• 한국음향학회지
• /
• 제19권3호
• /
• pp.20-25
• /
• 2000

본 논문에서는 3차원 물체 인식을 위한 표면 분류 및 임계치 선정 방법에 대해 제안 하고자 한다. 3차원 영상 처리는 크게 거리 영상의 획득과 특징 추출 그리고 정합 과정으로 이루어진다. 본 논문에서는 전체 3차원 영상 처리 시스템중 거리 영상을 입력으로 했을 시 형상 특징을 추출하는 방법에 대해 제안하고자 한다. 이를 위해 첫째, 거리 영상의 깊이 변화 부호 값의 분포 특성에 따라 표면을 분류하는 방법을 제안하고자 한다. 또한 평균 곡률과 가우스 곡률을 이용하여 표면을 분류했던 기존 방법을 토대로 그의 문제점이었던 실제 거리 영상에서의 임계치 선정 방법에 대하여 제안하고자 한다. 끝으로 제안한 방법의 유용성을 실험에 의해 입증하고자 한다.

• 사상체질의학회지
• /
• 제17권2호
• /
• pp.46-51
• /
• 2005

1. Objectives The purpose of this study is to objectify the diagnosis of Sasang Constitution. Using the methods of this study, it will improve to classificate Sasang Constitution. 2. Methods 1) Automatic feature extraction of human frontal faces for Sasang Constitution classification. 2) Color feature extraction of human frontal faces (1)Erosion filtering (skin-white, the other-black) (2) Median median 3. Results and Conclusions Observing a person's shape has been the major method for Sasang Constitution classification, which usually has been dependent upon doctor's intuition as of these days. We are developing an automatic system which provides objective basic data for Sasang Constitution classification. For this, in this paper, firstly, the signal processing techniques are applied to automatic feature extraction of human frontal faces for Sasang Constitution classification. The experiment is conducted to verify the effectiveness of the proposed system.

• 한국재료학회지
• /
• 제16권10호
• /
• pp.633-638
• /
• 2006

Silica powders with spherical shape and narrow size distribution were prepared by large-scale ultrasonic spray pyrolysis applying the droplet classification apparatus. On the other hand, silica powders prepared by large-scale ultrasonic spray pyrolysis without droplet classification apparatus had broad size distribution. Droplet classification apparatus used in this paper applied the principles of cyclone and dispersion plate with small holes. The droplets formed from the ultrasonic spray generator applying the droplet classification apparatus had narrow size distribution. The droplets with fine and large sizes were eliminated by droplet classification apparatus. The optimum flow rate of the carrier gas and diameter of the hole of the dispersion plate were studied to reduce the size distribution of the silica powders prepared by large-scale ultrasonic spray pyrolysis. The size distribution of the silica powders prepared by large-scale ultrasonic spray pyrolysis at the optimum preparation conditions was 0.76.

• International Journal of Computer Science & Network Security
• /
• 제22권10호
• /
• pp.271-281
• /
• 2022

Manual seed classification challenges can be overcome using a reliable and autonomous seed purity identification and classification technique. It is a highly practical and commercially important requirement of the agricultural industry. Researchers can create a new data mining method with improved accuracy using current machine learning and artificial intelligence approaches. Seed classification can help with quality making, seed quality controller, and impurity identification. Seeds have traditionally been classified based on characteristics such as colour, shape, and texture. Generally, this is done by experts by visually examining each model, which is a very time-consuming and tedious task. This approach is simple to automate, making seed sorting far more efficient than manually inspecting them. Computer vision technologies based on machine learning (ML), symmetry, and, more specifically, convolutional neural networks (CNNs) have been widely used in related fields, resulting in greater labour efficiency in many cases. To sort a sample of 3000 seeds, KNN, SVM, CNN and CNN-SVM hybrid classification algorithms were used. A model that uses advanced deep learning techniques to categorise some well-known seeds is included in the proposed hybrid system. In most cases, the CNN-SVM model outperformed the comparable SVM and CNN models, demonstrating the effectiveness of utilising CNN-SVM to evaluate data. The findings of this research revealed that CNN-SVM could be used to analyse data with promising results. Future study should look into more seed kinds to expand the use of CNN-SVMs in data processing.

• 한국지리정보학회지
• /
• 제17권2호
• /
• pp.121-135
• /
• 2014

최근 픽셀기반분류보다 더 많은 정보를 이용할 수 있는 객체기반에 대한 연구가 활발히 진행 중이다. 따라서, 본 연구는 인제군 Demilitarized Zone(DMZ)지역 일원을 대상으로 객체기반 분류기법을 이용한 토지피복분류를 실시하였다. 분류항목은 환경부 기준의 대분류 항목 7개, 중분류 항목 13개로 선정하였고, 사용된 인자는 분광 값의 평균과 표준편차, Grey Level Co-occurrence Matrix(GLCM)의 Homogeneity를 사용하여 감독분류방법 중 최근린기법을 이용하여 계층적 토지피복도를 구축하였다. 구축된 토지피복도를 이용하여 남방한계선으로부터의 거리와 Digital elevation model(DEM)을 통해 지형특성에 따른 분류항목 별 분포 특성을 분석하였다. 객체기반 분류를 위한 최적 가중치는 Scale 72, Shape 0.2. Color 0.8, Compactness 0.5, Smoothness 0.5로 선정하였고, 가중치 선정과정에서 Scale, Shape, Color가 가장 많은 영향을 주었다. 대분류 토지피복분류는 산림, 초지, 시가지의 순으로 각각 약 92%, 약 5%, 약 2%였으며, 중분류 토지피복분류는 활엽수림, 혼효림, 침엽수림의 순으로 각각 약 44%, 약 42%, 약 6% 순으로 분포하였다. 토지피복형태에 따른 분포특성을 보면, 남방한계선으로부터 2km이내의 지역에서 공공시설 지역과 도로의 이용이 높았으며, 남방한계선 6km이상의 지역에서는 논과 밭, 나지의 형태가 많았다. 산림은 표고 600m, 경사 $30^{\circ}$이상의 지역에서는 면적분포가 높았고, 농업지역과 나지, 초지는 표고 600m, 경사 $30^{\circ}$이하의 지역에서 면적분포가 높았다.

