• Journal of Broadcast Engineering
• /
• v.24 no.4
• /
• pp.564-572
• /
• 2019

With the development of deep learning, the performance of facial landmark detection methods has been greatly improved. The heat map regression method, which is a representative facial landmark detection method, is widely used as an efficient and robust method. However, the landmark coordinates cannot be directly obtained through a single network, and the accuracy is reduced in determining the landmark coordinates from the heat map. To solve these problems, we propose to combine integral regression with the existing heat map regression method. Through experiments using various datasets, we show that the proposed integral regression network significantly improves the performance of facial landmark detection.

• The Korean Journal of Food And Nutrition
• /
• v.13 no.1
• /
• pp.1-5
• /
• 2000

Detection for heat-labile enterotoxin(LT) of enterotoxigenic Escherichia coli(ETEC, EC81, O148:H28) and enterotoxin of enterotoxigentic Clostridium perfringents type A(CP, NCTC8238, Hobbs serotype 2) by use of a polymerase chain reaction (PCR) assay were positive reaction, which using LT gene-specific primers of ETEC with a detection limit equivalent from 100ng/${\mu}\ell$ to 1 pg of a DNA fragment of 417-bp in EC81 and enterotoxin gene-specific primers of CP with a detection limit equivalent from 100ng/${\mu}\ell$ to 10pg of a DNA fragment of 364-bp in NCTC8238. Detection for a LT gene of ETEC highly appeared 10-fold sensitivity than an enterotoxin gene of CP.

• Proceedings of the Korean Information Science Society Conference
• /
• 2012.06b
• /
• pp.369-371
• /
• 2012

Shadow detection and elimination is a critical issue in vision-based system to improve the detection performance of moving objects. However, traditional algorithms are useless at night time because they used the chromaticity and brightness information from the color image sequence. To obtain the high detection performance, we can use the thermal camera and there are shadows by the heat not the light. We proposed a novel algorithm to detect and eliminate the shadows using the thermal intensity and the locality property. By combining two results of the intensity-based and locality-based, we can detect the shadows by the heat and improve the detection performance of moving object.

• Proceedings of the KOSOMBE Conference
• /
• v.1994 no.12
• /
• pp.9-13
• /
• 1994

The Heat Anisotropic Diffusion Method has shown very effective for the contour detection of 2-D echocardiogram. To implement this algorithm, we have to choose the parameter C, K, and the threshold level. The choice of C and K are not very sensitive for the good edge detection of the echocardiogram, however the choice of the threshold level is very critical. Until now the threshold level is chosen by the trial and error method. In this paper, we present an automatic threshold decision method from the histogram of the gradient of boundary-like pixels.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.20 no.9
• /
• pp.828-833
• /
• 2010

Steam leakage is one of the major issues for the structural fracture of pipes of nuclear power plants. Therefore a method to inspect a large area of piping systems quickly and accurately is needed. In this paper, we proposed the method for the detecting steam leakage by using image signal processing. Our basic idea come from heat shimmer which shine with a soft light that looks as if it shakes slightly. To test the performance of this technique, experiments have been performed for simple heat source and steam generator. Results show that the proposed technique is quite powerful in the steam leak detection.

• Journal of Korea Multimedia Society
• /
• v.19 no.8
• /
• pp.1310-1319
• /
• 2016

Face detection is the first step in a wide range of face applications. However, detecting faces in the wild is still a challenging task due to the wide range of variations in pose, scale, and occlusions. Recently, many deep learning methods have been proposed for face detection. However, further improvements are required in the wild. Another important issue to be considered in the face detection is the computational complexity. Current state-of-the-art deep learning methods require a large number of patches to deal with varying scales and the arbitrary image sizes, which result in an increased computational complexity. To reduce the complexity while achieving better detection accuracy, we propose a fully convolutional network-based face detection that can take arbitrarily-sized input and produce feature maps (heat maps) corresponding to the input image size. To deal with the various face scales, a multi-scale network architecture that utilizes the facial components when learning the feature maps is proposed. On top of it, we design multi-task learning technique to improve detection performance. Extensive experiments have been conducted on the FDDB dataset. The experimental results show that the proposed method outperforms state-of-the-art methods with the accuracy of 82.33% at 517 false alarms, while improving computational efficiency significantly.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.35 no.11
• /
• pp.1111-1117
• /
• 2011

