• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.4 no.2
• /
• pp.505-516
• /
• 2000

Cellular Neural Networks(CNN) is a nonlinear information-processing system that has a locally connected characteristic and is widely used in the real-time high speed image processing. In this paper, a practical system approach of time-multiplexing CNN implementations suitable for processing large and complex images using small CNN arrays is presented and $6\times6$ CNN hardware is designed for the processing of a large image. While previous implementations are mostly suitable for black and white applications because of the thresholded outputs, our approach is especially suitable for applications in gray image processing due to the analog nature of the state node. CNN chip is designed using a 0.65${\mu}{\textrm}{m}$ 2P2M(double poly, double metal) N-Well CMOS process technology. It contains about 15,400 devices on an area of about $1.85\times1.75$ md. The designed $6\times6$ CNN is tested for the edge detection of a large image input and it's performance is verified.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.49 no.4
• /
• pp.95-104
• /
• 2012

In the image captured by the camera in a moving vehicle, there exist various motions due to the change of background, motion of objects in the image that make difficult to extract a pure vibrational motion by the camera. In this paper, we suggest an image stabilization with the elimination of various motion components based on the classification of motions in the image by their characteristics. After the elimination of various local motions, images are compensated and stabilized with the global motion caused by the camera. Also, we suggest an accurate and fast image stabilization by excluding regions of little information based on block differences and edge densities.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.43 no.6 s.312
• /
• pp.68-75
• /
• 2006

During cancer therapy by using high energy radiation, it is possible to improve the radiation therapy efficiency by performing a precise radiation therapy after verification of generated setup errors. In this paper, the video based electronic portal imaging device (EPID) which could display the portal image with near real time was developed to verify treatment position errors in radiation therapy instead of an analog typed portal film. This EPID system for applying QA tool of radiation therapy machine was consisted of a metal/fluorescent screen, $45^{\circ}$mirror, camera and image grabber. Radiation field verification has been performed to check quality assurance of the treatment machine itself by using this EPID system. The radiation field error was easily observed by edge detection of irradiated field size on EPID image when $0.6^{\circ}$ shift of collimator angle was generated. So, this implemented EPID system could be used as a radiation QA tool.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.10 no.5
• /
• pp.865-870
• /
• 2006

The cellular neural networks have shown a vast computing power for the image processing in spite of the simplicity of its structure. However, it is impossible to implement the CNN hardware which would require the same enormous amount of cells as that of the pixels involved in the practical large image. In this parer, the $5\times5$ CNN hardware and the pre post processor which can be used for processing the real large image with a time-multiplexing scheme are implemented. The implemented $5\times5$ CNN hardware and pre post processor is applied to the edge detection of $256\times256$ lena image to evaluate the performance. The total number of block. By the time-multiplexing process is about 4,000 blocks and to control pulses are needed to perform the pipelined operation or the each block. By the experimental resorts, the implemented $5\times5$ CNN hardware and pre post processor can be used to the real large image processing.

• The Transactions of the Korea Information Processing Society
• /
• v.6 no.8
• /
• pp.2271-2280
• /
• 1999

In this paper, a variable motion estimation scheme based on HBMA(Hierarchical Block Matching Algorithm) to improve the performance and to reduce heavy computational and transmission load, is presented. The proposed algorithm is composed of four steps. First, block activity for each block is defined using the edge information of differential image between two sequential images, and then average block activity of the present image is found by taking the mean of block activity. Secondly, camera pan compensation is carried out, according to the average activity of the image, in the hierarchical pyramid structure constructed by wavelet transform. Next, the LUT classifying each block into one among Moving, No Moving, Semi-Moving Block according to the block activity compensated camera pan is obtained. Finally, as varying the block size and adaptively selecting the initial search layer and the search range referring to LUT, the proposed variable HBMA can effectively carries out fast motion estimation in the hierarchical pyramid structure. The cost function needed above-mentioned each step is only the block activity defined by the edge information of the differential image in the sequential images.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.17 no.7
• /
• pp.1672-1686
• /
• 2013

