• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.41 no.4
• /
• pp.45-50
• /
• 2004

In this paper, we propose a new PCA-based LDA Mixture Algorithm(PLMA) for real-time face recognition system. This system greatly consists of the two parts: 1) face extraction part; 2) face recognition part. In the face extraction part we applied subtraction image, color filtering, eyes and mouth region detection, and normalization method, and in the face recognition part we used the method mixing PCA and LDA in extracted face candidate region images. The existing recognition system using only PCA showed low recognition rates, and it is hard in the recognition system using only LDA to apply LDA to the input images as it is when the number of image pixels ire small as compared with the training set. To overcome these shortcomings, we reduced dimension as we apply PCA to the normalized images, and apply LDA to the compressed images, therefore it is possible for us to do real-time recognition, and we are also capable of improving recognition rates. We have experimented using self-organized DAUface database to evaluate the performance of the proposed system. The experimental results show that the proposed method outperform PCA, LDA and ICA method within the framework of recognition accuracy.

• The Transactions of the Korea Information Processing Society
• /
• v.13 no.4
• /
• pp.189-198
• /
• 2024

Lipreading is one of the important parts of speech recognition, and several studies have been conducted to improve the performance of lipreading in lipreading systems for speech recognition. Recent studies have used method to modify the model architecture of lipreading system to improve recognition performance. Unlike previous research that improve recognition performance by modifying model architecture, we aim to improve recognition performance without any change in model architecture. In order to improve the recognition performance without modifying the model architecture, we refer to the cues used in human lipreading and set other regions such as chin and cheeks as regions of interest along with the lip region, which is the existing region of interest of lipreading systems, and compare the recognition rate of each region of interest to propose the highest performing region of interest In addition, assuming that the difference in normalization results caused by the difference in interpolation method during the process of normalizing the size of the region of interest affects the recognition performance, we interpolate the same region of interest using nearest neighbor interpolation, bilinear interpolation, and bicubic interpolation, and compare the recognition rate of each interpolation method to propose the best performing interpolation method. Each region of interest was detected by training an object detection neural network, and dynamic time warping templates were generated by normalizing each region of interest, extracting and combining features, and mapping the dimensionality reduction of the combined features into a low-dimensional space. The recognition rate was evaluated by comparing the distance between the generated dynamic time warping templates and the data mapped to the low-dimensional space. In the comparison of regions of interest, the result of the region of interest containing only the lip region showed an average recognition rate of 97.36%, which is 3.44% higher than the average recognition rate of 93.92% in the previous study, and in the comparison of interpolation methods, the bilinear interpolation method performed 97.36%, which is 14.65% higher than the nearest neighbor interpolation method and 5.55% higher than the bicubic interpolation method. The code used in this study can be found a https://github.com/haraisi2/Lipreading-Systems.

• Korean Journal of Food Science and Technology
• /
• v.32 no.2
• /
• pp.265-270
• /
• 2000

White ginseng powder, permitted to be irradiated in Korea for the purpose of microbial decontamination, was treated with electron beam at doses of $0{\sim}15\;kGy$ for a detection trial whether it is irradiated or not by measuring photostimulated luminescence for whole samples first and then (TL) for the mineral adhering to the samples. PSL values were less than threshold value (700, $T_{1}$) and were negative for nonirradiated samples but more than 5000 $(T_2)$ and were positive for irradiated ones. After PSL measurement mineral was separated from the whole samples using density separation. Mineral of nonirradiated samples was characterized by glow curves which have low intensity and were situated at the high temperature region (about $300^{\circ}C$) by the low level of natural radioactivity. Glow curves of minerals for all irradiated samples were observed at about $200^{\circ}C$. TL ratio by normalization was 0.01 for nonirradiated sample and more than 0.78 for irradiated samples, and it was possible to detect whether white ginseng powders were irradiated or not.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2003.05a
• /
• pp.165-170
• /
• 2003