This paper proposed a fault detection and diagnosis (FDD) algorithm using nonlinear regression models, focusing especially on heat exchanger faults. This research concerned four working modes: those with no fault, evaporator fault, condenser fault, and evaporator and condenser faults. This research used no fault mode data to create an FDD algorithm. Using the no fault mode data, correlation functions for predicting the degree of superheat or subcool of heat exchangers (an evaporator and a condenser) were derived. Each correlation function has five inputs and one output. Based on these correlation functions, it is possible to predict the degree of superheat or subcool of each heat exchanger under various working conditions. The FDD algorithm was developed by comparing the predicted value and the simulation value. The FDD algorithm works well in all four working modes.

• Journal of the Korean Society of Visualization
• /
• v.20 no.1
• /
• pp.11-17
• /
• 2022

Among the various factors that cause the performance decrease of heat exchangers used in many industries, flow path blocking is one of the important and serious factor. In order to solve this problem, proper maintenance and management of the heat exchanger is important and emphasized. In this study, we developed and algorithm that can quantitatively determine and diagnose the normal and blocked areas of fin-tube heat exchanger using pattern analysis, Gaussian Edge Detection, Image Processing and Affine Transformation techniques. The developed algorithms was applied to the actual heat exchanger and the performance was evaluated by comparing with the manual results. From these results, it was proved that the developed algorithm is effective in evaluating the pollution ratio of the fin-tube heat exchanger.

• Asian-Australasian Journal of Animal Sciences
• /
• v.9 no.6
• /
• pp.671-677
• /
• 1996

Erythrocytes from three different animal species were used to determine mannose-sensitive hemagglutination (MSHA) and mannose-resistant hemagglutination (MRHA) of 755 isolates obtained from rectal swabs of healthy pig. In addition, colony hybridization using digoxigenin-dUTP labeled polynucleotide probes was performed for the detection of heat-stable and heat-labile enterotoxin genes carried by MRHA positive isolates. Of 755 strains, 9, 4 and 28 strains gave a positive MRHA with bovine, equine and pig erythrocytes, respectively. Of these isolates, 28 (3.7%) were characterized for positive MRHA by at least one blood. Seven isolates gave a positive MRHA with two kinds of blood. Three gave a positive MRHA with three kinds of blood. Twenty-eight strains, while positive in MRHA, yielded negative signals in the colony hybridization assay for the detection of heat-stable (STaI and STaII) and heat-labile (LT) enterotoxin genes in E. coli.

• 한국생물공학회:학술대회논문집
• /
• 2000.04a
• /
• pp.355-356
• /
• 2000

Three 'stress probe' plasmids were constructed and characterized which utilize a green fluorescent protein (CFP) as a non-invasive reporter to elucidate Escherichia coli cellular stress responses in quiescent or 'resting' cells. Facile detection of cellular stress levels was achieved by fusion of three heat shock stress protein promoter elements, those of the heat shock transcription factor ${\sigma}^{32}$, pretense subunit ClpB, and chaperone DnaK, to the reporter gene $gfp_{uv}$. When perturbed by chemical or physical stress (such as heat shock, nutrient (amino acid) limitation, addition of IPTG, acetic acid, ethanol, phenol, antifoam, and salt (osmotic shock), the E. coli cells produced GFPuv which was easily detected from within the cells as emitted green fluorescence. A temporal and amplitudinal mapping of these responses was performed, demonstrating regions where quantitative delineation of cell stress was afforded.