In this paper, we propose a hybrid super-resolution algorithm robust to cut-change. Existing single-frame based super-resolution algorithms are usually fast, but quantity of information for interpolation is limited. Although the existing multi-frame based super-resolution algorithms generally robust to this problem, the performance of algorithm strongly depends on motions of input video. Furthemore at boundary of cut, applying of the algorithm is limited. In the proposed method, we detect a define boundary of cut using cut-detection algorithm. Then we adaptively apply a single-frame based super-resolution method to detected cut. Additionally, we propose algorithms of normalizing motion vector and analyzing pattern of edge to solve various problems of existing super-resolution algorithms. The experimental results show that the proposed algorithm has better performance than other conventional interpolation methods.

• Journal of the Institute of Convergence Signal Processing
• /
• v.11 no.2
• /
• pp.157-162
• /
• 2010

In this paper we propose a fractal analysis based on the discrete wavelet transform. It is well known that Fourier Transform is widely used for frequency analysis of random signal. However, the frequency domain is not used for expressing the sudden signal change and non-stationary signal at the time-axis by this method. Maximum value in the wavelet modules can be expressed by the Lipschitz exponent, which is useful to represent the characteristics of signal or the edge of an image. It is possible to reconstruct the original image only by using the few maximum points. The v possible image It iusing oil was acquired to interpret the maximum value. ufter that, it was applied to the v possible image of a ship model. In addition, the fractal dimens by by the conlapse process of the sediment particle was examined. In this paper, the fractal dimens by has been obtained by the maximum value and the experiment obtained from the visualized image also acquired the same result as existing methods.

• The Journal of the Korea Contents Association
• /
• v.11 no.5
• /
• pp.84-92
• /
• 2011

The automated segmentation technique of cell region in Bio Images helps biologists understand complex functions of cells. It is mightly important in that it can process the analysis of cells automatically which has been done manually before. The conventional methods for segmentation of cell and nuclei from multi-channel images consist of two steps. In the first step nuclei are extracted from DNA channel, and used as initial contour for the second step. In the second step cytoplasm are segmented from Actin channel by using Active Contour model based on intensity. However, conventional studies have some limitation that they let the cell segmentation performance fall by not considering inhomogeneous intensity problem in cell images. Therefore, the paper consider correlation between DNA and Actin channel, and then proposes the preprocessing algorithm by which the brightness of cell inside in Actin channel can be compensated homogeneously by using DNA channel information. Experiment result show that the proposed preprocessing method improves the cell segmentation performance compared to the conventional method.

• 전자공학회논문지 IE
• /
• v.48 no.2
• /
• pp.32-39
• /
• 2011

The output image of the uncooled thermal infrared camera is difficult the identification of target because of the limited dynamic range and the various noises. Retinex algorithm based on the theory of the human visual perception is known to be effective contrast enhancement technique. However, the image quality is insufficient when it is adopted to the narrow dynamic range image as the infrared image. In this paper, we propose the revised retinex algorithm to enhance the contrast of the infrared image. To improve the contrast enhancement performance, we designed the new dynamic range compression function instead of log function. To reduce the noise and compensate the loss of edge, we added the contrast compensation procedure in the MSR image generation process. According to the output picture comparing and numerical analysis, the proposed algorithm shows the better contrast enhancement performance and the more suitable method for the infrared image enhancement.

• Journal of Korea Game Society
• /
• v.14 no.5
• /
• pp.97-106
• /
• 2014

This study proposes an efficient method to model various and diverse leaves required to express digital plants such as flowers and trees in virtual landscape easily and intuitively. The proposed procedural method divides a leaf mainly into a blade and vein thereby detecting contours from binary images that correspond to blades and generating leaves by modeling leaf veins procedurally based on the detected contours. First of all, a complicated leaf vein structure is divided into main veins, lateral veins, and tertiary vein while all veins grow procedurally directing from start auxin to destination auxin. Here, to calculate destination auxin required for growth automatically, approximated contours from binary images that correspond to blades are found thereby calculating candidate destination auxin. Finally, natural digital leaves are generated by applying a color combination method. Through the proposed method, natural and various leaves can be generated and whether the proposed method is efficient or not is verified through the experiment.