• Journal of Intelligence and Information Systems
• /
• v.26 no.2
• /
• pp.1-25
• /
• 2020

In this paper, we suggest an application system architecture which provides accurate, fast and efficient automatic gasometer reading function. The system captures gasometer image using mobile device camera, transmits the image to a cloud server on top of private LTE network, and analyzes the image to extract character information of device ID and gas usage amount by selective optical character recognition based on deep learning technology. In general, there are many types of character in an image and optical character recognition technology extracts all character information in an image. But some applications need to ignore non-of-interest types of character and only have to focus on some specific types of characters. For an example of the application, automatic gasometer reading system only need to extract device ID and gas usage amount character information from gasometer images to send bill to users. Non-of-interest character strings, such as device type, manufacturer, manufacturing date, specification and etc., are not valuable information to the application. Thus, the application have to analyze point of interest region and specific types of characters to extract valuable information only. We adopted CNN (Convolutional Neural Network) based object detection and CRNN (Convolutional Recurrent Neural Network) technology for selective optical character recognition which only analyze point of interest region for selective character information extraction. We build up 3 neural networks for the application system. The first is a convolutional neural network which detects point of interest region of gas usage amount and device ID information character strings, the second is another convolutional neural network which transforms spatial information of point of interest region to spatial sequential feature vectors, and the third is bi-directional long short term memory network which converts spatial sequential information to character strings using time-series analysis mapping from feature vectors to character strings. In this research, point of interest character strings are device ID and gas usage amount. Device ID consists of 12 arabic character strings and gas usage amount consists of 4 ~ 5 arabic character strings. All system components are implemented in Amazon Web Service Cloud with Intel Zeon E5-2686 v4 CPU and NVidia TESLA V100 GPU. The system architecture adopts master-lave processing structure for efficient and fast parallel processing coping with about 700,000 requests per day. Mobile device captures gasometer image and transmits to master process in AWS cloud. Master process runs on Intel Zeon CPU and pushes reading request from mobile device to an input queue with FIFO (First In First Out) structure. Slave process consists of 3 types of deep neural networks which conduct character recognition process and runs on NVidia GPU module. Slave process is always polling the input queue to get recognition request. If there are some requests from master process in the input queue, slave process converts the image in the input queue to device ID character string, gas usage amount character string and position information of the strings, returns the information to output queue, and switch to idle mode to poll the input queue. Master process gets final information form the output queue and delivers the information to the mobile device. We used total 27,120 gasometer images for training, validation and testing of 3 types of deep neural network. 22,985 images were used for training and validation, 4,135 images were used for testing. We randomly splitted 22,985 images with 8:2 ratio for training and validation respectively for each training epoch. 4,135 test image were categorized into 5 types (Normal, noise, reflex, scale and slant). Normal data is clean image data, noise means image with noise signal, relfex means image with light reflection in gasometer region, scale means images with small object size due to long-distance capturing and slant means images which is not horizontally flat. Final character string recognition accuracies for device ID and gas usage amount of normal data are 0.960 and 0.864 respectively.

• Geophysics and Geophysical Exploration
• /
• v.6 no.3
• /
• pp.126-133
• /
• 2003

High-resolution shallow marine seismic surveys have been carried out for the resources exploration, engineering applications and Quaternary mapping. To improve the resolution of subsurface structure image, multichannel digital technique has been applied. The quality of the image depends on the vertical and horizontal resolution and signal to noise (S/N) ratio which are associated with the data acquisition parameters such as sample interval, common midpoint (CMP) interval and CMP fold. To understand the effect of the acquisition parameters, a test survey was carried out off Yeosu and the acquired data were analyzed. A 30 $in^3$ small air gun was used as a seismic source and 8 channel streamer cable with a 5 m group interval was used as a receiver. The data were digitally recorded with a shot interval of 2 s and sample interval of 0.1 ms. The acquired data were resampled with various sample intervals, CMP intervals and CMP folds. The resampled data were processed, plotted as seismic sections and compared each other. The analysis results show that thin bed structure with ${\~}1m$ thickness and ${\~}6^{\circ}$ slope can be imaged with good resolution and continuity and low noise using the acquisition parameters with a sample interval shorter than 0.2 ms, CMP interval shorter than 2.5 m and CMP fold more than 4. Because seismic resolution is associated with the acquisition parameters, the quality of the subsurface structure can be imaged successfully using suitable and optimum acquisition parameters.

