• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.22 no.4
• /
• pp.675-684
• /
• 1997

The bit rate control algorithm which is capable of handing scene change is proposed. In MPEG-2 TM5, block variance is used to measure block activity. But block variance is not consistent with human visual system and does not differenciate the distribution of pixel values within the block. In target bit allocation process of TM5, global complexity, obtained by results of previous coded pictures, is used. Since I pictures are spaced relatively far apart, their complexity estimate is not very accurate. In the proposed algorithm local variance is used to measure block activity and detect scene change. Local variance, using deviation from the mean of neighboring pixels, well represents the distribution of pixel values within the block. If scene change is detected, the local variance information is used for target bit allocation process. Allocating target bits for I picture, the average local variance difference between previous and current I picture is considered. The experimental results show that the proposed algorithm can detect scene change very precisely and gives better picture quality and higher PSNR values than MPEG-2 TM5.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2014.10a
• /
• pp.802-804
• /
• 2014

Edge detection of image for performing the road lane recognition is an essential preprocessing. Various studies are being performed in order to detect such edge and there are conventional edge detection methods such as Sobel, Prewitt and Roberts. Such methods regardless of pixel distribution are processed by applying the same weighted value to the entire pixels and show a somewhat insufficient edge detection results. Therefore, this paper has proposed an algorithm that detects the edge using the suitable weighted value for the road lane recognition considering the pixel distribution shape of the image.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2008.06a
• /
• pp.697-698
• /
• 2008

• Journal of radiological science and technology
• /
• v.14 no.1
• /
• pp.61-72
• /
• 1991

In this paper, the rotating plate method extracting signal and reconstructing original image was proposed. The rotating methode has cell detector array each of which has used in the medical diagnosis X-ray photography. The major problem using the simple horizontal moving or non-moving methode is the size and number of detector cells which have the considerable affection on the sharpness and resolution of the reconstructed image. Secondary, the estimated pixel values of non-detected real points which are placed between detector cells will be the distorted pixels in the reconstructed image. Therefore, the proposed rotating plate method has the exact distribution on the uncertain pixels which were reconstructed by conventional methods to solve there problems. And then, the image using the rotated plate's cell out put signal was reconstructed on the computer simulation. The method will rotated the detector array plate to solve the reconstruction from the detector size and number of conventional methods. The result of simulation has estimated the original pixel position and 81 pixel/mm resolution which the reconsiderlation of the detector's moving orientation, the proposed method has 25 pixel/mm resolution. These results have been represented by 3-D computer graphics.

• 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.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.05a
• /
• pp.428-430
• /
• 2022

Image processing is playing an important part in automation and artificial intelligence systems, such as object tracking, object recognition and classification, and the importance of IoT technology and automation is emphasizing as interest in automation increases. However, in a system that requires detailed data such as an image boundary, a precise noise removal algorithm is required. Therefore, in this paper, we propose a filtering algorithm based on the pixel value distribution pattern to minimize the information loss in the filtering process. The proposed algorithm finds the distribution pattern of neighboring pixel values with respect to the pixel values of the input image. Then, a weight mask is calculated based on the distribution pattern, and the final output is calculated by applying it to the filtering mask. The proposed algorithm has superior noise removal characteristics compared to the existing method and restored the image while minimizing blurring.

• Smart Media Journal
• /
• v.4 no.3
• /
• pp.50-55
• /
• 2015

In this paper, we present background modeling method based on Gaussian mixture model to subtract background for night-time video surveillance. In night-time video, it is hard work to distinguish the object from the background because a background pixel is similar to a object pixel. To solve this problem, we change the pixel of input frame to more advantageous value to make the Gaussian mixture model using scaled histogram stretching in preprocessing step. Using scaled pixel value of input frame, we then exploit GMM to find the ideal background pixelwisely. In case that the pixel of next frame is not included in any Gaussian, the matching test in old GMM method ignores the information of stored background by eliminating the Gaussian distribution with low weight. Therefore we consider the stacked data by applying the difference between the old mean and new pixel intensity to new mean instead of removing the Gaussian with low weight. Some experiments demonstrate that the proposed background modeling method shows the superiority of our algorithm effectively.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2007.06a
• /
• pp.328-328
• /
• 2007

Large area matrix-addressed image detectors are a recent technology for x-ray imaging with medical diagnostic and other applications. The imaging properties of x-ray pixel detectors depend on the quantum efficiency of x-rays, the generated signal of each x-ray photon and the distribution of the generated signal between pixels. In a phosphor coated detector the light signal is generated by electrons captured in the phosphor screen. In our study we simulated the lateral spread distributions for phosphor coupled detector by Monte Carlo simulations. Most simulations of such detectors simplify the setup by only taking the conversion layer into account neglecting behind. The Monte Carlo code MCNPX has been used to simulate the complete interaction and subsequent charge transport of x-ray radiation. This has allowed the analysis of charge sharing between pixel elements as an important limited factor of digital x-ray imaging system. The parameters are determined by lateral distribution of x-ray photons and x-ray induced electrons. The primary purpose of this study was to develop a design tool for the evaluation of geometry factor in the phosphor coupled optical imaging detector. In order to evaluate the spatial resolution for different phosphor material, phosphor geometry we have developed a simulation code. The developed code calculates the energy absorption and spatial distribution based on both the signal from the scintillating layer and the signal from direct detection of x-ray in the detector. We show that internal scattering contributes to the so-called spatial resolution drop of the image detector. Results from the simulation of spatial distribution in a phosphor pixel detector are presented. The spatial resolution can be increased by optimizing pixel size and phosphor thickness.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2013.10a
• /
• pp.73-76
• /
• 2013

This paper proposes a tongue and its direction recognition algorithm which compares and estimates pixel distribution in the mouth area. As the size of smart phones grows, facial gesture control technology for a smart phone is required. Firstly, the nose area is detected and the mouth area is detected based on the ratio of the nose to mouth. After detecting the mouth area, it is divided by a pattern of grid and the distribution of pixels having the similar color to the tongue is tested for each segment. The recognition rate was nearly 80% in the experiments performed with five researchers among our laboratory members.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2019.05a
• /
• pp.426-428
• /
• 2019

Digital images are compromised by noise for various reasons, such as camera sensor malfunctions and hardware errors. Since AWGN can be found in most of electronic equipment, AWGN removal is essential in various image processing processes. In this paper, we propose a filter algorithm that eliminates noise considering the pixel characteristics in AWGN environments. In order to compensate this, the filtering range is set considering the distribution of the pixels inside the mask. The output of the filter suitable for each component is adjusted by adding or subtracting the weight according to the normal distribution. Set the output. To evaluate the performance of the proposed algorithm, we compared it with the existing method using simulation.