• Journal of the Korea Society of Computer and Information
• /
• v.18 no.11
• /
• pp.31-38
• /
• 2013

Feature normalization as a pre-processing step has been widely used in classification problems to reduce the effect of different scale in each feature dimension and error as a result. Most of the existing methods, however, assume some distribution function on feature distribution. Even worse, existing methods do not use the labels of data points and, as a result, do not guarantee the optimality of the normalization results in classification. In this paper, proposed is a supervised rank normalization which combines rank normalization and a supervised learning technique. The proposed method does not assume any feature distribution like rank normalization and uses class labels of nearest neighbors in classification to reduce error. SVM, in particular, tries to draw a decision boundary in the middle of class overlapping zone, the reduction of data density in that area helps SVM to find a decision boundary reducing generalized error. All the things mentioned above can be verified through experimental results.

• Journal of KIISE:Databases
• /
• v.30 no.3
• /
• pp.336-349
• /
• 2003

Histograms have been getting a lot of attention recently. Histograms are commonly utilized in commercial database systems to capture attribute value distributions for query optimization Recently, in the advent of researches on approximate query answering and stream data, the interests in histograms are widely being spread. The simplest approach assumes that the attributes in relational tables are independent by AVI(Attribute Value Independence) assumption. However, this assumption is not generally valid for real-life datasets. To alleviate the problem of approximation on multi-dimensional data with multiple one-dimensional histograms, several techniques such as wavelet, random sampling and multi-dimensional histograms are proposed. Among them, GENHIST is a multi-dimensional histogram that is designed to approximate the data distribution with real attributes. It uses overlapping buckets that allow more efficient approximation on the data distribution. In this paper, we propose a scheme, OPT that can determine the optimal frequencies of overlapped buckets that minimize the SSE(Sum Squared Error). A histogram with overlapping buckets is first generated by GENHIST and OPT can improve the histogram by calculating the optimal frequency for each bucket. Our experimental result confirms that our technique can improve the accuracy of histograms generated by GENHIST significantly.

• Journal of Korean Ophthalmic Optics Society
• /
• v.20 no.4
• /
• pp.501-509
• /
• 2015

Purpose: To evaluate the reliability of refractive power by comparing the marked refractive power in an automatic phoropter and actually measured spherical/cylindrical refractive power. Methods: Actual refractive power of minus spherical lens and cylindrical lens in an automatic phoropter was measured by a manual lensmeter and compared with the accuracy of marked refractive power. Furthermore, combined refractive power and spherical equivalent refractive power of two overlapped lenses were compared and evaluated with the refractive power of trial lens. Results: An error of 0.125 D and more against the marked degree was observed in 70.6% of spherical refractive power of spherical lens which is built in phoropter, and the higher error was shown with increasing refractive power. Single cylindrical refractive power of cylindrical lens is almost equivalent to the marked degree. Combined spherical refractive power was equivalent to spherical refractive power of single lens when spherical lens and cylindrical lens were overlapped in a phoropter. Thus, there was no change in spherical refractive power by lens overlapping. However, there was a great difference, which suggest the effect induced by overlapping between cylindrical refractive power and the marked degree when spherical lens and cylindrical lens were overlapped. Spherical equivalent refractive power measured by using a phoropter was lower than that estimated by trial glasses frame and marked degree. The difference was bigger with higher refractive power. Conclusions: When assessment of visual acuity is made by using an automatic phoropter for high myopes or myopic astigmatism, some difference against the marked degree may be produced and they may be overcorrected which suggests that improvement is required.

• Journal of Wetlands Research
• /
• v.16 no.3
• /
• pp.383-392
• /
• 2014

The purpose of this study is to simulate inundation and evaluate the applicability of LISFLOOD model to the streams in South Korea by comparing with the inundation map using FLUMEN. The suggested levee breaching scenarios were applied to the LISFLOOD model, and the results obtained from scenarios were evaluated. The modeling results using LISFLOOD by appling the levee breaching scenarios showed 0.2% ~ 42% relative error with FLUMEN model in inundation area. But the relative error of maximum inundation area by overlapping all the flood analysis results from levee breaching scenarios such as the way making flood risk map was approximately 1.2% between two models. Meanwhile, LISFLOOD model was easy to construct input data, DEM as topographic data and discharge hydrograph as upper boundary conditions. And computing time of LISFLOOD was shorter than FLUMEN. Therefore LISFLOOD model can be applied usefully in the region that needs immediate inundation modeling.

• The Journal of Korean Academy of Prosthodontics
• /
• v.37 no.2
• /
• pp.167-184
• /
• 1999

The purpose of this study was to evaluate and compare the at of denture bases processed by injection pressing technic using laser scanner of reverse engineering technic. The auther duplicated 20 maxillary edentulous models and 20 mandibular edentulous models, which were scanned on HYSCAN 45C 3D BCANNER(Hymarc Co., Canada). The scanned data were stored in the personal computer using SURFACER (Imageware Co. U.S.A.) software program. After 40 dentures were cured by PERform Inkovac system, SR-Ivocap system, Palajet system, and Sulfon system, they were stored in water at room temperature fir 24 hours. The dentures were scanned on HYSCAN 45C 3D SCANNER(Hymarc Co., Canada). The scanned data were stored in the personal computer using SURFACER (Imageware Co., U.S.A.) software program. By overlapping two images using the same program, the fit between two surfaces was scaled by positive and negative errors. The obtained results were as follows 1. In the upper denture, most of the positive errors occurred on the lingual side of anterior alveolar ridge and the negative errors were on the flange of denture bases. 2. In the lower denture, most of the positive errors occurred on the inner side of lingual flange and the negative errors were on the border of anterior labial flange areas, 3. There were no statistical differences among the positive errors of the four types of injection denture curing methods and also no statistical differences between negative errors except only in negative maximum errors. 4. In PERform system and SR-Ivocap system, they have the tendency of inaccurate at of lower denture bases comparing to that of upper denture bases. 5. The negative error scales were greater than the positive error scales in all types of injection denture curing methods.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.13 no.12
• /
• pp.2677-2684
• /
• 2009