• The Journal of Korean Society for Radiation Therapy
• /
• v.18 no.1
• /
• pp.1-5
• /
• 2006

Purpose: IMRT quality assurance(Q.A) is consist of the absolute dosimetry using ionization chamber and relative dosimetry using the film. We have in general used 0.015 cc ionization chamber, because small size and measure the point dose. But this ionization chamber is too small to give an accurate measurement value. In this study, we have examined the degree of calculated to measured dose difference in intensity modulated radiotherapy(IMRT) based on the observed/expected ratio using various kinds of ion chambers, which were used for absolute dosimetry. Materials and Methods: we peformed the 6 cases of IMRT sliding-window method for head and neck cases. Radiation was delivered by using a Clinac 21EX unit(Varian, USA) generating a 6 MV x-ray beam, which is equipped with an integrated multileaf collimator. The dose rate for IMRT treatment is set to 300 MU/min. The ion chamber was located 5cm below the surface of phantom giving 100cm as a source-axis distance(SAD). The various types of ion chambers were used including 0.015cc(pin point type 31014, PTW. Germany), 0.125 cc(micro type 31002, PTW, Germany) and 0.6 cc(famer type 30002, PTW, Germany). The measurement point was carefully chosen to be located at low-gradient area. Results: The experimental results show that the average differences between plan value and measured value are ${\pm}0.91%$ for 0.015 cc pin point chamber, ${\pm}0.52%$ for 0.125 cc micro type chamber and ${\pm}0.76%$ for farmer type 0.6cc chamber. The 0.125 cc micro type chamber is appropriate size for dose measure in IMRT. Conclusion: IMRT Q.A is the important procedure. Based on the various types of ion chamber measurements, we have demonstrated that the dose discrepancy between calculated dose distribution and measured dose distribution for IMRT plans is dependent on the size of ion chambers. The reason is small size ionization chamber have the high signal-to-noise ratio and big size ionization chamber is not located accurate measurement point. Therefore our results suggest the 0.125 cc farmer type chamber is appropriate size for dose measure in IMRT.

• Geophysics and Geophysical Exploration
• /
• v.26 no.2
• /
• pp.52-61
• /
• 2023

The use of ultra-high-resolution (UHR) seismic surveys to preceisly characterize coastal and shallow structures have increased recently. UHR surveys derive a spatial resolution of 3.125 m using a high-frequency source (80 Hz to 1 kHz). A digital streamer system is an essential module for acquiring high-quality UHR seismic data. Localization studies have focused on reducing purchase costs and decreasing maintenance periods. Basic performance verification and application tests of the developed streamer have been successfully carried out; however, a comparative analysis with the existing benchmark model was not conducted. In this study, we characterized data obtained by using a developed streamer and a benchmark model simultaneously. Tamhae 2 and auxiliary equipment of the Korea Institute of Geoscience and Mineral Resources were used to acquire 2D seismic data, which were analyzed from different perspectives. The data obtained using the developed streamer differed in sensitivity from that obtained using benchmark model by frequency band.However, both type of data had a very high level of similarity in the range corresponding to the central frequency band of the seismic source. However, in the low frequency band below 60 Hz, data obtained using the developed streamer showed a lower signal-to-noise ratio than that obtained using the benchmark model.This lower ratio can hinder the quality in data acquisition using low-frequency sound sources such as cluster air guns. Three causes for this difference were, and streamers developed in future will attempt to reflect on these improvements.

