• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.34 no.4C
• /
• pp.388-396
• /
• 2009

This paper proposes a new fire-smoke detection method by using extracted features from camera images and pattern recognition technique. First, moving regions are detected by analyzing the frame difference between two consecutive images and generate candidate smoke regions by applying smoke color model. A smoke region generally has a few characteristics such as similar color, simple texture and upward motion. From these characteristics, we extract brightness, wavelet high frequency and motion vector as features. Also probability density functions of three features are generated using training data. Probabilistic models of smoke region are then applied to observation nodes of our proposed Dynamic Bayesian Networks (DBN) for considering time continuity. The proposed algorithm was successfully applied to various fire-smoke tasks not only forest smokes but also real-world smokes and showed better detection performance than previous method.

• Journal of the Institute of Convergence Signal Processing
• /
• v.20 no.4
• /
• pp.205-211
• /
• 2019

Steganalysis is a technique that aims to detect and recover data hidden by steganography. Steganalytic methods detect hidden data by analyzing visual and statistical distortions caused during data embedding. However, for recovering the hidden data, they need to know which steganographic methods the hidden data has been embedded by. Therefore, we propose a hierarchical convolutional neural network (CNN) structure that identifies a steganographic method applied to an input image through multi-level classification. We trained four base CNNs (each is a binary classifier that determines whether or not a steganographic method has been applied to an input image or which of two different steganographic methods has been applied to an input image) and connected them hierarchically. Experimental results demonstrate that the proposed hierarchical CNN structure can identify four different steganographic methods (LSB, PVD, WOW, and UNIWARD) with an accuracy of 79%.

• Journal of the Korean earth science society
• /
• v.30 no.6
• /
• pp.683-691
• /
• 2009

Geomagnetic variation around May 2, 2009 when Angdong earthquake broke out was analyzed using the data recorded at the Cheong-yang geomagnetic observatory, KMA. Firstly, we predict the geomagnetic variation by PCA analysis of geomagnetic data, and then compare the predicted value with the observed data to find any significant differences in residuals. Secondly, wavelet semblance technique is applied to compare the time series before and after the earthquake. Some meaningful change is detected in the Z-field. Thirdly, eigen value analysis for the 3 component geomagnetic data is performed. The location of the observatory was too far from the epicenter and the magnitude was too small to find decisive precursory phenomenon. Nevertheless we can detect some significant correlation between the earthquake and the variation of the geomagnetic field. Various signal processing methods applied in this study will give some opportunity to find precursory effects in the future.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2001.10a
• /
• pp.487-490
• /
• 2001

With the establishment of the optimized copyright, digital image watermarking technique is demended to identify the owner of a certain image and to avoid the unauthorized distribution of digital image copies. Also, a robust watermarking approach should survive several possible attacks, such as image processing and lossy image compession. The proposed scheme distributes the 256 gray-level signature information in discrete wavelet transform domain of the host image where is very little visible distortion. While much of the privious work used signature data that is a small fraction of th e host images the proposed approach can easily handle gray-scale Images. As the result, stable reconstruction can be obtained even when the images are transformed, JPEG lossy compression or otherwise modified by low-pass filtering operations.

• Journal of the Korea Computer Graphics Society
• /
• v.6 no.4
• /
• pp.35-42
• /
• 2000

3D texture mapping generates highly natural visual effects in which objects appear carved from lumps of materials rather than laminated with thin sheets as in 2D texture mapping. Storing 3D texture images in a table for fast mapping computations, instead of evaluating procedures on the fly, however, has been considered impractical due to the extremely high memory requirement. Recently, a practical real-time 3D texture mapping technique was proposed in [11], where they attempt to resolve the potential texture memory problem by compressing 3D textures using a wavelet-based encoding method. In this paper, we consider two other encoding schemes that could also be applied to the compression-based 3D texture mapping. In particular, we extend the vector quantization and FXT1 for 3D texture compression, and compare their performance with the wavelet-based encoding scheme.

