• Journal of Korea Multimedia Society
• /
• v.8 no.11
• /
• pp.1510-1519
• /
• 2005

We proposed a digital watermarking method that it is possible to identify the copyright because the watermark is detected in the first print-scan and to protect a forgery because the watermark is not detected in the second print-scan. The proposed algorithm uses LPM and DFT transform for the robustness to the distortion of pixel value and geometrical distortion. This methods could improve watermark detection performance and image quality by selecting maximum sampling radius in LPM transform. After analyzing the characteristics of print-scan process, we inserted the watermark in the experimentally selected frequency bands that survives robustly to the first print-scan and is not detected in the second print-scan, using the characteristic of relatively large distortion in high frequency bands of DFT As the experimental result, the original proof is possible because average similarity degree 5.13 is more than the critical value 4.0 in the first print-scan. And the detection of forgery image is also possible because average similarity degree 2.76 is less than the critical value 4.0 in the second print-scan.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.21 no.11
• /
• pp.2029-2036
• /
• 2017

Currently, the demand for multimedia services is increasing with the rapid development of the digital age. Image data is corrupted by various noises and typical noise is mainly AWGN, salt and pepper noise and the complex noise that these two noises are mixed. Therefore, in this paper, the noise is processed by classifying AWGN and salt and pepper noise through noise judgment. In the case of AWGN, the outputs of spatial weighted filter and pixel change weighted filter are composed and processed, and the composite weights are applied differently according to the standard deviation of the local mask. In the case of salt and pepper noise, cubic spline interpolation and local histogram weighted filters are composed and processed. This study suggested the multiple image restoration filter algorithm which is processed by applying different composite weights according to the salt and pepper noise density of the local mask.

• Korean Journal of Remote Sensing
• /
• v.30 no.2
• /
• pp.219-226
• /
• 2014

Fractional values resulted from the spectral mixture analysis could be used to classify not only urban area with various materials but also forest area in more detailed spatial scale. Especially South Korea is largely consist of mixed forest, so the spectral mixture analysis is suitable as a classification method. For the successful classification using spectral mixture analysis, extraction of optimal endmembers is prerequisite process. Though geometric endmember selection has been widely used, it is barely suitable for forest area. Therefore, in this study, we modified Iterative Error Analysis (IEA), one of the most famous algorithms of image endmember selection which extracts pure pixel directly from the image. The endmembers which represent deciduous and coniferous trees are automatically extracted. The experiments were implemented on two sites of Compact Airborne Spectrographic Imager (CASI) and classified forest area into two types. Accuracies of each classification results were 86% and 90%, which mean proposed algorithm effectively extracted proper endmembers. For the more accurate classification, another substances like forest gap should be considered.

• Korean Journal of Remote Sensing
• /
• v.30 no.2
• /
• pp.207-218
• /
• 2014

The information on the land cover along stream corridor is important for stream restoration and maintenance activities. This study aims to review the different classification methods for mapping the status of stream corridors in Seom River using airborne RGB and CIR digital ortho imagery with a ground pixel resolution of 0.2m. The maximum likelihood classification, minimum distance classification, parallelepiped classification, mahalanobis distance classification algorithms were performed with regard to the improvement methods, the skewed data for training classifiers and filtering technique. From these results follows that, in aerial image classification, Maximum likelihood classification gave results the highest classification accuracy and the CIR image showed comparatively high precision.

• The Korean Journal of Applied Statistics
• /
• v.31 no.5
• /
• pp.539-554
• /
• 2018

Effective interaction between user and device is considered an important ability of IoT devices. For some applications, it is necessary to recognize human facial expressions in real time and make accurate judgments in order to respond to situations correctly. Therefore, many researches on facial image analysis have been preceded in order to construct a more accurate and faster recognition system. In this study, we constructed an automatic recognition system for facial expressions through two steps - a facial recognition step and a classification step. We compared various models with different sets of data with pixel information, landmark coordinates, Euclidean distances among landmark points, and arctangent angles. We found a fast and efficient prediction model with only 30 principal components of face landmark information. We applied several prediction models, that included linear discriminant analysis (LDA), random forests, support vector machine (SVM), and bagging; consequently, an SVM model gives the best result. The LDA model gives the second best prediction accuracy but it can fit and predict data faster than SVM and other methods. Finally, we compared our method to Microsoft Azure Emotion API and Convolution Neural Network (CNN). Our method gives a very competitive result.

