• 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.

• Journal of Powder Materials
• /
• v.27 no.2
• /
• pp.103-110
• /
• 2020

Aluminum (Al) - based powders have attracted attention as key materials for 3D printing because of their excellent specific mechanical strength, formability, and durability. Although many studies on the fabrication of 3D-printed Al-based alloys have been reported, the influence of the size of raw powder materials on the bulk samples processed by selective laser melting (SLM) has not been fully investigated. In this study, AlSi10Mg powders of 65 ㎛ in average particle size, prepared by a gas atomizing process, are additively manufactured by using an SLM process. AlSi10Mg powders of 45 ㎛ average size are also fabricated into bulk samples in order to compare their properties. The processing parameters of laser power and scan speed are optimized to achieve densified AlSi10Mg alloys. The Vickers hardness value of the bulk sample prepared from 45 ㎛-sized powders is somewhat higher than that of the 65 ㎛m-sized powder. Such differences in hardness are analyzed because the reduction in melt pool size stems from the rapid melting and solidification of small powders, compared to those of coarse powders, during the SLM process. These results show that the size of the powder should be considered in order to achieve optimization of the SLM process.

• Journal of the Korean Solar Energy Society
• /
• v.36 no.2
• /
• pp.41-51
• /
• 2016

A solar air heater was designed for supplementary domestic heating. The absorber plate had a series of protruding notches which had triangular openings on the front surface of the absorber plate to direct partial air flow to the rear surface and to enhance the convective heat transfer to the flowing air. The height of the opening as well as the opening configuration was determined by preceding numerical simulations. The experimental model had an absorber plate of 0.78-m width and 1.0-m length which was coated with black paint. The air temperature increased as much as $18^{\circ}C$ for $90-m^3/h$ flow rate when the absorber plate was inclined by $45^{\circ}$ for a clear-day solar irradiation of about $906W/m^2$. The collector efficiency ranged from 69 to 74%. Considering the simplicity of the structure and low manufacturing cost, the solar air heater might have competence as an auxiliary heating device for domestic use. On-site experimental results are presented with discussion for various solar irradiations and air flow conditions.

• Journal of ILASS-Korea
• /
• v.20 no.3
• /
• pp.149-155
• /
• 2015

With increasing of GDP, the registration number of passenger cars has exceeded 20 million last year in Korea. Especially, the registration number of the diesel engine vehicles has been increasing. However, the WHO(World Health Organization) IARC (International Agency for Research on Cancer) has reported that diesel engine exhaust gas is an one of HAPs, which has carcinogenic for human, and they have designated it to Group 1. To solve this problem, exhaust gas from diesel engines has to be controlled. Thus, it has been controlling by European regulatory standard in Korea. On the other hand, in order to meet the enhanced emission regulations, all manufacturing company applied $NO_x$ control device to vehicles such as EGR (Exhaust Gas Recirculation), SCR (Selective Catalytic Reduction) and so on. However, these devices (EGR, SCR) were operated by difference reaction mechanism respectively, and the composition of exhaust gas would be differenced from that of them. In this study, it was conducted to evaluate variety characteristics on changing of exhaust gas composition by each $NO_x$ control device, and the heavy duty diesel trucks were chosen as experimental vehicles. From the result, it revealed that vehicles (with EGR) were discharged higher THC as 52.5% than that of others (with SCR). However, it did not followed that trend, in the case of CO; it was discharged as 57.2% lower than that of others (with SCR). In the future, these data would be used to apply to efficient $NO_x$ control device for meeting to EURO 6.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2003.06a
• /
• pp.1869-1872
• /
• 2003

As SLA(Sterealithography), SLS(Selective Laser Sintering), LOM(Laminated Object Manufacturing), FDM(Fused Deposition Modeling) etc. The FDM system the heart of a study and is developed by Stratasys co. ltd, in US., is small and cheap R.P. The material filament is heated until the material reaches a near-liquid state, it is pumped through a nozzle and become hand with a shape required, and this nozzle move pumping on the previously deposited material. Such FDM system that choice deposition type with X-Y plouter obtain in the thin continue layer by decreasing amount of extrusion or to central the injection amount when the head slow down at the corner, but in the process that fusion wax or resin become hand, deformation occur and it will affect the shape accuracy and the surface roughness. Such effect will depreciate quality and reliability of the product. Therefore, when the product made in actuality, the fundamental study on the basis geometry(surface, volume, line, angle) must be preceded and it have been research by many Free Form Fabrication. So, this basic object study purpose to obtain the fundamental geometry data and to enhance the surface roughness of the shape. And an operant can use the data for the progress of the surface roughness. This study research the estimation and application of the prototype surface roughness by adjustment the injection amount. And basie of this research, describe the pattern of prototype surface roughness and also used the result to estimate the surface of prototype.