This paper presents an integrated approach to unconstrained face recognition in arbitrary scenes. The front end of the system comprises of a scale and pose tolerant face detector. Scale normalization is achieved through novel combination of a skin color segmentation and log-polar mapping procedure. Principal component analysis is used with the multi-view approach proposed in[10] to handle the pose variations. For a given color input image, the detector encloses a face in a complex scene within a circular boundary and indicates the position of the nose. Next, for recognition, a radial grid mapping centered on the nose yields a feature vector within the circular boundary. As the width of the color segmented region provides an estimated size for the face, the extracted feature vector is scale normalized by the estimated size. The feature vector is input to a trained neural network classifier for face identification. The system was evaluated using a database of 20 person's faces with varying scale and pose obtained on different complex backgrounds. The performance of the face recognizer was also quite good except for sensitivity to small scale face images. The integrated system achieved average recognition rates of 87% to 92%.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.7 no.6
• /
• pp.1312-1317
• /
• 2003

This paper presents an integrated approach to unconstrained face recognition in arbitrary scenes. The front end of the system comprises of a scale and pose tolerant face detector. Scale normalization is achieved through novel combination of a skin color segmentation and log-polar mapping procedure. Principal component analysis is used with the multi-view approach proposed in[10] to handle the pose variations. For a given color input image, the detector encloses a face in a complex scene within a circular boundary and indicates the position of the nose. Next, for recognition, a radial grid mapping centered on the nose yields a feature vector within the circular boundary. As the width of the color segmented region provides an estimated size for the face, the extracted feature vector is scale normalized by the estimated size. The feature vector is input to a trained neural network classifier for face identification. The system was evaluated using a database of 20 person's faces with varying scale and pose obtained on different complex backgrounds. The performance of the face recognizer was also quite good except for sensitivity to small scale face images. The integrated system achieved average recognition rates of 87% to 92%.

• Molecules and Cells
• /
• v.40 no.6
• /
• pp.410-417
• /
• 2017

Estimation of postmortem interval (PMI) is a key issue in the field of forensic pathology. With the availability of quantitative analysis of RNA levels in postmortem tissues, several studies have assessed the postmortem degradation of constitutively expressed RNA species to estimate PMI. However, conventional RNA quantification as well as biochemical and physiological changes employed thus far have limitations related to standardization or normalization. The present study focuses on an interesting feature of the subdomains of certain RNA species, in which they are site-specifically cleaved during apoptotic cell death. We found that the D8 divergent domain of ribosomal RNA (rRNA) bearing cell death-related cleavage sites was rapidly removed during postmortem RNA degradation. In contrast to the fragile domain, the 5' terminal region of 28S rRNA was remarkably stable during the postmortem period. Importantly, the differences in the degradation rates between the two domains in mammalian 28S rRNA were highly proportional to increasing PMI with a significant linear correlation observed in mice as well as human autopsy tissues. In conclusion, we demonstrate that comparison of the degradation rates between domains of a single RNA species provides quantitative information on postmortem degradation states, which can be applied for the estimation of PMI.

• New & Renewable Energy
• /
• v.8 no.2
• /
• pp.14-23
• /
• 2012

The TMY (Typical Meteorological Year) for the solar energy study is generated using observation data with 22 solar sites from KMA (Korea Meteorological Administration) during 11 years (2000-2010). The meteorological data for calculation the TMY are used solar radiation, temperature, dew point temperature, wind speed and humidity data. And the TMY is calculated to apply the FS (Finkelstein and Schafer) statistics and RMSE (Root Mean Squared Error) methods. FS statistics performed with each point and each variable and then selected top five candidate TMM months with statistical analysis and normalization. Finally TMY is generated to select the highest TMM score with evaluation the average errors for the 22 whole points. The TMY data is represented average state and long time variations with 22 sites and meteorological data. When TMY validated with the 11-year daily solar radiation data, the correlation coefficient was about 0.40 and the highest value is 0.57 in April and the lowest value is 0.23 in May. Mean monthly solar radiation of TMY is 411.72 MJ which is 4 MJ higher than original data. Average correlation coefficient is 0.71, the lowest correlation is 0.43 in May and the highest correlation is 0.90 in January. Accumulated annual solar radiation by TMY have higher value in south coast and southwestern region and have relatively low in middle regions. And also, differences between TMY and 11-year mean of is distributed lower 100 MJ in Kyeongbuk, higher 200 MJ in Jeju and higher 125 MJ in Jeonbuk and Jeonnam, respectively.