• Journal of the Society of Naval Architects of Korea
• /
• v.51 no.4
• /
• pp.311-320
• /
• 2014

Test uncertainty of a towed underwater Stereoscopic Particle Image Velocimetry (SPIV) system was assessed in a towing tank. To estimate the systematic error and random error of mean velocity and turbulence properties measurement, velocity field of uniform flow was measured. Total uncertainty of the axial component of mean velocity was 1.45% of the uniform flow speed and total uncertainty of turbulence properties was 3.03%. Besides, variation of particle displacement was applied to identify the change of error distribution. In results for variation of particle displacement, the error rapidly increases with particle movement under one pixel. In addition, a nominal wake of a model ship was measured and compared with existing experimental data by five-hole Pitot tubes, Pitot-static tube, and hot wire anemometer. For mean velocity, small local vortex was identified with high spatial resolution of SPIV, but has serious disagreement in local maxima of turbulence properties due to limited sampling rate.

• Journal of the Korea Society of Computer and Information
• /
• v.3 no.2
• /
• pp.57-67
• /
• 1998

In order to extract the contour of interesting object in the image, Kass suggested the Active Contour Model called "Snakes" The speed of this model is slow and this model is sensitive of initialization. In order to improve these problems, Gunn extracted the accurate contour by using two initializations, and operated to less sensitive of initialization. This method could extract more accurate contour than the existing method, but it had no effect in the speed and it was sensitive of noise. This paper applied to the Energy Minimization Algorithm about only the pixel within the window applying the window of $8{\times}8$ size at each contour point consisting Snakes in order to solve these problems. In order to less sensitive of noise which exists within image, it suggests a method that moves the window to vertical direction for the gradient of each contour point.our point.

• Journal of the Korea Society of Computer and Information
• /
• v.17 no.11
• /
• pp.27-32
• /
• 2012

In this paper, we propose a fuzzy method that improves the existing problem of the fuzzy filtering algorithm. The proposed fuzzy filtering algorithm separates R, G, and B channels from the color image. Mask information was extracted from separated channels and the brightness of the mean value and median value for channels was applied in the function of the proposed fuzzy method to calculate the membership and achieve application in the inference rule. Also, the membership degrees of R, G, and B were used to distinguish the possibility of noise. The proposed fuzzy method selected three membership functions. If noise is distinguished, the noise is eliminated by selecting the median value or mean value as the relevant pixel value according to the degree of noise. By applying the proposed method in color images, it was verified that the proposed method is more effective in eliminating noise when compared with the conventional fuzzy filtering method.

• Geophysics and Geophysical Exploration
• /
• v.6 no.1
• /
• pp.7-12
• /
• 2003

The lineament extraction from satellite images is important in the geologic studies including groundwater and mineral exploration, groundwater survey, natural hazard analysis, and many others. The lineaments in remote sensing images are identified by the difference of pixel values or brightness. Since the visual interpretation is apt to be influenced by the knowledges and experiences, many of the automatic lineament detection algorithms are developed to ensure the objectives and efficient outputs. DSTA (dynamic segment tracing algorithm) is one of such algorithms, which can be applied to not only mountainous area but also alluvial area. However, when the alluvial area is wider than mountain region, somewhat severe noises are generated. To reduce such noises, AERA (alluvial effect reducing algorithm) is proposed and tested for the image which contains mountains, cultivated land and urban area. Upon the application of AERA, alluvial effects in lineament extraction from satellite image are substantially reduced.

• The KIPS Transactions:PartB
• /
• v.10B no.7
• /
• pp.777-784
• /
• 2003

In this Paper. we propose a recognition method of machine printed characters based on the combination of recognition units using multiple neural networks. In our recognition method, the input character is classified into one of 7 character types among which the first 6 types are for Hangul character and the last type is for non-Hangul characters. Hangul characters are recognized by several MLP (multilayer perceptron) neural networks through two stages. In the first stage, we divide Hangul character image into two or three recognition units (HRU : Hangul recognition unit) according to the combination fashion of graphemes. Each recognition unit composed of one or two graphemes is recognized by an MLP neural network with an input feature vector of pixel direction angles. In the second stage, the recognition aspect features of the HRU MLP recognizers in the first stage are extracted and forwarded to a subsequent MLP by which final recognition result is obtained. For the recognition of non-Hangul characters, a single MLP is employed. The recognition experiments had been performed on the character image database collected from 50,000 real letter envelope images. The experimental results have demonstrated the superiority of the proposed method.