• KIPS Transactions on Software and Data Engineering
• /
• v.2 no.12
• /
• pp.889-898
• /
• 2013

Due to the popularization of high-performance capturing equipments and the emergence of powerful image-editing softwares, it is easy to make high-quality counterfeit money. However, the probability of detecting counterfeit money to the general public is extremely low and the detection device is expensive. In this paper, a counterfeit money detection algorithm using a general purpose scanner and computer system is proposed. First, the printing features of color printers are calculated using morphological operations and gray-level co-occurrence matrix. Then, these features are used to train a support vector machine classifier. This trained classifier is applied for identifying either original or counterfeit money. In the experiment, we measured the detection rate between the original and counterfeit money. Also, the printing source was identified. The proposed algorithm was compared with the algorithm using wiener filter to identify color printing source. The accuracy for identifying counterfeit money was 91.92%. The accuracy for identifying the printing source was over 94.5%. The results support that the proposed algorithm performs better than previous researches.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.47 no.3
• /
• pp.11-22
• /
• 2010

This paper is a study on implementation of intelligent image surveillance system using context information and supplements temporal-spatial constraint, the weak point in which it is hard to process it in real time. In this paper, we propose scene analysis algorithm which can be processed in real time in various environments at low resolution video(320*240) comprised of 30 frames per second. The proposed algorithm gets rid of background and meaningless frame among continuous frames. And, this paper uses wavelet transform and edge histogram to detect shot boundary. Next, representative key-frame in shot boundary is selected by key-frame selection parameter and edge histogram, mathematical morphology are used to detect only motion region. We define each four basic contexts in accordance with angles of feature points by applying vertical and horizontal ratio for the motion region of detected object. These are standing, laying, seating and walking. Finally, we carry out scene analysis by defining simple context model composed with general context and emergency context through estimating each context's connection status and configure a system in order to check real time processing possibility. The proposed system shows the performance of 92.5% in terms of recognition rate for a video of low resolution and processing speed is 0.74 second in average per frame, so that we can check real time processing is possible.

• Journal of Korean Society of Transportation
• /
• v.33 no.3
• /
• pp.284-293
• /
• 2015

The objective of this paper is to determine the conditions of road surface by utilizing the images collected from closed-circuit television (CCTV) cameras installed on roadside. First, a technique was examined to detect wet surfaces at nighttime. From the literature reviews, it was revealed that image processing using polarization is one of the preferred options. However, it is hard to use the polarization characteristics of road surface images at nighttime because of irregular or no light situations. In this study, we proposes a new discriminant for detecting wet and dry road surfaces using CCTV image data at night. To detect the road surface conditions with night vision, we applied the wavelet packet transform for analyzing road surface textures. Additionally, to apply the luminance feature of night CCTV images, we set the intensity histogram based on HSI(Hue Saturation Intensity) color model. With a set of 200 images taken from the field, we constructed a detection criteria hyperplane with SVM (Support Vector Machine). We conducted field tests to verify the detection ability of the wet road surfaces and obtained reliable results. The outcome of this study is also expected to be used for monitoring road surfaces to improve safety.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.35 no.12
• /
• pp.1301-1307
• /
• 2011

Electrosurgical knee wand (EKW) is a high-frequency thermocautery instrument and is often used for coagulation, ablation, excision, and extirpation of knee ligaments and tissues. In order to maximize the success rate, ease, and safety of knee surgery using EKW and radiofrequency ablation, it is necessary to ensure that the EKW selectively approaches the lesion with utmost accuracy and safety. The key feature of this instrument is its excellent maneuverability. Hence, the authors constructed a tensile spring model based on a shape memory alloy (SMA), which exhibits the shape memory effect. This model can be used in knee surgery as it is considered the most biocompatible femorotibial surgical actuator. The changes in external temperature with current and the thermoelectric characteristics of the SMA were investigated. The relationship between the restoring force and the typical stroke (TS) in response to the conditions in the SMA tensile spring design were evaluated. In conclusion, as the diameter of the SMA tensile spring decreased, the maximum temperature increased. The strain in the actuator caused a stable and proportional increase in the force and induced current for up to 15s, but this increase became very unstable after 30s. Moreover, the relationship between the current and the TS was more stable than that between the current and the restoring force.

• Journal of KIISE
• /
• v.42 no.4
• /
• pp.475-482
• /
• 2015

Developing a high-performance Semantic Role Labeling (SRL) system for a domain requires manually annotated training data of large size in the same domain. However, such SRL training data of sufficient size is available only for a few domains. Performances of Korean SRL are degraded by almost 15% or more, when it is directly applied to another domain with relatively small training data. This paper proposes two techniques to minimize performance degradation in the domain transfer. First, a domain adaptation algorithm for Korean SRL is proposed which is based on the prior model that is one of domain adaptation paradigms. Secondly, we proposed to use simplified features related to morphological and syntactic tags, when using small-sized target domain data to suppress the problem of data sparseness. Other domain adaptation techniques were experimentally compared to our techniques in this paper, where news and Wikipedia were used as the sources and target domains, respectively. It was observed that the highest performance is achieved when our two techniques were applied together. In our system's performance, F1 score of 64.3% was considered to be 2.4~3.1% higher than the methods from other research.