• Clinical and Experimental Pediatrics
• /
• v.61 no.6
• /
• pp.194-199
• /
• 2018

Purpose: Sacral dimples are a common cutaneous anomaly in infants. Spine ultrasonography (USG) is an effective and safe screening tool for patients with a sacral dimple. The aim of this study was to determine the clinical manifestations in patients with an isolated sacral dimple and to review the management of spinal cord abnormalities identified with USG. Methods: We reviewed clinical records and collected data on admissions for a sacral dimple from March 2014 through February 2017 that were evaluated with spine USG by a pediatric radiologist. During the same period, patients who were admitted for other complaints, but were found to have a sacral dimple were also included. Results: This study included 230 infants under 6-months-old (130 males and 100 females; mean age $52.8{\pm}42.6days$). Thirty-one infants with a sacral dimple had an echogenic filum terminale, and 57 children had a filar cyst. Twenty-seven patients had a low-lying spinal cord, and only one patient was suspected of having a tethered cord. Follow-up spine USG was performed in 28 patients, which showed normalization or insignificant change. Conclusion: In this study, all but one infant with a sacral dimple had benign imaging findings. USG can be recommended in infants with a sacral dimple for its convenience and safety.

• Proceedings of the Plant Resources Society of Korea Conference
• /
• 2000.10a
• /
• pp.116-122
• /
• 2000

The molecular modification of antiepileptic drugs and direct synthesis of new drugs with the predetermined antiepileptic properties are perspective. New neurochemical attacking to solve the problem including prevention and inhibition of seizures seems to be related to ginsenosides and ginseng polypeptides. The main study based on the severity of febrile convulsions of rat pups has been done from the earlier investigations of antiepileptical action of ginsenosides between KGTRI and MSU (Chepurnov, Park et al., 1995) with different kinds of experimental models of epilepsy. From the cultured cell line DAN25 of ginseng root, the extracts of ginsenosides made in "BIOKHIMMASH" were studied by the project of preclinical anticonvulsant screening (Stables, Kupferberg, 1997). The inhibition of severity of convulsions, decrease of seizures threshold, decrease of audiogenic seizures in rats of different strains and normalization of cerebral blood flow (measured by hydrogen test) were demonstrated in rats after i.c.v., intraperitoneally and orally administration, respectively. The antiepileptical effects by the combination of compounds from ginseng; were compared with the iuluence of Rg1, Rb1, Rc and with the well known antiepileptical drugs such as carbamazepine, valproic acid. The base for the research is obtained by using the WAG/Rij strain (Luijtelaar, Coenen, Kuznetcova), an excellent genetic model for human generalized absence epilepsy. The improving action of gensinosides was effectively demonstrated on the model of electrical kindling of amygdala of WAG/Rij rats with genetically determined absences, and the influences of ginsenosides on the slow wave discharges have also been being investigated. The different characteristics of a kindling process exerted in the sex-different region of the amygdala and demonstrated that the level of sex steroids and content of neurosteroids in amygdaloid tissue can modify the development of seizures. The chemical structures of ginsenosides not only have some principal differences from well-known antiepileptical drugs but the Plant Pharmacology gives us unique possibility to develop new class of antiepileptic drugs and to improve its biological activity.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.46 no.6
• /
• pp.26-35
• /
• 2009

In this paper, we present morphology-based step region extraction and regularized iterative point-spread-function (PSF) estimation methods. The proposed PSF estimation method uses canny edge detector to extract the edge of the input image. We extract feasible vertical and horizontal edges using morphology analysis, such as the hit-or-miss transform. Given extracted edges we estimate the optimal step-response using flattening and normalization processes. The PSF is finally characterized by solving the equation which relates the optimal step response and the 2D isotropic PSF. We shows the restored image by the estimated PSF. The proposed algorithm can be applied a fully digital auto-focusing system without using mechanical focusing parts.