• The Journal of the Korea Contents Association
• /
• v.12 no.12
• /
• pp.517-523
• /
• 2012

Biometrics is a study of identifying individual by using the features of human body. It has been studied for an alternative or complementary method for the classical method based on password, ID card, etc. In comparison with the fingerprint, iris, ear, palmprint, finger-knuckle print has been recently studied. This paper proposes an effective method for recognizing finger-knuckle print based on the feature of Gradient orientation. The main features of finger-knuckle print are the size and direction of winkles. In order to extract these features stably, we make a feature vector consisted of Gradient orientations after the preprocessing of enhancing non-uniform brightness and low contrast. Total 790 images acquired from 158 persons have been used at the experiment for evaluating the performance of the proposed method. The experimental results show the recognition rate of 99.69% and the relatively high decidability index of 1.882. These results demonstrate that the proposed method is effective in recognizing finger-knuckle print.

• The Bulletin of The Korean Astronomical Society
• /
• v.39 no.1
• /
• pp.56.2-56.2
• /
• 2014

We present [Fe II] 1.64 ${\mu}m$ imaging observations for jets and outflows from young stellar objects over the northern part (${\sim}24^{\prime}{\times}45^{\prime}$) of the Carina Nebula, a massive star forming region. The observations were performed with IRIS2 of Anglo-Australian Telescope and the seeing was ~1.5". Eleven jets and outflows features are detected at eight different regions, and are termed as Ionized Fe Objects (IFOs). The [Fe II] features have knotty or elongated shapes, and the detection rate of IFOs against previously identified YSOs is 1.4%. Four IFOs show anti-correlated peak intensities in [Fe II] and $H{\alpha}$, where the ratio I([Fe II])/I($H{\alpha}$) is higher for longish IFOs than for knotty IFOs. We estimate the outflow mass loss rate from the [Fe II] flux using two different methods. The jet-driving objects are identified for three IFOs (IFO-2, -4, and -7). The ratios of the outflow mass loss rate over the disk accretion rate for IFO-4 and -7 are consistent with the previously reported values ($10^{-2}-10^{+1}$), while the ratio is higher for IFO-2. This excess may result from underestimating the disk accretion rate. Other YSO physical parameters show reasonable relations or trends.

• Journal of Acupuncture Research
• /
• v.26 no.2
• /
• pp.41-48
• /
• 2009

Objectives : The conventional direct and indirect moxibustion therapies for prostate treatment could not been applied to the acupuncture point of $CV_1$(Conception Vessel Meridian 1, 會陰) because of its boring body region. The position of $CV_1$(會陰) is the back side of hard part between the anus and the genital organ. The conventional moxibustion methods have many troubles in operating to the acupuncture point of $CV_1$(會陰). In order to get rid of these problems, we have suggested the special heat generating terminal especially for prostate. The features of the special heat generating terminal for prostate are the low temperature infrared heater and the adhesive moxa-pad. These features are no burnt, no fiery and especially suitable for the point of $CV_1$(會陰). Methods: The heat generating terminal which is a part of the moxa-extract moxibustion cauterizer is composed of a PTC(Positive Temperature Coefficients) ceramic heater and the adhesive moxa-pad We had got the experimental demonstrations by the stimulating the acupuncture points which are $CV_1$(會陰), $BL_{28}$(Bladder Meridian 28, 膀胱兪), and $CV_3$(Conception Vessel Meridian 3, 中極) with the special heat generating terminal for the prostatitis and the benign prosthetic hypertrophy. And the stimulation level was 43$^{\circ}C$ infrared heat for one hour. The type of thermography is IRIS-5000. Results : With one subject suffering the prostatitis and another subject suffering the benign prosthetic hypertrophy, we cauterized the acupuncture points $CV_1$(會陰), $BL_{28}$(膀胱兪) and $CV_3$(中極) with the special heat generating terminal for prostate. We measured the temperature variations by the thermography before and after stimulations. Finally we estimated the tendency of temperature decreasing in the region of post-stroke urinary symptoms and the improvement of nocturnal enuresis after the stimulations. Conclusions : We suggest that the special heat generating terminal of moxa-extract moxibustion cauterizer proposed herein is effective for the treatment of prostate by NIH-CPSI and IPSS.

• Korean Journal of Veterinary Research
• /
• v.35 no.2
• /
• pp.375-381
• /
• 1995

Two intraocular tumors were identified in low and medium dosed groups of a carcinogenicity study using 200 males and 200 females Sprague-Dawley rats in Screening & Toxicology Research Center, Korea Research Institute of Chemical Technology. The tumors were grossly observed as white or yellow, unilateral nodules. They were approximately $1-2{\times}3-5mm$ in size. The tumors located in the region of iris and/or ciliary body invaded peripheral cornea. The microscopic features were usually composed of spindle cells arranged in parallel, forming gently curving bundles or whorls. The spindle cells had poorly defined cell boundaries, scant to moderate cytoplasm, fusiform nuclei and indistinct nucleoli. Mitotic figures were rare and areas of necrosis were present. The spindle cells had positive immunoreactivity for S-100 protein and vimentin but negative for desmin, collagen and HMB-45 antibody. In special histochemical studies, the spindle cells react with Gomori's stain for argyrophih fibers, Prussian blue stain for iron but negative with Masson-Fontana's stain for melanin granules. Ultrastructurally, cytoplasmic premelanosomes were not observed in the tumor cells due to the poor preservation of tumor masses. Based on the results, the tumors were diagnosed as amelanotic melanoma.