• Journal of the Korean Society of Radiology
• /
• v.11 no.1
• /
• pp.27-35
• /
• 2017

The purpose of this study was to investigate the features of CsI:Tl and $Gd_2O_2S$ detectors with an indirect conversion method using phantom in the DR (digital radiography) system by obtaining images of thick chest phantom, medium thickness thigh phantom, and thin hand phantom and by analyzing the SNR and CNR. As a result of measuring the SNR and CNR according to the thickness change of the subject, the SNR and CNR were higher in CsI:Tl detector than in $Gd_2O_2S$ detector when the medium thickness thigh phantom and thin hand phantom were scanned. However, when the thick chest phantom was used, for the SNR at 80~125 kVp and the CNR at 80~110 kVp in the $Gd_2O_2S$ detector, the values were higher than those of CsI:Tl detector. The SNR and CNR both increased as the tube voltage increased. The SNR and CNR of CsI:Tl detector in the medium thickness thigh phantom increased at 40~50 kVp and decreased as the tube voltage increased. The SNR and CNR of $Gd_2O_2S$ detector increased at 40~60 kVp and decreased as the tube voltage increased. The SNR and CNR of CsI:Tl detctor in the thin hand phantom decreased at the low tube voltage and increased as the tube voltage increased, but they decreased again at 100~110 kVp, while the SNR and CNR of $Gd_2O_2S$ detector were found to decrease as the tube voltage increased. The MTF of CsI:Tl detector was 6.02~90.90% higher than that of $Gd_2O_2S$ detector at 0.5~3 lp/mm. The DQE of CsI:Tl detector was 66.67~233.33% higher than that of $Gd_2O_2S$ detector. In conclusion, although the values of CsI:Tl detector were higher than those of $Gd_2O_2S$ detector in the comparison of MTF and DQE, the cheaper $Gd_2O_2S$ detector had higher SNR and CNR than the expensive CsI:Tl detector at a certain tube voltage range in the thick check phantom. At chest X-ray, if the $Gd_2O_2S$ detector is used rather than the CsI:Tl detector, chest images with excellent quality can be obtained, which will be useful for examination. Moreover, price/performance should be considered when determining the detector type from the viewpoint of the user.

• Resources Recycling
• /
• v.30 no.6
• /
• pp.43-52
• /
• 2021

In this study, the optimal nitration process for selective lithium leaching from powder of a spent battery cell (LiNi_xCo_yMn_zO₂, LiCoO₂) was studied using Taguchi method. The nitration process is a method of selective lithium leaching that involves converting non-lithium nitric compounds into oxides via nitric acid leaching and roasting. The influence of pretreatment temperature, nitric acid concentration, amount of nitric acid, and roasting temperature were evaluated. The signal-to-noise ratio and analysis of variance of the results were determined using L₁₆(4⁴) orthogonal arrays. The findings indicated that the roasting temperature followed by the nitric acid concentration, pretreatment temperature, and amount of nitric acid used had the greatest impact on the lithium leaching ratio. Following detailed experiments, the optimal conditions were found to be 10 h of pretreatment at 700℃ with 2 ml/g of 10 M nitric acid leaching followed by 10 h of roasting at 275℃. Under these conditions, the overall recovery of lithium exceeded 80%. X-ray diffraction (XRD) analysis of the leaching residue in deionized water after roasting of lithium nitrate and other nitrate compounds was performed. This was done to determine the cause of rapid decrease in lithium leaching rate above a roasting temperature of 400℃. The results confirmed that lithium manganese oxide was formed from lithium nitrate and manganese nitrate at these temperatures, and that it did not leach in deionized water. XRD analysis was also used to confirm the recovery of pure LiNO₃ from the solution that was leached during the nitration process. This was carried out by evaporating and concentrating the leached solution through solid-liquid separation.

• Investigative Magnetic Resonance Imaging
• /
• v.12 no.2
• /
• pp.115-122
• /
• 2008

