• Proceedings of the Korean Institute of Building Construction Conference
• /
• 2008.05a
• /
• pp.35-41
• /
• 2008

Qualitative enhancement of dwelling life has changing the recognition for the environment friendly wood which is being highlighted for its usage as an interior materials. This trend may prove the excellent performance of wood whose inherent characteristics has its comfortable, mild feeling of material, sound resistance and stabilities and the market of interior woods including floor, moulding and wooden panel as finishing interior materials is growing sustainably. However, since this materials is vulnerable to humidity and flame, waterproofing and flame retarding stability, an essential condition for interior materials, together with maintenance, are the main topics to be resolved. From the above-mentioned results, as a result of waterdrop contact angle, wood absorption volume and water content percentage test and the performance test of the processed materials after flame retardant, though there was some submerging time changes among types of woods for ensuring waterproofing performance improvement but as time passes, similar tendency was noticed to be formulated. As the submerging time is increased, so does the absorption volume and accordingly optimal level of range is judged to be drawn in order to ensure excellent performance, taking optimal economy into consideration. Therefore, it is considered that above-mentioned woods could be utilized for waterproof and flame retardant processed interior materials using uniform microwave and in order to put this technology into practical application, a research by way of diversified performance proving is required to be carried out.

• Proceedings of the KIPE Conference
• /
• 2018.07a
• /
• pp.275-277
• /
• 2018

Recently, the importance of an early response has been emphasized due to the large fire. The most efficient method of extinguishing a large fire is early response to a small flame. To implement this solution, we propose a fire detection mechanism based on a deep learning artificial intelligence. In this study, a small amount of data sets is manipulated by an image augmentation technique using rotating, tilting, blurring, and distorting effects in order to increase the number of the data sets by 5 times, and we study the flame detection algorithm using faster R-CNN.

• Journal of the Korean Society of Safety
• /
• v.37 no.5
• /
• pp.14-21
• /
• 2022

The purpose of this study was to apply the signal pattern recognition technique to the digital wireless fire-alarm system and to reduce unwanted fire alarms. In this study, the fire alarms of the K Institute, which operates the largest digital wireless fire-alarm system in Korea, were classified into normal operations and unwanted fire alarms, and these were analyzed and compared with actual fire signals. In addition, by designing a non-fire signal filter and applying it to the K Institute, we confirmed that the monthly unwanted fire alarm rate of all 5,713 detectors decreased sharply. In particular, the unwanted fire alarm rate for flame decreased from 1.09% to 0.11% and the unwanted fire alarm rate for smoke decreased from 0.65% to 0.035%.

• Proceedings of the KSME Conference
• /
• 2008.11b
• /
• pp.3163-3168
• /
• 2008

Pulverized coal (PC) has become an important auxiliary fuel in the iron and steel industry since the technique of pulverized coal injection (PCI) system was developed for iron making. Combustion efficiencies of pulverized coal in blowpipes and tuyeres under various operational are numerically predicted to recognize the performance with the locations of nozzles in a blast furnace. A variety of parameters including the pulverized coal quantities, oxygen amounts, inlet temperature of the tuyeres and mass flow rate of coal carrier gas are taken into consideration. Also In order to develop more efficient than existing coal injection system, this study applies a flame measurement system using a charge couple device (CCD) camera and frame grabber. And it has used algorithms of auto sampling from flame shape information and composed the device for location control of PCI. This study find to further improve the blast furnace performance by the control of PCI locations.

• KIPS Transactions on Software and Data Engineering
• /
• v.5 no.11
• /
• pp.607-614
• /
• 2016

In this paper, we propose an intelligent fire learning and detection system using convolutional neural networks (CNN). Through the convolutional layer of the CNN, various features of flame and smoke images are automatically extracted, and these extracted features are learned to classify them into flame or smoke or no fire. In order to detect fire in the image, candidate fire regions are first extracted from the image and extracted candidate regions are passed through CNN. Experimental results on various image shows that our system has better performances over previous work.

• Journal of the Society of Disaster Information
• /
• v.13 no.4
• /
• pp.529-536
• /
• 2017

In this study, the perceptions were surveyed fire risk and flash flames concerning the firefighters. The results were statistically evaluated according to age, experience and rank. More than 70% of the respondents answered that there is a possibility of unexpected flame exposure in the field of fire, but there was no recognition difference according to age, experience and rank. However, if there is an emergency situation in the field of fire, the survey on the ability to cope with crises showed that there is a difference in perception depending on the age, career, and rank of respondents. From these results, it is expected that strengthening simulation training of unexpected situation will be more urgently required in the future, and measures should be taken to minimize human accidents through improvement of standard operation procedures or supplement of fire suppression education according to unexpected situation.

• Journal of IKEEE
• /
• v.22 no.4
• /
• pp.1079-1087
• /
• 2018

Recently, computer vision field based deep learning artificial intelligence has become a hot topic among various image analysis boundaries. In this study, flames are detected in fire images using the Faster R-CNN algorithm, which is used to detect objects within the image, among various image recognition algorithms based on deep learning. In order to improve fire detection accuracy through a small amount of data sets in the learning process, we use image augmentation techniques, and learn image augmentation by dividing into 6 types and compare accuracy, precision and detection rate. As a result, the detection rate increases as the type of image augmentation increases. However, as with the general accuracy and detection rate of other object detection models, the false detection rate is also increased from 10% to 30%.