• Journal of Information Processing Systems
• /
• v.14 no.5
• /
• pp.1114-1135
• /
• 2018

A multiple classification system based on a new boosting technique has been approached utilizing different biometric traits, that is, color face, iris and eye along with fingerprints of right and left hands, handwriting, palm-print, gait (silhouettes) and wrist-vein for person authentication. The images of different biometric traits were taken from different standard databases such as FEI, UTIRIS, CASIA, IAM and CIE. This system is comprised of three different super-classifiers to individually perform person identification. The individual classifiers corresponding to each super-classifier in their turn identify different biometric features and their conclusions are integrated together in their respective super-classifiers. The decisions from individual super-classifiers are integrated together through a mega-super-classifier to perform the final conclusion using programming based boosting. The mega-super-classifier system using different super-classifiers in a compact form is more reliable than single classifier or even single super-classifier system. The system has been evaluated with accuracy, precision, recall and F-score metrics through holdout method and confusion matrix for each of the single classifiers, super-classifiers and finally the mega-super-classifier. The different performance evaluations are appreciable. Also the learning and the recognition time is fairly reasonable. Thereby making the system is efficient and effective.

• Korean Journal of Artificial Intelligence
• /
• v.11 no.4
• /
• pp.29-34
• /
• 2023

The flow from developing a machine learning model to deploying it in a production environment suffers challenges. Efficient and reliable deployment is critical for realizing the true value of machine learning models. Bridging this gap between development and publication has become a pivotal concern in the machine learning community. FastAPI, a modern and fast web framework for building APIs with Python, has gained substantial popularity for its speed, ease of use, and asynchronous capabilities. This paper focused on leveraging FastAPI for deploying machine learning models, addressing the potentials associated with integration, scalability, and performance in a production setting. In this work, we explored the seamless integration of machine learning models into FastAPI applications, enabling real-time predictions and showing a possibility of scaling up for a more diverse range of use cases. We discussed the intricacies of integrating popular machine learning frameworks with FastAPI, ensuring smooth interactions between data processing, model inference, and API responses. This study focused on elucidating the integration of machine learning models into production environments using FastAPI, exploring its capabilities, features, and best practices. We delved into the potential of FastAPI in providing a robust and efficient solution for deploying machine learning systems, handling real-time predictions, managing input/output data, and ensuring optimal performance and reliability.

• Economic and Environmental Geology
• /
• v.31 no.1
• /
• pp.21-29
• /
• 1998

Broadband receiver functions are developed from teleseismic P waveforms recorded at Wonju (KSRS), Inchon (IRIS), and Pohang (PHN), and are analyzed to examine the crustal structure beneath the three stations. The teleseismic receiver functions are inverted in the time domain to the vertical P wave velocity structure beneath the stations. Clear P-to-S converted phases from the Moho interface are observed in teleseismic seismograms recorded at the three stations. We estimated the crustal velocity structures beneath the stations using the receiver function inversion. The general features of inversion results are as follows: (1) For Pohang station, there is a high velocity gradient at a 4~5 km deep for SE and NW back azimuth and a low velocity zone at around 10 km deep. The Moho depth is 28 km for NW direction. (2) The shallow crustal structure beneath Wonju station is somewhat complex and there is a high-velocity zone ($V_p{\simeq}6.8km/sec$) at 3 to 4 km deep. The average crustal thickness is 33 km, and a transition zone exists at a 30~33 km deep of lower crust, of which velocity is abruptly changed 6.4 to 7.9 km/sec. (3) For Inchon station, the crustal velocity gradient monotonously increases up to the Moho discontinuity and the velocity is abruptly changed from 6.2 km/sec to 7.9 km/sec at 29 km deep.

• Journal of the Korea Convergence Society
• /
• v.9 no.8
• /
• pp.1-8
• /
• 2018

In the field of biometric security such as face and iris recognition, it is essential to extract facial features such as eyes and lips. In this paper, we have studied a method of detecting eye and lip region in face image using faster R-CNN. The faster R-CNN is an object detection method using deep running and is well known to have superior performance compared to the conventional feature-based method. In this paper, feature maps are extracted by applying convolution, linear rectification process, and max pooling process to facial images in order. The RPN(region proposal network) is learned using the feature map to detect the region proposal. Then, eye and lip detector are learned by using the region proposal and feature map. In order to examine the performance of the proposed method, we experimented with 800 face images of Korean men and women. We used 480 images for the learning phase and 320 images for the test one. Computer simulation showed that the average precision of eye and lip region detection for 50 epoch cases is 97.7% and 91.0%, respectively.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.12 no.5
• /
• pp.2339-2345
• /
• 2011

Gait recognition is the method to identify the person who walks in front of camera using characteristics of individuals by a sequence of images of walking people. The accuracy of biometric such as fingerprint or iris is very high; however, to provide information needs downsides which allow users to direct contact or close-up, etc. There have been many studies in gait recognition because it could capture images and analysis characteristics far from a person. In order to recognize the gait of person needs a continuous sequence of walking which can be distinguished from the individuals should be extracted features rather than an single image. Therefore, this paper proposes a method of gait recognition that the motion of objects in sequence is described the characteristics of a shape sequence descriptor, and through a variety of experiments can show possibility as a recognition technique.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.22 no.5
• /
• pp.734-739
• /
• 2018

Fingerprint recognition and iris recognition, which are one of the biometric methods, are easily influenced by external factors such as sunlight. Recently, finger vein recognition is used as a method utilizing internal features. However, for accurate finger vein recognition, it is important to clearly separate vein and background regions. However, it is difficult to separate the vein region and background region due to the abnormalized illumination, and a method of separating the vein region and the background region after normalized the illumination of the input image has been proposed. In this paper, we proposed a method to enhance the quality improvement and improve the processing time compared to the existing finger vein recognition system binarization and labeling method of the image including the image stretching process based on the existing illumination normalization method.