• Journal of Korea Water Resources Association
• /
• v.47 no.3
• /
• pp.285-295
• /
• 2014

We applied the ISPSO-GLUE method, which integrates the Isolated-Speciation-based Particle Swarm Optimization (ISPSO) with the Generalized Likelihood Uncertainty Estimation (GLUE) method, to the uncertainty analysis of the Topography Model (TOPMODEL) and compared its performance with that of the GLUE method. When we performed the same number of model runs for the both methods, we were able to identify the point where the performance of ISPSO-GLUE exceeded that of GLUE, after which ISPSOGLUE kept improving its performance steadily while GLUE did not. When we compared the 95% uncertainty bounds of the two methods, their general shapes and trends were very similar, but those of ISPSO-GLUE enclosed about 5.4 times more observed values than those of GLUE did. What it means is that ISPSOGLUE requires much less number of parameter samples to generate better performing uncertainty bounds. When compared to ISPSO-GLUE, GLUE overestimated uncertainty in the recession limb following the maximum peak streamflow. For this recession period, GLUE requires to find more behavioral models to reduce the uncertainty. ISPSO-GLUE can be a promising alternative to GLUE because the uncertainty bounds of the method were quantitatively superior to those of GLUE and, especially, computationally expensive hydrologic models are expected to greatly take advantage of the feature.

• The Journal of the Acoustical Society of Korea
• /
• v.39 no.5
• /
• pp.414-423
• /
• 2020

In this paper, we propose a Dilated Convolution Gate Linear Unit (DCGLU) to mitigate the lack of sparsity and small receptive field problems caused by the segmentation map extraction process in sound event detection with weak labels. In the advent of deep learning framework, segmentation map extraction approaches have shown improved performance in noisy environments. However, these methods are forced to maintain the size of the feature map to extract the segmentation map as the model would be constructed without a pooling operation. As a result, the performance of these methods is deteriorated with a lack of sparsity and a small receptive field. To mitigate these problems, we utilize GLU to control the flow of information and Dilated Convolutional Neural Networks (DCNNs) to increase the receptive field without additional learning parameters. For the performance evaluation, we employ a URBAN-SED and self-organized bird sound dataset. The relevant experiments show that our proposed DCGLU model outperforms over other baselines. In particular, our method is shown to exhibit robustness against nature sound noises with three Signal to Noise Ratio (SNR) levels (20 dB, 10 dB and 0 dB).

• The Journal of the Acoustical Society of Korea
• /
• v.22 no.5
• /
• pp.375-382
• /
• 2003

Lip-reading technique based on bimodal is to enhance speech recognition rate in noisy environment. It is most important to detect the correct lip-image. But it is hard to estimate stable performance in dynamic environment, because of many factors to deteriorate Lip-reading's performance. There are illumination change, speaker's pronunciation habit, versatility of lips shape and rotation or size change of lips etc. In this paper, we propose the IIR filtering in time-domain for the stable performance. It is very proper to remove the noise of speech, to enhance performance of recognition by digital filtering in time domain. While the lip-reading technique in whole lip image makes data massive, the Principal Component Analysis of pre-process allows to reduce the data quantify by detection of feature without loss of image information. For the observation performance of speech recognition using only image information, we made an experiment on recognition after choosing 22 words in available car service. We used Hidden Markov Model by speech recognition algorithm to compare this words' recognition performance. As a result, while the recognition rate of lip-reading using PCA is 64%, Time-domain filter applied to lip-reading enhances recognition rate of 72.4%.

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.11
• /
• pp.147-155
• /
• 2022

In this study, we developed a system to dynamically balance a daily stock portfolio and performed trading simulations using gradient boosting and genetic algorithms. We collected various stock market data from stocks listed on the KOSPI and KOSDAQ markets, including investor-specific transaction data. Subsequently, we indexed the data as a preprocessing step, and used feature engineering to modify and generate variables for training. First, we experimentally compared the performance of three popular gradient boosting algorithms in terms of accuracy, precision, recall, and F1-score, including XGBoost, LightGBM, and CatBoost. Based on the results, in a second experiment, we used a LightGBM model trained on the collected data along with genetic algorithms to predict and select stocks with a high daily probability of profit. We also conducted simulations of trading during the period of the testing data to analyze the performance of the proposed approach compared with the KOSPI and KOSDAQ indices in terms of the CAGR (Compound Annual Growth Rate), MDD (Maximum Draw Down), Sharpe ratio, and volatility. The results showed that the proposed strategies outperformed those employed by the Korean stock market in terms of all performance metrics. Moreover, our proposed LightGBM model with a genetic algorithm exhibited competitive performance in predicting stock price movements.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.43 no.3
• /
• pp.321-335
• /
• 2023

Monitoring sediment loads in natural rivers is the key process in river engineering, but it is costly and dangerous. In practice, suspended loads are directly measured, and total loads, which is a summation of suspended loads and bed loads, are estimated. This study proposes a real-time sediment discharge monitoring system using the horizontal acoustic Doppler current profiler (H-ADCP) and support vector regression (SVR). The proposed system is comprised of the SVR model for suspended sediment concentration (SVR-SSC) and for total loads (SVR-Q_TL), respectively. SVR-SSC estimates SSC and SVR-Q_TL mimics the modified Einstein procedure. The grid search with K-fold cross validation (Grid-CV) and the recursive feature elimination (RFE) were employed to determine SVR's hyperparameters and input variables. The two SVR models showed reasonable cross-validation scores (R²) with 0.885 (SVR-SSC) and 0.860 (SVR-Q_TL). During the time-series sediment load monitoring period, we successfully detected various sediment transport phenomena in natural streams, such as hysteresis loops and sensitive sediment fluctuations. The newly proposed sediment monitoring system depends only on the gauged features by H-ADCP without additional assumptions in hydraulic variables (e.g., friction slope and suspended sediment size distribution). This method can be applied to any ADCP-installed discharge monitoring station economically and is expected to enhance temporal resolution in sediment monitoring.