• Proceedings of the Korean Society for Agricultural Machinery Conference
• /
• 2000.11b
• /
• pp.227-235
• /
• 2000

Agricultural products are easily deformable its shape because of some external forces. However, these force behavior is difficult to measure quantitatively. Until now, many researches on the mechanical property was performed with various methods such as material testing, chemical analysis and non-destructive methods. In order to investigate force behavior on the cellular unit of agricultural products, electro-microscope based 3D image processing method will contribute to analysis of plant cells behavior. Before image measurement of plant cells, plant sample was cut off cross-sectioned area in a size of almost 300-400 ${\mu}$ m units using the micron thickness device, and some of preprocessing procedure was performed with fixing and dyeing. However, the wall structure of plant cell is closely neighbor each other, it is necessary to separate its boundary pixel. Therefore, image merging and shrinking algorithm was adopted to avoid disconnection. After then, boundary pixel was traced through thinning algorithm. Each image from the electro-microscope has a information of x,y position and its height along the z axis cross sectioned image plane. 3D image was constructed using the continuous image combination. Major feature was acquired from a fault image and measured area, thickness of cell wall, shape and unit cell volume. The shape of plant cell was consist of multiple facet shape. Through this measured information, it is possible to construct for structure shape of unit plant cell. This micro unit image processing techniques will contribute to the filed of agricultural mechanical property and will use to construct unit cell model of each agricultural products and information of boundary will use for finite element analysis on unit cell image.

• Korean Journal of Optics and Photonics
• /
• v.11 no.1
• /
• pp.58-64
• /
• 2000

Assuming that the performance of holographic storage media is ideal, we estimate the area storage density of disk-type holographic memories, when the method of either angle multiplexing, or rotational multiplexing, or both are used. The area storage density is strongly dependent on the f numbers (ratio of focal length to diameter) of both the Fourier transform lens in the signal arm, denoted by $F/#_2$, and the angle range over which the reference beam is incident (or, the equivalent f number corresponding to the angle range denoted by $F/#_1$). The area storage density is largely independent of the pixel pitch of the spatial light modulator when the Fourier plane holograms are recorded, while it is sensitive to the pixel pitch when the image plane holograms are recorded. In general, to obtain high area storage density, the Fourier or at least near Fourier plane holograms rather than the image plane holograms should be recorded. In addition, when the thickness of the recording materials are less than approximately $500\mu\extrm{m}$, rotational multiplexing gives higher area storage densities than angle multiplexing does. To increase the storage density further, however, it is desirable to use both of the two multiplexing methods in combination.nation.

• Journal of the Korea Convergence Society
• /
• v.11 no.12
• /
• pp.15-22
• /
• 2020

In this paper, we proposed a deep learning based super-resolution method that combines Channel Attention and Spatial Attention feature enhancement methods. It is important to restore high-frequency components, such as texture and features, that have large changes in surrounding pixels during super-resolution processing. We proposed a super-resolution method using feature enhancement that combines Channel Attention and Spatial Attention. The existing CNN (Convolutional Neural Network) based super-resolution method has difficulty in deep network learning and lacks emphasis on high frequency components, resulting in blurry contours and distortion. In order to solve the problem, we used an emphasis block that combines Channel Attention and Spatial Attention to which Skip Connection was applied, and a Residual Block. The emphasized feature map extracted by the method was extended through Sub-pixel Convolution to obtain the super resolution. As a result, about PSNR improved by 5%, SSIM improved by 3% compared with the conventional SRCNN, and by comparison with VDSR, about PSNR improved by 2% and SSIM improved by 1%.

• Tunnel and Underground Space
• /
• v.29 no.3
• /
• pp.184-196
• /
• 2019

This study aims to extract a 3D image of micro-cracks generated by hydraulic fracturing tests, using the deep learning method and X-ray computed tomography images. The pixel-level cracks are difficult to be detected via conventional image processing methods, such as global thresholding, canny edge detection, and the region growing method. Thus, the convolutional neural network-based encoder-decoder network is adapted to extract and analyze the micro-crack quantitatively. The number of training data can be acquired by dividing, rotating, and flipping images and the optimum combination for the image augmentation method is verified. Application of the optimal image augmentation method shows enhanced performance for not only the validation dataset but also the test dataset. In addition, the influence of the original number of training data to the performance of the deep learning-based neural network is confirmed, and it leads to succeed the pixel-level crack detection.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.30 no.6C
• /
• pp.465-474
• /
• 2005

In this paper, a directional reconstruction of lost block in image over noisy channel is presented. DCT coefficients or pixel values in the lost blocks are recovered by using the linear interpolation with available neighboring blocks that are adaptively selected by the directional measure that are composed of the DDC (Difference of DC opposite blocks)and SAC(Similarity of AC opposite blocks) between opposite blocks around lost blocks. The proposed directional recovery method is effective for the strong edge and texture regions because we do not make use of the fixed 4-neighboring blocks but exploit the varying neighboring blocks adaptively by the directional information in the local image. In this paper, we describe the novel directional measure(CDS: Combination of DDC and SAC) composed of the DDC and the SAC and select the usable block to recover the lost block with the directional measure. The proposed method shows about 0.6dB PSNR improvement in average compared to the conventional methods.

