• ETRI Journal
• /
• v.46 no.3
• /
• pp.526-537
• /
• 2024

Monitoring fetal growth in utero is crucial to anomaly diagnosis. However, current computer-vision models struggle to accurately assess the key metrics (i.e., head circumference and occipitofrontal and biparietal diameters) from ultrasound images, largely owing to a lack of training data. Mitigation usually entails image augmentation (e.g., flipping, rotating, scaling, and translating). Nevertheless, the accuracy of our task remains insufficient. Hence, we offer a U-Net fetal head measurement tool that leverages a hybrid Dice and binary cross-entropy loss to compute the similarity between actual and predicted segmented regions. Ellipse-fitted two-dimensional ultrasound images acquired from the HC18 dataset are input, and their lower feature layers are reused for efficiency. During regression, a novel region of interest pooling layer extracts elliptical feature maps, and during segmentation, feature pyramids fuse field-layer data with a new scale attention method to reduce noise. Performance is measured by Dice similarity, mean pixel accuracy, and mean intersection-over-union, giving 97.90%, 99.18%, and 97.81% scores, respectively, which match or outperform the best U-Net models.

• Journal of applied mathematics & informatics
• /
• v.31 no.1_2
• /
• pp.99-110
• /
• 2013

In this article, sensitivity analysis of atmospheric dispersion model RIMPUFF is considered. Uncertain parameters are taken to be triangular fuzzy numbers, and sensitivity analysis is carried out by using the Hartley-like measure. Codes for evaluating membership function using the Vertex method and the Hartley-like measure are prepared using Matlab.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1997.10a
• /
• pp.419-423
• /
• 1997

This paper is concerned about the classification of objects together with muti-attributes such as remote sensing image data by using tolerance rough set. To produce more reliable relations from given attributes in the data, we define new similarity measures by using scaling. Our Method will be applied to classify multi-spectral image data.

• The Bulletin of The Korean Astronomical Society
• /
• v.42 no.2
• /
• pp.65.3-65.3
• /
• 2017

With the advent of high energy/power lasers, extreme conditions, such as those found in astrophysical environments, can be reproduced in laboratory. The scaling between laboratory and astrophysical environments, especially for viscosity and resistivity that govern dissipation processes, is not perfect. Yet, the similarity is close enough to make laboratory experiments relevant for astrophysics. The results have been encouraging, in the sense of suggesting the possibility of exploring fundamental physical processes at play in astrophysical phenomena. In this talk, I will review a few successfully performed and ongoing experiments, such as those for turbulence and magnetic field generation in fluid regime and collisionless shock wave in plasma regime.

• 제어로봇시스템학회:학술대회논문집
• /
• 1990.10b
• /
• pp.917-922
• /
• 1990

Pattern classification is an essential step in automatic robotic assembly which joins together finite number of seperated industrial parts. In this paper, a fast and systematic algorithm for classifying occlusion-free objects is proposed, using the notion of incremental circle transform which describes the boundary contour of an object as a parametric vector function of incremental elements. With similarity transform and line integral, normalized determinant curve of an object classifies each object, independent of position, orientation, scaling of an object and cyclic shift of the stating point for the boundary description.

• Journal of the Korean Home Economics Association
• /
• v.38 no.10
• /
• pp.85-95
• /
• 2000

The purpose of this study was to identify brand loyally of women's wear and construct brand positioning maps by using multidimensional scaling(MDS). There were significant differences between brand loyal and indifferent group in ages, income, occupation status and level of education. Significant differences were found between groups in the degree of importance of attributes (design/color, fashion, quality, store image, salesperson's attitude and brand reputation) when evaluating brands. The positioning maps upon the similarity and preference of brand image were composed by use of MDS.

• Journal of Microbiology and Biotechnology
• /
• v.9 no.6
• /
• pp.773-777
• /
• 1999