• The Journal of Information Technology
• /
• v.7 no.1
• /
• pp.155-161
• /
• 2004

During normal machine operation, partial discharge(PD) measurements were performed with turbine generator analyzer(TGA) in imitation stator winding of high voltage motors. The motor was energized to 4.47kV, 6.67, respectively. Applied voltage to Imitation winding was used two voltage level, 4.47[kV] and 6.67[kV]. Motors having imitation stator winding were installed with 80pF capacitive couplers at the terminal box. Case of PD Pattern regarding applied voltage phase angel, the PD patterns were displayed two dimensional and three dimensional. TGA summarizes each plot with two quantities such as the normalized quantity number(NQN) and the peak PD magnitude(Qm). As the result, we could discrimidate using TGA the difference of internal and surface discharge for imitation stator winding. We have used the other technique, in order to feature extraction of faulty signals on stator winding, Daubechies Discrete wavelet transform and Harmonics analysis(FFT) about faulty signals.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.22 no.8
• /
• pp.1790-1802
• /
• 1997

A wavelet transform coding method that fulfills the requirements of HD-VCR(high definition video casstte recorder) for studio applications in proposed. A constant bit rate is achieved by a forward rate control technique whcih determins the quantizer stepsize based on the coding results fo the previous frame. We also propose a two-level coder that consists of both the IDC(independently decodable code) and the DDC(dependently decodable code). To minimize error propagation, the transformed coefficients are restructured into transform blocks which are represented by a tree structure. The result shows thta the proposed coding scheme produces better picture quality with block effects than that of DCT(discrete cosine transform) based coding schemes at the same compression ratio. The proposed method meets most of the requirements of HD-VCR.

• Journal of Korea Multimedia Society
• /
• v.6 no.2
• /
• pp.247-258
• /
• 2003

Principal component analysis (PCA) based algorithms form the basis of numerous algorithms and studies in the face recognition literature. PCA is a statistical technique and its incorporation into a face recognition system requires numerous design decisions. We explicitly take the design decisions by in-troducing a generic modular PCA-algorithm since some of these decision ate not documented in the literature We experiment with different implementations of each module, and evaluate the different im-plementations using the September 1996 FERET evaluation protocol (the do facto standard method for evaluating face recognition algorithms). We experiment with (1) changing the illumination normalization procedure; (2) studying effects on algorithm performance of compressing images using JPEG and wavelet compression algorithms; (3) varying the number of eigenvectors in the representation; and (4) changing the similarity measure in classification process. We perform two experiments. In the first experiment, we report performance results on the standard September 1996 FERET large gallery image sets. The result shows that empirical analysis of preprocessing, feature extraction, and matching performance is extremely important in order to produce optimized performance. In the second experiment, we examine variations in algorithm performance based on 100 randomly generated image sets (galleries) of the same size. The result shows that a reasonable threshold for measuring significant difference in performance for the classifiers is 0.10.

• Korean Journal of Remote Sensing
• /
• v.22 no.1
• /
• pp.49-61
• /
• 2006

The lifting scheme has been found to be a flexible method for constructing scalar wavelets with desirable properties. In this paper, it is extended to the UDAR data compression. A newly developed data compression approach to approximate the UDAR surface with a series of non-overlapping triangles has been presented. Generally a Triangulated Irregular Networks (TIN) are the most common form of digital surface model that consists of elevation values with x, y coordinates that make up triangles. But over the years the TIN data representation has become an important research topic for many researchers due its large data size. Compression of TIN is needed for efficient management of large data and good surface visualization. This approach covers following steps: First, by using a Delaunay triangulation, an efficient algorithm is developed to generate TIN, which forms the terrain from an arbitrary set of data. A new interpolation wavelet filter for TIN has been applied in two steps, namely splitting and elevation. In the splitting step, a triangle has been divided into several sub-triangles and the elevation step has been used to 'modify' the point values (point coordinates for geometry) after the splitting. Then, this data set is compressed at the desired locations by using second generation wavelets. The quality of geographical surface representation after using proposed technique is compared with the original UDAR data. The results show that this method can be used for significant reduction of data set.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.29 no.5A
• /
• pp.421-431
• /
• 2009

This paper is the study on the verification of structural health monitoring (SHM) algorithm based on the ultrasonic guided wave. An active inspection system using Lamb wave (LW) for SHM was considered. The basic study about the application of this algorithm was performed for detecting the circular notch defect in steel plate. LW testing technique, pitch-catch method, was used for interpretation of circular notch defect with depth of 50% of plate thickness and 7 mm width. Damage characterization takes place by comparing $S_0$ mode sensor signals collected before and after the damage event. By subtracting the signals of both conditions from each other, a scatter signal is produced which can be used for damage localization. The continuous Gabor wavelet transform is used to attain the time between the arrivals of the scatter and sensor signals. A new practical damage monitoring algorithm, based on damage monitoring polygon and pitch-catch method, has been proposed and verified with good accuracy. The possible damage location can be estimated by the average on calculated location points and the damage extent by the standard deviation.