Purpose : To evaluate the correlation between the radiological non-invasive hepatic fibrosis index (RNHFI), as determined by SPIO-enhanced MRI, and the laboratory non-invasive hepatic fibrosis index. Materials and Methods : Patients (99 total: 61 men and 38 women; mean age: 58 years) who underwent SPIO-enhanced MRI (1.5T) during 5 years included. These patients were subdivided into a liver cirrhosis group (LCG) and a non-liver cirrhosis group (non-LCG). Using PACS view, we measured the RNHFI (mean standard deviation of hepatic signal intensity (SD), noise-corrected coefficient of variation (CV)) of three ROIs in the liver parenchyma by SPIO-enhanced MRI. The laboratory non-invasive hepatic fibrosis index (AST-platelet ratio index (APRI)) of all patients was calculated from the laboratory data. We compared the RNHFI and APRI of LCG with those of non-LC group using Student's t-test. A bivariate correlation was performed to investigate the relationship between the RNHFI and APRI in the LCG. Results : For the LCG, mean values of SD and CV by SPIO-enhanced MRI were $10.3{\pm}3.7$ and $0.19{\pm}0.08$, respectively. For the non-LCG, mean values of SD and CV were $6.5{\pm}1.6$ and $0.08{\pm}0.05$, respectively. The mean APRI of the LCG and the non- LCG were $2.04{\pm}1.7$ and $0.32{\pm}0.32$, respectively. The RNHFI and APRI were significantly different between both groups (p<0.05). For the LCG, the bivariate correlation between SD and APRI revealed a statistically significant positive correlation (r=0.5, p<0.001). In both groups, there was no statistically significant correlation between CV and APRI. Conclusion: A measurement of SD can be a simple and useful method for the evaluation of hepatic fibrosis.

• Korean Journal of Biological Psychiatry
• /
• v.23 no.4
• /
• pp.148-156
• /
• 2016

Objectives According to previous studies, the cannabinoid receptor 1 (CNR1) gene could be an important candidate gene for schizophrenia. Some studies have linked the (AAT)n trinucleotide repeat polymorphism in CNR1 gene with the risk of schizophrenia. Meanwhile, smooth pursuit eye movement (SPEM) has been regarded as one of the most consistent endophenotypes of schizophrenia. In this study, we investigated the association between the (AAT)n trinucleotide repeats in CNR1 gene and SPEM abnormality in Korean patients with schizophrenia. Methods We measured SPEM function in 167 Korean patients with schizophrenia (84 male, 83 female) and they were divided according to SPEM function into two groups, good and poor SPEM function groups. We also investigated allele frequencies of (AAT)n repeat polymorphisms on CNR1 gene in each group. A logistic regression analysis was performed to find the association between SPEM abnormality and the number of (AAT)n trinucleotide repeats. Results The natural logarithm value of signal/noise ratio (Ln S/N ratio) of the good SPEM function group was $4.34{\pm}0.29$ and that of the poor SPEM function group was $3.21{\pm}0.70$. In total, 7 types of trinucleotide repeats were identified, each containing 7, 10, 11, 12, 13, 14, and 15 repeats, respectively. In the patients with $(AAT)7$ allele, the distributions of the good and poor SPEM function groups were 18 (11.1%) and 19 (11.0%) respectively. In the patients with $(AAT)_{10}$ allele, $(AAT)_{11}$ allele, $(AAT)_{12}$ allele, $(AAT)_{13}$ allele, $(AAT)_{14}$ allele and $(AAT)_{15}$ allele, the distributions of good and poor SPEM function groups were 13 (8.0%) and 12 (7.0%), 4 (2.5%) and 6 (3.5%), 31 (19.8%) and 35 (20.3%), 51 (31.5%) and 51 (29.7%), 36 (22.2%) and 45 (26.2%), 9 (5.6%) and 4 (2.3%) respectively. As the number of (AAT) n repeat increased, there was no aggravation of abnormality of SPEM function. Conclusions There was no significant aggravation of SPEM abnormality along with the increase of number of (AAT)n trinucleotide repeats in the CNR1 gene in Korean patients with schizophrenia.

• The Korean Journal of Nuclear Medicine Technology
• /
• v.13 no.3
• /
• pp.102-109
• /
• 2009