In this paper, we study a watermarking method based on the informed coding and embedding by means of trellis code and entropy masking. An image is divided as $8{\times}8$ block with no overlapping and the discrete cosine transform(DCT) is applied to each block. Then the 16 medium-frequency AC terms of each block are extracted. Next it is compared with gaussian random vectors having zero mean and unit variance. As these processing, the embedding vectors with minimum value of linear combination between linear correlation and Watson distance can be obtained by Viterbi algorithm at each stage of trellis coding. For considering the image characteristics, we apply different weight value between the linear correlation and the Watson distance using the entropy masking. To evaluate the performance of proposed method, the average bit error rate of watermark message is calculated from different several images. By the experiments the proposed method is improved in terms of the average bit error rate.

• The Journal of Korea Robotics Society
• /
• v.16 no.4
• /
• pp.336-344
• /
• 2021

This paper presents a robust deep learning-based human tracking framework in crowded environments. For practical human tracking applications, a target must be robustly tracked even in undetected or overcrowded situations. The proposed framework consists of two parts: robust deep learning-based human detection and tracking while recognizing the aforementioned situations. In the former part, target candidates are detected using Detectron2, which is one of the powerful deep learning tools, and their weights are computed and assigned. Subsequently, a candidate with the highest weight is extracted and is utilized to track the target human using a Kalman filter. If the bounding boxes of the extracted candidate and another candidate are overlapped, it is regarded as a crowded situation. In this situation, the center information of the extracted candidate is compensated using the state estimated prior to the crowded situation. When candidates are not detected from Detectron2, it means that the target is completely occluded and the next state of the target is estimated using the Kalman prediction step only. In two experiments, people wearing the same color clothes and having a similar height roam around the given place by overlapping one another. The average error of the proposed framework was measured and compared with one of the conventional approaches. In the error result, the proposed framework showed its robustness in the crowded environments.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.36 no.5
• /
• pp.355-369
• /
• 2018

LiDAR (Light Detection And Ranging) strip adjustment is process to improve geo-referencing of the ALS (Airborne Laser Scanner) strips that leads to seamless LiDAR data. Multiple strips are required to collect data over the large areas, thus the strips are overlapped in order to ensure data continuity. The LSA (LiDAR Strip Adjustment) consists of identifying corresponding features and minimizing discrepancies in the overlapping strips. The corresponding features are utilized as control features to estimate transformation parameters. This paper applied SURF (Speeded Up Robust Feature) to identify corresponding features. To improve determination of the corresponding feature, false matching points were removed by applying three schemes: (1) minimizing distance of the SURF feature vectors, (2) selecting reliable matching feature with high cross-correlation, and (3) reflecting geometric characteristics of the matching pattern. In the strip adjustment procedure, corresponding points having large residuals were removed iteratively that could achieve improvement of accuracy of the LSA eventually. Only a few iterations were required to reach reasonably high accuracy. The experiments with simulated and real data show that the proposed method is practical and effective to airborne LSA. At least 80 % accuracy improvement was achieved in terms of RMSE (Root Mean Square Error) after applying the proposed schemes.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.14 no.11
• /
• pp.2548-2554
• /
• 2010

This paper describes a design of fast Fourier transform(FFT) processor for EEG(electroencephalogram) signal analysis for health care services. Hamming window function with 1/2 overlapping is adopted to perform short-time FFT(ST-FFT) of a long period EEG signal occurred in real-time. In order to analyze efficiently EEG signals which have frequency characteristics in the range of 0 Hz to 100 Hz, a 256-point FFT processor is designed, which is based on a single-memory bank architecture and the radix-4 algorithm. The designed FFT processor has been verified by FPGA implementation, and has high accuracy with arithmetic error less than 2%.

• Journal of information and communication convergence engineering
• /
• v.12 no.3
• /
• pp.173-180
• /
• 2014

This paper investigates the robustness of watermarking techniques for stereo or multi-view images generated from texture and depth images. A three-dimensional (3D) warping technique is applied to texture and depth images to generate stereo or multi-view images for a 3D display. By using the 3D warping technique, in this paper, we developed watermarking techniques and evaluated the robustness of these techniques that can extract watermarks from texture images even when the viewpoints are moved. A depth image is used to generate a stereo image with the largest viewpoint difference to the left and right. The overlapping region in the stereo image that does not disappear after warping is then obtained, and DWT is applied to this region to embed a watermark in the LL sub-band. The proposed watermarking techniques were found to yield bit error rates of about 3%-16% when they were applied to stereo images generated from texture and depth images. Furthermore, the results showed that the copyright could be seen when the extracted watermark was visually confirmed.