• Journal of the Korea Organic Resources Recycling Association
• /
• v.9 no.4
• /
• pp.81-88
• /
• 2001

This experiment was carried out to evaluate the effect of residual food compost application on plant growth and physicochemical properties of soil when residual food compost made from the composting process as part of organic waste recycling was applied in soil as variable rate. The treatments were composed non-fertilizer treatment, control treatment applied with pig compost $20Mg(D.W.)ha^{-1}$, and residual food compost treatments each of applied with 20, 40, 60, $80Mg(D.W.)ha^{-1}$ and were randomized complete block design with three replication. Soil pH after experiment was more increased as more increase application of residual food compost. Other chemical properties of soil including EC were also showed the similar results. The plant growth in treatments applied much of residual food compost at the early growth stage was very damaged and these results were proved at the first growth measurement. Production of the red pepper especially reduced in treatments taken growth demage at the early stage and that of $20Mgha^{-1}$ treatment almost same as control treatment. Bulk density was reduced but porosity was increased according to increase of residual food compost application. Considering the reduction of red pepper production and the demage of growth at the early stage by plenty of salt and water soluble application volume of residual food compost was not permitted over of $30Mgha^{-1}$. Additively, to settle the application volume of residual food compost and to evaluate the plant growth and changes of physicochemical properties of soil, the results taken from continual applying is concluded more important than single application.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.20 no.1
• /
• pp.163-169
• /
• 2020

Recently, research on automation and unmanned operation of machines in the industrial field has been conducted with the advent of AI, Big data, and the IoT, which are the core technologies of the Fourth Industrial Revolution. The machines for these automation processes are controlled based on the data collected from the sensors attached to them, and further, the processes are managed. Conventionally, the abnormalities of sensors are periodically checked and managed. However, due to various environmental factors and situations in the industrial field, there are cases where the inspection due to the failure is not missed or failures are not detected to prevent damage due to sensor failure. In addition, even if a failure occurs, it is not immediately detected, which worsens the process loss. Therefore, in order to prevent damage caused by such a sudden sensor failure, it is necessary to identify the failure of the sensor in an embedded system in real-time and to diagnose the failure and determine the type for a quick response. In this paper, a deep neural network-based fault diagnosis system is designed and implemented using Raspberry Pi to classify typical sensor fault types such as erratic fault, hard-over fault, spike fault, and stuck fault. In order to diagnose sensor failure, the network is constructed using Google's proposed Inverted residual block structure of MobilieNetV2. The proposed scheme reduces memory usage and improves the performance of the conventional CNN technique to classify sensor faults.

• Journal of radiological science and technology
• /
• v.46 no.4
• /
• pp.277-285
• /
• 2023

In this study, pneumonia identification networks with the small number of layers were constructed by using chest X-ray images. The networks had similar trainable-parameters, and the performance of the trained models was quantitatively evaluated with the modification of the network architectures. A total of 6 networks were constructed: convolutional neural network (CNN), VGGNet, GoogleNet, residual network with identity blocks, ResNet with bottleneck blocks and ResNet with identity and bottleneck blocks. Trainable parameters for the 6 networks were set in a range of 273,921-294,817 by adjusting the output channels of convolution layers. The network training was implemented with binary cross entropy (BCE) loss function, sigmoid activation function, adaptive moment estimation (Adam) optimizer and 100 epochs. The performance of the trained models was evaluated in terms of training time, accuracy, precision, recall, specificity and F1-score. The results showed that the trained models with the small number of layers precisely detect pneumonia from chest X-ray images. In particular, the overall quantitative performance of the trained models based on the ResNets was above 0.9, and the performance levels were similar or superior to those based on the CNN, VGGNet and GoogleNet. Also, the residual blocks affected the performance of the trained models based on the ResNets. Therefore, in this study, we demonstrated that the object detection networks with the small number of layers are suitable for detecting pneumonia using chest X-ray images. And, the trained models based on the ResNets can be optimized by applying appropriate residual-blocks.

• Journal of the Korean Association of Oral and Maxillofacial Surgeons
• /
• v.26 no.6
• /
• pp.644-651
• /
• 2000

In cases of severe alveolar bony resorption in the edentulous posterior maxillae, implant placement is limited anatomically due to maxillary sinus. If the ridge is atrophic, the various bone grafting methods are required for the ridge augmentation. But the result of the onlay grafting procedure is not always promising. On the posterior maxilla, maxillary sinus mucosa lifting and bone grafting into the sinus floor, subantral augmentation(SA) technique are recommended. Various SA procedures have been developed for implant installation. We perfer to simultaneous block bone graft and implant installation through the residual alveolar ridge into the grafted block bone to fix the grafted bone and to gain the primary stability of the installed fixture. When a sagittal skeletal discrepancy in present due to the severe alveolar bony resorption of the maxilla, the advancement of the maxilla by Le Fort I osteotomy simultaneously with installation of implant fixtures combined with sinus lifting and interpositional bone graft procedure can be indicated. We applied various SA techniques for implant installtion to the 46 edentulous posterior maxillae, and total 154 implants were installed at our department from 1992 to 1999. Various SA techniques were classified in detail and the indications of each techniques were discussed. The changes of residual bony height following SA procedure were studied. The results were as follows. 1. The SA procedure combined with bone graft and simultaneous fixture installation were performed in 41 cases, 126 fixtures were installed and 5 fixtures were removed out of them. Le Fort I osteotomy procedure combined with sinus lifting and interpositional bone graft simultaneous with fixture installation were performed in 5 cases. Total 28 fixtures were installed and 2 fixtures were removed so far. 2. Autogenous block bone graft into sinus floor were performed in 35 cases, autogenous particulated marrow cancellous bone(PMCB) graft in 9 cases, and demineralized human bone powder in 2 cases. The donor site for bone graft were anterior iliac bone in 39 cases, posterior iliac bone in 3 cases and mandibular symphysis in 1 case and mandibular ramus in 1 case. 3. In 9 cases with which SA procedure had been performed with the block bone graft, the change of pre- and postoperative residual bony height were measured using MPR(multiplanar reformatted)-CT. The mean residual bony height was 8.0mm preoperatively, 20.2mm at 6 months following up operation and we gained average 12.2mm alveolar bony height. So, we can recommend this one-stage subantral augmentation and fixture installation technique as a time conserving, safe and useful method for compromised posterior edentulous maxilla.