Scale-up experiments for hydrolysis of beef tallow, fat, and palm kernel with lipase derived from Candida cylindracea were carried out in 1-1, 100-1, and 10,000-1 reactors. The optimum agitation speed for the hydrolysis of the 1-1 reactor was investigated and found to be 350rpm, and this was a basis for the scale-up of agitation speed. The hydrolysis system in this work was the oil-water system in which the hydrolysis seems to process a heterogeneous reaction. An emulsion condition was the most important factor for determining the reaction rate of hydrolysis. Therefore, the scale-up of agitation speed was performed by using the power n = 1/3 in an equation of the rules of thumb method. The geometrical similarity for scaling-up turned out to be unsatisfactory in this study. Thus, the working volume per one agitator was used for the scale-up. In the case of scale-up from a 1-1 reactor to a 100-1 reactor, the hydrolysis of palm kernel was very much scaled-up by initiating the rules of thumb method. However, the hydrolysis of fat and beef tallow in a 100-1 reactor was a little higher than that of the 1-1 reactor because of the difference of geometrical similarity. The scale-up of hydrolysis from the 100-1 reactor to the 10,000-1 reactor was improved compared to that of the 1-1 to 100-1 reactor. The present results indicated that the scale-up of hydrolysis in the oil-water system by the rules of thumb method was more satisfactory under the condition of geometrical similarity. Even in the case where geometrical similarity was not satisfactory, the working volume per one agitator could be used for the scale-up of a heterogeneous enzyme reaction.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.14 no.12
• /
• pp.1223-1232
• /
• 2004

This study is directed toward determining the number and characteristics of psychologically meaningful perceptual dimensions required for assessing the sound quality with respect to vehicle noises, and toward identifying the acoustical and/or psychoacoustical bases underlying the preference and similarity judgments. For the purpose of analyzing the paired comparison data produced by subjective ratings we used nonmetric multidimensional scaling(MDS). The perceptual dimensions based upon preference ratings could explain 76.3 % of the variance by maximum dB(A) and sharpness acum. The correlation between objective and subjective positions of the stimuli is $R^2$=0.97(F(1,13)=195.45, p < .01), corrected $R^2$=0.93. The less the intensity of the stimulus the more becomes the subjective Position would be over-estimated relative to the objective one. The same is valid for the opposite case. The perceptual dimensions based upon similarity judgments could be accounted for 47.8 % and 23.5% of the variance, each of which might be a match for the maximum dB(A) and the sharpness acum, respectively. The correlation between objective and subjective positions of the stimuli is $R^2$=0.94(F(1,13)=92.38, p < .01), corrected $R^2$=0.87. The more the intensity of the stimulus the more becomes the subjective position would be over-estimated relative to the objective one. The same is valid for the opposite case. In other words, it is likely that the larger the amount of two stimuli which to compare would be judged similar. So far it should be further clarified that whether the relationship between preference ratings and psychological distances nay be optimized through which psycho-physical models.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.25 no.1
• /
• pp.59-66
• /
• 2021

The earthquakes are the natural disasters that can cause the most serious damage to civil structures. Therefore, various studies are being conducted to secure the safety of structures against earthquakes. Most studies on the safety or mechanism of civil structures during earthquakes are being conducted based on lab scale test, because real structure tests are impossible when considering the scale of civil structures. The scaling factor proposed by Iai is mainly cited, but when applying the scaling factor proposed by Iai, there are many difficulties in selecting the structural members necessary for the production of the lab scale model. This is because when applying the scaling factor proposed by Iai, the scaling factor must be applied to the elastic modulus, which is the material property of the structure. Therefore, a new method based on Iai's 's similarity law for determining scale factor is applied in this study where the material property of real structure is same as that of lab-scale model. Through the results of this study, it is considered that the characteristics of the structure calculated through the lab scale model test can more accurately reflect the characteristics of the real structure.

• East Asian Journal of Business Economics (EAJBE)
• /
• v.7 no.2
• /
• pp.21-30
• /
• 2019

Purpose - In this work, we categorize the 21 shopping items which foreign tourists purchase in South Korea and monitor the level of dissimilarity (or similarity) between each item by utilizing distance matrix, and both hierarchical and k-means cluster analyses, respectively, based on several purpose of visit attributes in 2017. In addition, multidimensional scaling (MDS) method is applied for mining visual appearance of proximities among shopping items based on purpose of visit attributes. Research design and methodology - This study is carried out in 2017 by Ministry of Culture, Sports and Tourism and conduct a face-to-face survey of foreign tourists from 20 countries who purchase shopping items in South Korea. CLUSTER, PROXIMITIES and ALSCAL modules in IBM SPSS 23.0 are used to perform this work. Results - We ascertain that 21 shopping items can be classified into five similar groups which have homogeneous traits by going through two-step cluster analysis. We can position homogeneous places of cluster and shopping items joining each cluster. Conclusions - We can relatively assess patterns and characteristics of each shopping item, come by useful information in activating shopping tour based on the actual state of recognition of foreign tourists and practically apply to each tourism industry on underlying results.