of subjective and objective image quality.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.35 no.11A
• /
• pp.1027-1037
• /
• 2010

In this paper, we propose the linear interpolation-based BD (Block Diagonalization) precoding approximation algorithm for low complexity in downlink multiuser MIMO-OFDM (Multiple-input Multiple-output Orthogonal Frequency Division Multiplexing) systems. In the case of applying the general BD precoding algorithm to multiuser MIMO-OFDM systems, the computational complexity increases in proportional to the number of subcarriers. The proposed interpolation-based BD precoding approximation algorithm can be achieved similar SER performance with general BD algorithm and can decrease the computational complexity. It is proved that proposed algorithm can achieve the significantly decreased computational complexity by computer simulation.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.27 no.1A
• /
• pp.51-57
• /
• 2002

In this paper, a simple interpolation technique in a frequency domain is proposed for the discrete Fourier transform(DFT) algorithm. Frequency and phase resolution capabilities of the DFT algorithm can be significantly improved by the proposed interpolation technique without increase of a DFT size(the number of points for the DFT). The new technique uses a diving point in amplitude and phase spectrums. As an application, the technique can be used for joint estimation of fine frequency and phase offsets in burst mode digital transmission. Simulation results show that the joint estimator using the technique is robust to estimation errors.

• Journal of the Korea Society of Computer and Information
• /
• v.17 no.2
• /
• pp.59-67
• /
• 2012

In this paper, an efficient interpolation scheme using gradient information between adjacent pixel values was proposed to estimate the value of interpolated pixel to have the locality which exists in real image and the characteristic of simple convex surface and simple concave surface which exist partially in the real image. PSNR(Peak Signal to Noise Ratio) was used to evaluate the performance of the proposed scheme. The PSNR values of the magnified images using the proposed scheme are greater than those of the magnified images using the previous interpolation schemes.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2015.07a
• /
• pp.435-438
• /
• 2015

본 논문에서는 영상 보간의 결과 측면에서 깊이를 보정하는 방법 중 하나인 유도 영상 필터링 기법을 비교한다. 실험을 위해 초기 깊이 영상을 두 종류의 유도 영상 필터링 기법으로 개선을 하였다. 초기 깊이 영상과, 각각의 필터링 기법에 의해 개선된 변이 영상을 이용하여 영상 보간을 하였고, 그 결과를 비교하였다. 결과로서 한 시점 영상의 텍스처 정보만을 이용하여 변이를 개선하는 유도 영상 필터링 기법으로 변이 영상을 개선하게 되는 경우에 육안으로는 구분이 갈 정도로 변이가 개선이 되지만, 영상 보간의 측면에서 보았을 때는 크게 차이가 없거나 오히려 품질이 저하되는 경우를 확인할 수 있었다.

• Journal of the Korea Society of Computer and Information
• /
• v.19 no.8
• /
• pp.45-54
• /
• 2014

In this paper, we proposed an effective quality improvement scheme of magnified image using the surface characteristics in image. If the surface in image is estimated as simple convex surface or simple concave surface, the interpolated value can be calculated to have the surface characteristics by using the other method in the proposed scheme. The calculated value becomes the interpolated pixel value inmagnified image. So, themagnified image reflects the surface characteristics of the real image. If the surface is not estimated as simple convex surface or simple concave surface, the interpolated value is calculated more accurately than bilinear interpolation by using the method of the proposed scheme. The PSNR values of the magnified images using the proposed schemes are greater than those of the magnified images using the previous interpolation schemes.

• Journal of the Korea Computer Industry Society
• /
• v.4 no.10
• /
• pp.581-586
• /
• 2003

In this paper, we propose a region-based deinterlacing algorithm which is computationally efficient and shows effective visual performance. Computer simulation results indicate that the proposed algorithm shows better visual performance than conventional deinterlacing algorithms.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2005.11a
• /
• pp.173-176
• /
• 2005

본 논문에서 영역 분할을 기반으로 하는 새로운 scan rate of-conversion 기법을 제안한다. 제안하는 공간적 보간 방법과 기존의 시간적 보간 방법을 이용하여 시공간 보간을 한다. 제안된 방법에서는 먼저 분수령 알고리즘을 이용한 영역 분할하고, 분할된 영역을 기반으로 하여 영역간의 에지 방향성을 결정한다. 기존 알고리즘과 같이 화소간 값 차이를 이용하여 에지의 방향을 구하지 않고 분할된 영역에서 영역구분에 따른 방향에 따라 보간을 함으로써 에지를 유지하면서 scan rate up-conversion을 할 수 있다.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.18 no.7
• /
• pp.734-740
• /
• 2017

A frame rate up-conversion scheme is needed when moving pictures with a low frame rate is played on appliances with a high frame rate. Frame rate up-conversion methods interpolate the frame with two consecutive frames of the original source. This can be divided into the frame repetition method and motion estimation-based the frame interpolation one. Frame repetition has very low complexity, but it can yield jerky artifacts. The interpolation method based on a motion estimation and compensation can be divided into pixel or block interpolation methods. In the case of pixel interpolation, the interpolated frame was classified into four areas, which were interpolated using different methods. The block interpolation method has relatively low complexity, but it can yield blocking artifacts. The proposed method is the frame rate up-conversion method based on a block motion estimation and compensation using the linearity of motion. This method uses two previous frames and one next frame for motion estimation and compensation. The simulation results show that the proposed algorithm effectively enhances the objective quality, particularly in a high resolution image. In addition, the proposed method has similar or higher subjective quality than other conventional approaches.