• Journal of Chest Surgery
• /
• v.23 no.1
• /
• pp.41-48
• /
• 1990

Thirteen patients underwent repair of atrioventricular septal defect [AVSD] from January 1980 to July 1989 at Kyungpook National University Hospital. Two patients had complete AVSD [Rastelli type A] and eleven patients had partial AVSD [ostium primum atrial septal defect and cleft of anterior mitral leaflet]. In all the patients of partial AVSD, atrial septal defect was closed with Dacron patch and the mitral cleft was approximated with interrupted simple sutures. In one patient of complete AVSD, one patch technique was used to close the atrial and ventricular septal defect, and in the other patient of complete AVSD, two patch technique was used. In six patients, there were associated anomalies; four had isolated ostium secundum ASD, two had patent foramen ovale. Postoperative complete A-V block was noted in a patient of partial AVSD, but it was returned to 1st degree A-V block 30 months later and in another case of partial AVSD, severe congestive heart failure [NYHA functional class IV] due to residual mitral insufficiency was developed postoperatively, but this patient was recovered to the state of functional class I after receiving mitral valve replacement. There was one hospital death [8 %] resulting from low cardiac output.

• 한국전산유체공학회:학술대회논문집
• /
• 2001.05a
• /
• pp.38-44
• /
• 2001

An efficient Gauss-Seidel time integration scheme is developed for solving the Euler and Navier-Stokes equations on unstructured meshes. Roe's FDS is used for the explicit residual computations and van Leer's FVS for evaluating implicit flux Jacobian. To reduce the memory requirement to a minimum level, off-diagonal flux Jacobian contributions are repeatedly calculated during the Gauss-Seidel sub-iteration process. Computational results based on the present scheme show that approximately $15\%$ of CPU time reduction is achieved while maintaining the memory requirement level to $50-60\%$ of the original Gauss-Seidel scheme.

• Korean Journal of Air-Conditioning and Refrigeration Engineering
• /
• v.6 no.4
• /
• pp.406-416
• /
• 1994

In this study the reactor design procedure and method of solid-solid-gas chemical heat pump system using STELF technology were investigated. For manufacturing IMPEX block which is the kernel of reactor, proper salt pair should be selected, and equilibrium temperature drop and COP should be examined for selected salt pair. Moreover, apparent density, residual porosity, and graphite ratio should be calculated to give minimum block volume and mass, and maximum energy density without causing heat and mass transfer problems. Since heat exchange area can be changed with operating condition, reactor diameter, length, and stainless steel thickness should be decided for desired specifications. These procedure and method were applied to the case study of 6kW cold production and 8 hours storage capacity reactor.

• Proceedings of the KOSOMBE Conference
• /
• v.1998 no.11
• /
• pp.175-176
• /
• 1998

The automated ECG diagnostic systems that are used in hospitals have low performance of P wave detection when faced with some diseases such as conduction block. So, the aim of this study is the improvement of detection performance in conduction block which is low in P wave detection. Median QRS-T segments were subtracted from the raw data and residual in the QRS-T regions was zeroed. After band pass filtering, we applied approximated length transformation to detect P wave.

• Proceedings of the KIEE Conference
• /
• 2000.07c
• /
• pp.1877-1879
• /
• 2000

This paper deals with the change of the micro-structure of ZnO associated with lightning surge current and ageing test. In this work, a surge current generator which can produce 8/20 [${\mu}s$], 6 [kA] impulse current is designed and fabricated to simulate the lightning impulse current. The residual voltage and leakage current flowing to ZnO blocks are observed. Also a compensation circuit was used in resistive current measurement. The micro-structures of ZnO arrester block were significantly changed by lightning surge current and accelerated temperature ageing test.

• Proceedings of the KSRS Conference
• /
• 2003.11a
• /
• pp.888-890
• /
• 2003

The objective of this investigation is to establish a simple yet effective block adjustment procedure for the orthorectification of multi-orbit satellite images. The major works of the proposed scheme are: (1) adjustment of satellite‘s orbit accurately, (2) calculation of the error vectors for each tie point using digital terrain model and ray tracing technique, (3) refining the orbit using the Least Squares Filtering technique and (4) generation of the orthophotos. In the process of least squares filtering, we use the residual vectors on ground control points and tie points to collocate the orbit. In orthorectification, we use the indirect method to generate the orthoimage. Test areas cover northern Taiwan. Test images are from SPOT 5 satellite. Experimental results indicate that proposed method improves the relative accuracy significantly.