• 패션비즈니스
• /
• 제12권2호
• /
• pp.87-99
• /
• 2008

This study aims to classify lower trunk body types of Chinese men in their middle age (30s and 40s) and suggest the standard for them to design slacks pattern. Mollison's relation deviations were used to analyze the direct measurement items in this research compared to those of Size Korea 2004. Though Korean middle-aged men were higher in most items than Chinese counterparts, all items except mid-thigh circumference and side hip length were merely within the range of ${\pm}1\sigma$. According to the results of size classification by absolute values, factor analysis extracted to 2 factors(horizontal size of lower body and vertical size of lower body), and cluster analysis brought about 3 types(type1: tall and thin trunk (36.9%), type 2 : normal height and thick trunk(45.5%), and type 3 : short and normal trunk(17.6%) with some significant differences among them. Also, the results of shape classification by index values, they were extracted 4 factors (waist-hip flatness, waist-hip cross section, vertical waist-groin and hip-surface length) by factor analysis and revealed 3 types(type 1: different waist-hip width, thick, long waist, long upper hip, and short hip-surface length(27.8%), type 2 : different waist-hip width, flat, short upper hip, high hip and groin, and average hip-surface length(29.4%), and type 3: small waist-hip width, thick, average upper hip, and short hip-surface length (42.8%)) with significant differences among them by cluster analysis. The results of standard body types by shape-size combination, 19subjects(10.16%) under these values are regarded as standard body types. Significance was not seen in all items in the t-test results between the total group and the standard group. The latter had lower variation coefficients and smaller individual differences than the former. However, in-depth research is required for generalization since this research is limited to a small number of subjects in Ningbo of Zhejiang.

• Journal of Electrical Engineering and information Science
• /
• 제3권2호
• /
• pp.131-138
• /
• 1998

We derive a back-propagation learning algorithm of fuzzy neural networks using fuzzy operations, which preserves the shapes of fuzzy numbers, in order to utilize fuzzy if-then rules as well as numerical data in the learning of neural networks for classification problems and for fuzzy control problems. By introducing the shape preseving fuzzy operation into a neural network, the proposed network simplifies fuzzy arithmetic operations of fuzzy numbers with exact result in learning the network. And we illustrate our approach by computer simulations on numerical examples.

• 한국공작기계학회:학술대회논문집
• /
• 한국공작기계학회 2004년도 추계학술대회 논문집
• /
• pp.280-285
• /
• 2004

In this study, we made a comparative study of backpropagation neural network and probabilistic neural network and bayesian classifier and perceptron as shape recognition algorithm of welding flaws. For this purpose, variables are applied the same to four algorithms. Here, feature variable is composed of time domain signal itself and frequency domain signal itself. Through this process, we comfirmed advantages/disadvantages of four algorithms and identified application methods of four algorithms.

• 패션비즈니스
• /
• 제8권1호
• /
• pp.76-90
• /
• 2004

The purpose of this study was to classify women's face types and to analyze the measurement of face types. For study, 180 adult women(aged between 20 and 29) in Pusan and Ulsan area was sampled to be measured for facial types. Data were analyzed by Frequencies, Means, Duncan's Multiple Range Test, Distinction analysis. The major results were as followed. Women's face types were classified by 6 types and there were round shape(29.4%), oblong shape(18.9%), inverted triangle shape(16.1%), square shape(13.9%), egg shape(11.7%), diamond shape(10.0%) in the subject. Phyiognomic facial height was 182.38mm, the upper face length was 59.82mm, the middle face length 60.82mm, the lower face length 61.76mm, and the index of face length to face breadth was 1.35. The face width was 134.90mm, interocular distance 34.75mm, the nose width 33.93mm, and mouth width was 43.87mm. And also, differences from those measurements like forehead breadth, face length/bizygion breadth, forehead slopper, bigonion breadth, bignathion breadth, bignathion slopper.

• 한국안전학회지
• /
• 제31권6호
• /
• pp.93-98
• /
• 2016

This study analyzed the inflow characteristics of debris flow according to shape of defensive structure and computed risk index. In order to simulate debris flow, two shapes of defensive structure were considered. Initial mass distribution was set with a rectangular shape and defensive structures were set semi-circular shape and rectangular shape, respectively. It was found that a defensive structure with semicircular shape was more vulnerable to debris impact compared with rectangular shape because the flow mass became concentrated in quadrant part of the inner circle. If the velocity of the debris flow was less than 1 m/s, the risk assessment by FII (Flood Intensity Index) was much appropriate. However, when the movement of debris runout was faster than 1 m/s, the risk index of FHR (Flood Hazard Rating) provided improved classification due to its subdivided hazardous range.