• The Journal of Korean Institute of Information Technology
• /
• v.17 no.9
• /
• pp.99-112
• /
• 2019

Cracks on pavement surfaces are critical signs and symptoms of the degradation of pavement structures. Image-based pavement crack detection is a challenging problem due to the intensity inhomogeneity, topology complexity, low contrast, and noisy texture background. In this paper, we address the problem of pavement crack detection and segmentation at pixel-level based on a Deep Neural Network (DNN) using gray-scale images. We propose a novel DNN architecture which contains a modified U-net network and a high-level features network. An important contribution of this work is the combination of these networks afforded through the fusion layer. To the best of our knowledge, this is the first paper introducing this combination for pavement crack segmentation and detection problem. The system performance of crack detection and segmentation is enhanced dramatically by using our novel architecture. We thoroughly implement and evaluate our proposed system on two open data sets: the Crack Forest Dataset (CFD) and the AigleRN dataset. Experimental results demonstrate that our system outperforms eight state-of-the-art methods on the same data sets.

• Journal of the Korean Society of Clothing and Textiles
• /
• v.34 no.3
• /
• pp.540-552
• /
• 2010

This research forms preliminary data for the coordination of environmental color planning in industry through a color analysis of work clothes and the work environment. A digital camera was used to study the work environment of two major shipbuilding companies located in Geoje city and Goseong county. The picture data was analyzed as G (ground: environment) and F (figure: clothes), and analyzed hue, value, and the chroma value through a Muncell conversion 9.0.6 from the color cluster, number of pixel, and RGB value. The results are as follows: First, GY, Y color were mostly used in the shipbuilding environment and work clothes. The color value was used in a relatively wide range but very low chroma (0-3), dark grayish, grayish tone dominated both fields. Second, the use of limited colors cannot be secured for safety in attention of the shipbuilding field. Third, unclear and vogue colors lessened the optical tiredness of workers that helped in the prevention of industrial accidents. Color combination and color selection should be considered for a secure safety color coordination between work clothes and the work environment when it comes to complicated color principles.

• Journal of Korean Society for Geospatial Information Science
• /
• v.13 no.3 s.33
• /
• pp.33-40
• /
• 2005

The aim of remotely sensed data classification is to produce the best accuracy map of the earth surface assigning each pixel to its appropriate category of the real-world. The classification of satellite multi-spectral image data has become tool for generating ground cover map. Many classification methods exist. In this study, MLC(Maximum Likelihood Classification), ANN(Artificial neural network), SVM(Support Vector Machine), Naive Bayes classifier algorithms are compared using IKONOS image of the part of Dalsung Gun, Daegu area. Two preprocessing methods are performed-PCA(Principal component analysis), ICA(Independent Component Analysis). Boosting algorithms also performed. By the combination of appropriate feature selection pre-processing and classifier, the best results were obtained.

• Proceedings of the Korean Vacuum Society Conference
• /
• 1999.07a
• /
• pp.244-244
• /
• 1999

Liquid Crystal Displays(LCDs) 투명성 전도막으로 사용하는 Indium Tin Oxide (ITO)의 고밀도 식각특성을 조사하였다. 특히 ITO식각의 경우, pixel electrode 전극에서 사용되는 underlayer인 SiO2, Si3N4와의 최적의 선택비를 얻는데 중점을 두고 있다. 따라서 본 실험에서는 Inductively Coupled Plasma(ICP)를 이용하여 source power, gas combination, bias voltage, pressure 및 기판온도에 따른 ITO의 식각 특성과 이의 underlayer인 SiO2, Si3N4와의 선택비를 조사하였다. Ar과 CH4를 주된 식각가스로서 사용하였으며 첨가가스로는 O2와 HBr를 사용하였다. ITO의 식각특성을 이해하기 위하여 Quadruple Mass Spectrometry(QMS), Optical emission spectroscopy(OES) 이용하였으며, 식각된 sample의 잔류물을 조사하기 위하여 X-ray photoelectron spectroscopy(XPS)를 이용하여 분석하였다. Ar gas에 적정량의 CH4 혼합이 순수한 Ar 가스로 식각한 경우에 비하여 ITO와 SiO2, Si3N4의 선택비가 높았으며, 더 높은 식각 선택비를 얻기 위하여 Ar/CH 분위기에서 첨가가스 O2, HBr을 사용하였다. Source power 및 bias 증가에 따라 ITO의 식각률은 증가하나, underlayer와의 선택비는 감소함을 보였다. 본 실험에서 측정된 ITO의 high 식각률은 약 1500$\AA$/min이며, SiO2, Si3N4와의 high selectivity는 각각 7:1, 12:1로 나타났다. ITO의 etchrate 및 선택비는 source power, bias, pressure, CH 가스첨가에 의존하였지만 기판온도에는 큰 변화가 없음을 관찰하였다. 또한 적정량의 가스조합으로 식각된 시편의 잔류물을 줄일 수 있었다.

• Journal of Forest and Environmental Science
• /
• v.24 no.3
• /
• pp.151-154
• /
• 2008

Dramatic price increases of fossil fuels and the economic development of emerging nations accelerates the transformation of forest lands into monocultures, e.g. for biofuel production. On this account, cost efficient methods to enable the monitoring of land resources has become a vital ambition. The application of remote sensing techniques has become an integral part of forest attribute estimation and mapping. The aim of this study was to evaluate the potentials of the kNN method by combining terrestrial with remotely sensed data for the development of a pixel-based monitoring system for the small scaled mosaic of different land use types of the off-reserve forests of the Goaso forest district in Ghana, West Africa. For this reason, occurrence and distribution of land use types like cocoa and non-timber forest resources, such as bamboo and raphia palms, were estimated, applying the kNN method to ASTER satellite data. Averaged overall accuracies, ranging from 79% for plantain, to 83% for oil palms, were found for single-attribute classifications, whereas a multi-attribute approach showed overall accuracies of up to 70%. Values of k between 3 and 6 seem appropriate for mapping bamboo. Optimisation of spectral bands improves results considerably.