• Journal of the Korean Wood Science and Technology
• /
• v.38 no.6
• /
• pp.480-489
• /
• 2010

We investigated the functional characteristics of commercial wood charcoal in Korea and their application as functional raw materials. The areas of analysis were anatomical features, elementary composition, mineral composition, caloric values, anion and far-infrared ray emission, and moisture absorption capacity. Based on the analyses as above mentioned, it is considered that charcoal can be evaluated as functional raw material. In commercial wood charcoal in Korea, there were highly varied depending on manufacturing methods as black charcoal, white charcoal and mechanical charcoal and manufactures for elementary composition, mineral composition, anion emission, far infrared ray emission. Especially, black charcoal showed lower moisture absorption capacity than white charcoal and mechanical charcoal. For charcoal as functional raw material, selective usage are needed based on the analyses of anatomical features, elementary composition, mineral composition, caloric values, anion and far-infrared ray emission, and moisture absorption capacity. Specific charcoal making methods for improving specific functionality, required as functional raw material, are necessary in further research.

• Applied Chemistry for Engineering
• /
• v.23 no.1
• /
• pp.119-123
• /
• 2012

This study presents the effect of oxygen on the $NH_3$ selective catalytic reduction (SCR) by Mn/$TiO_2$ catalyst. The lattice oxygen of catalysts is participate in the low temperature SCR, and the gaseous oxygen directly takes part in the rexoidtion of reduced catalyst. These redox properties of oxygen an play important role in SCR activity and the available capability of lattice oxygen depends on the manganese oxidation state of the catalyst surface. $MnO_2$ species has a higher redox property than that of $Mn_2O_3$ species on deposited $TiO_2$ surface and these manganese oxide states strongly depend on the $TiO_2$ surface area.

• Clean Technology
• /
• v.4 no.1
• /
• pp.6-23
• /
• 1998

Recently, semiconductor industry has grown rapidly because of the large demand for electronic devices and equipment. The semiconductor industries have also played an important role on the economic growth in Korea. As the environmental regulations become strict, the proper environmental management and the well-developed waste minimization technologies in semiconductor industries are two of urgent problems to be solved. The semiconductor manufacturing process consists of a series of continuous chemical processes, such as cleaning, oxidation, diffusion, photolithography, etching and film deposition. During the processes, various environmentally hazardous wastes are produced. The wastes may be classified as wastewater, gaseous pollutants, and solid wastes. For waste minimization, the substitution of raw materials and process optimization techniques are used, while the selective destruction technologies of toxic chemicals contained in the wastes have been reported. Also, new technologies have been developed for source reduction and waste reduction, such as reduction of toxic chemical use and substitution of hazardous liquids with gaseous reactants or solvent.

• Current Photovoltaic Research
• /
• v.1 no.1
• /
• pp.17-26
• /
• 2013

In contrast to conventional crystalline cells, back-contact solar cells feature high efficiencies, simpler module assembly, and better aesthetics. The highest commercialized cell and module efficiency was recorded by n-type back-contact solar cells. However, the mainstream PV industry uses a p-type substrate instead of n-type due to the high costs and complexity of the manufacturing processes in the case of the latter. P-type back-contact solar cells such as metal wrap-through and emitter wrap-through, which are inexpensive and compatible with the current PV industry, have consequently been developed. In this paper the characteristics of EWT (emitter wrap-through) solar cells and their status and prospects for development are discussed.

• Journal of the Korean association of regional geographers
• /
• v.17 no.5
• /
• pp.536-553
• /
• 2011

New urbanization different from the manufacturing based urban growth has been widely discussed in post-modern city, consumption city and others to reflect the urban changes in qualitative manner. Urbanization stage models consider large cities to be in mature stage, reurbanization and the stages are assumed to be experienced successively in lower sized cities. However, since the industrial restructuring implies new urbanization experiences, this study examines new urbanization in diverse aspects such as the population change and the geographic and social characteristics of commuters and migrants for the 6 large cities in Korea. Seoul follows the urbanization stages in order while other 5 large cities are divergent in their transition from the industrial city. Regional large cities reveal broad reverse commuting and selective in-migration of economically active households for better work and housing opportunities available in central cities. Similar to the consumption city thesis with enhanced cultural and recreational amenities in western cities, the widening urban residential function including housing and other services is a new urbanization characteristic of large cities in Korea.