Purpose: Flash 3D (pixon(R) method; 3D OSEM) was developed as a software program to shorten exam time and improve image quality through reconstruction, it is an image processing method that usefully be applied to nuclear medicine tomography. If perfoming brain diamox perfusion scan by reconstructing subtracted images by Flash 3D with shortened image acquisition time, there was a problem that SNR of subtracted image is lower than basal image. To increase SNR of subtracted image, we use LEAP collimators, and we emphasized on sensitivity of vessel dilatation than resolution of brain vessel. In this study, our purpose is to confirm possibility of application of LEAP collimators at brain diamox perfusion tomography, identify proper reconstruction factors by using Flash 3D. Materials and methods: (1) The evaluation of phantom: We used Hoffman 3D Brain Phantom with $^{99m}Tc$. We obtained images by LEAP and LEHR collimators (diamox image) and after 6 hours (the half life of $^{99m}Tc$: 6 hours), we use obtained second image (basal image) by same method. Also, we acquired SNR and ratio of white matters/gray matters of each basal image and subtracted image. (2) The evaluation of patient's image: We quantitatively analyzed patients who were examined by LEAP collimators then was classified as a normal group and who were examined by LEHR collimators then was classified as a normal group from 2008. 05 to 2009. 01. We evaluate the results from phantom by substituting factors. We used one-day protocol and injected $^{99m}Tc$-ECD 925 MBq at both basal image acquisition and diamox image acquisition. Results: (1) The evaluation of phantom: After measuring counts from each detector, at basal image 41~46 kcount, stress image 79~90 kcount, subtraction image 40~47 kcount were detected. LEAP was about 102~113 kcount at basal image, 188~210 kcount at stress image and 94~103 at subtraction image kcount were detected. The SNR of LEHR subtraction image was decreased than LEHR basal image about 37%, the SNR of LEAP subtraction image was decreased than LEAP basal image about 17%. The ratio of gray matter versus white matter is 2.2:1 at LEHR basal image and 1.9:1 at subtraction, and at LEAP basal image was 2.4:1 and subtraction image was 2:1. (2) The evaluation of patient's image: the counts acquired by LEHR collimators are about 40~60 kcounts at basal image, and 80~100 kcount at stress image. It was proper to set FWHM as 7 mm at basal and stress image and 11mm at subtraction image. LEAP was about 80~100 kcount at basal image and 180~200 kcount at stress image. LEAP images could reduce blurring by setting FWHM as 5 mm at basal and stress images and 7 mm at subtraction image. At basal and stress image, LEHR image was superior than LEAP image. But in case of subtraction image like a phantom experiment, it showed rough image because SNR of LEHR image was decreased. On the other hand, in case of subtraction LEAP image was better than LEHR image in SNR and sensitivity. In all LEHR and LEAP collimator images, proper subset and iteration frequency was 8 times. Conclusions: We could archive more clear and high SNR subtraction image by using proper filter with LEAP collimator. In case of applying one day protocol and reconstructing by Flash 3D, we could consider application of LEAP collimator to acquire better subtraction image.

• 한국지구물리탐사학회:학술대회논문집
• /
• 2002.09a
• /
• pp.139-162
• /
• 2002

Since weak zones or geological lineaments are likely to be eroded, weak zones may develop beneath rivers, and a careful evaluation of ground condition is important to construct structures passing through a river. Dc resistivity surveys, however, have seldomly applied to the investigation of water-covered area, possibly because of difficulties in data aquisition and interpretation. The data aquisition having high quality may be the most important factor, and is more difficult than that in land survey, due to the water layer overlying the underground structure to be imaged. Through the numerical modeling and the analysis of case histories, we studied the method of resistivity survey at the water-covered area, starting from the characteristics of measured data, via data acquisition method, to the interpretation method. We unfolded our discussion according to the installed locations of electrodes, ie., floating them on the water surface, and installing at the water bottom, since the methods of data acquisition and interpretation vary depending on the electrode location. Through this study, we could confirm that the dc resistivity method can provide the fairly reasonable subsurface images. It was also shown that installing electrodes at the water bottom can give the subsurface image with much higher resolution than floating them on the water surface. Since the data acquired at the water-covered area have much lower sensitivity to the underground structure than those at the land, and can be contaminated by the higher noise, such as streaming potential, it would be very important to select the acquisition method and electrode array being able to provide the higher signal-to-noise ratio data as well as the high resolving power. The method installing electrodes at the water bottom is suitable to the detailed survey because of much higher resolving power, whereas the method floating them, especially streamer dc resistivity survey, is to the reconnaissance survey owing of very high speed of field work.