에지 검출을 위한 변형된 top-hat 및 bottom-hat 변환 알고리듬에 관한 연구

백운석(Woon Seok Baek),이하운(Ha Woon Lee) 한국전자통신학회 2016 한국전자통신학회 논문지 Vol.11 No.9

에지는 영상의 가장 기본적인 특징을 나타내며, 에지 검출은 영상처리 분야 및 컴퓨터 비전 영역에서 매우 중요한 역할을 한다. 이러한 에지를 검출하기 위한 연구들이 국내․외적으로 많이 수행되고 있다. 기존의 에지 검출 방법에는 로버츠, 소벨, 프리윗, 라플라시안 등 고정된 값의 마스크를 사용하는 방법들이 있으며 모폴로지 처리 기술 가운데 팽창과 침식을 이용하는 모폴로지 그라디언트 방법 등이 있다. 그러나 이러한 방법들은 대각선 방향이나 완만한 영상의 변화가 있는 경우 에지 검출이 잘 되지 않는 문제가 있다 따라서 본 논문에서는 이러한 경우에도 에지 검출이 잘 되는 변형된 top-hat 및 bottom-hat 변환 방식의 에지 검출 알고리듬을 제안하였다. 제안된 알고리듬을 기존의 방법들과 비교하여 에지 검출 영상을 제시하였으며 코사인 기반의 유사도를 사용하여 성능 및 유사성을 평가하였다. Edge is the basic characteristic of image, edge detection is very important in image processing applications and computer vision area. Many studies are being performed to detect these edges by domestic and foreign researchers. The conventional edge detection methods such as Roberts, Sobel, Prewitt, and Laplacian etc, which are using a fixed value of mask are widely used and morphological gradient which uses dilation and erosion among morphology process techniques is also widely used. But these methods does not detect edges well in the diagonal direction or gradually changing image parts. Accordingly, in this paper, the modified top-hat and bottom-hat transform algorithms which are detecting edges well in the parts of diagonal direction or gradually changing image are proposed. The proposed algorithms present the detected edge images compared with the conventional methods and are evaluated performance by using cosine similarity.

Four-Leg Object Recognition for Service Robot Based on Top-hat Transformation

Hao Wu,Guohui Tian,Xinran Wang,Fei Lu 보안공학연구지원센터 2014 International Journal of Control and Automation Vol.7 No.12

A method of object recognition for objects with special topological structure is developed based on top-hat transformation for service robot. The objects with four legs are easily detected and recognized by service robot with laser ranger array by this method. First, the top-hat transformation in one dimension is reviewed, then the recognition strategy of self-adapting threshold for objects with special topological structure is proposed, and the general data process for object recognition and position is proposed and analyzed. Experimental results show that the process of object recognition based on top-hat transformation proposed in this article is an effective and accurate application.

초저비트율 동화상 압축을 위한 post-it 변환을 이용한 수리형태학적 상세부분(details)추출의 효율개선

허시행,어진우,Huh, Si-Heng,Eo, Jin-Woo 한국전기전자학회 1998 전기전자학회논문지 Vol.2 No.2

영상 압축을 위한 방법으로서 수리형태학적 상세부분(details)의 추출을 위한 방법인 post-it 변환 과정 중에서 기존에는 상세부분(details)를 추출하는 단계의 참조영상을 일반 top-hat 결과로 취함으로서 상세부분(details)의 정보량이 많아 지고 밝고 어두운 상세부분(details)을 뺀 smoothed 영상이 평활화 되지 않고 오히려 공간주파수가 높아지는 단점이 있으므로 본 논문에서는 그러한 단점을 극복하기 위해 참조영상을 일반 top-hat이 아닌 재생 top-hatd의 사용을 제안함으로서 상세부분(details) 영상과 smoothed 영상의 압축효율 증대와 재생과정상의 반복연산수 감소 효과가 발생함을 보인다. 그러나 이러한 결과는 단독적으로 사용되는것이 아니라 동화상의 고효율 압축시 그 일부 과정으로 이용하며 이러한 변형 post-it 변환을 분할기반 부호화와 연계하여 동화상 비디오 시퀸스에 적용함으로서 본래 목적하였던 초저속 동화상 전송을 위한 고효율 압축을 실현하고자 한다. In this paper, a new morphological details extraction algorithm is proposed. It is known that separate transmission of details and background smoothed image is a powerful technique for very low bit-rate video data transmission. Several details extraction algorithms show relatively large variation of grayscale levels in details, smoothed image, which is the difference between original and details images, provides highly distortedand complicated result. In order to remedy those pitfails, and thus to improve the coding efficiency, we propose a new algorithm using the reconstruction top-hat result as the reference image in process of obtaining details, instead of top-hat result, which is used fur existing a1gorithms. Experimental results show that details, extracted using the proposed algorithm, are much similar to original image, and thus the smoothed image is not too complicated.

적외선 영상에서 다층 구조 요소 NWTH 변환을 이용한 소형 표적 검출 방법

김병익(Byoung-Ik Kim),배태욱(Tae-Wuk Bae),김영춘(Young-Choon Kim),안상호(Sang-Ho Ahn),김덕규(Duk-Gyoo Kim) 한국정보기술학회 2011 한국정보기술학회논문지 Vol.9 No.7

In this paper, we propose a small target detection method using multi-structuring element new white top-hat (NWTH) transformation in the infrared (IR) images. The white top-hat (WTH) transformation has not effective performance in case of dim targets in IR images with many clutters. Because the NWTH transformation uses structuring elements considering target region and background region, the NWTH transformation is superior than the WTH transformation in viewpoint of target enhancement. But the performance of NWTH transformation depends on size of structuring element. To solve this detect in top-hat based transformations, a automatic decision method of structuring element is needed. Because this selection standard of structuring element must be objective and adaptive in various IR images, that is performed by candidate target-to-clutter ratio gain (CTCRG). The proposed method decides multi-structuring element by using the CTCRG and its recursive method in the NWTH transformation. And then small targets can be detected through a sum image of the NWTH transformation images by multi-structuring element. We confirmed that the proposed method has superior image enhancement performance in target region, compared with existing WTH and NWTH transformation.

Whole-brain imaging with receive-only multichannel top-hat dipole antenna RF coil at 7 T MRI

KUMARSUCHIT,Yoon Jun-Sik,Kim Jong-Min,Lee Chulhyun,오창현 한국물리학회 2022 THE JOURNAL OF THE KOREAN PHYSICAL SOCIETY Vol.80 No.9

This work investigates the construction and performance of an eight-channel top-hat dipole receiver RF coil with a capacitive plate to increase the longitudinal whole-brain coverage and receiver sensitivity gain in the brain at 7 T MRI. The construction method for top-hat dipole-based receiver RF coil by adjusting the length and structure corresponding to each channel consists of tuning, matching, balun, and detuning circuitry. Electromagnetic simulations were analyzed on a 3-D human model to evaluate B1 + efficiency and specific absorption rate deposition. Coil performance was evaluated in the human head imaging in vivo. EM simulation results indicated a higher B1 − sensitivity in the brain and z-directional coverage of the proposed eight-channel receiver RF coil. The MR images were acquired with an identical field of view showing the receiver coverage improvement in the brain when capacitive plates are used. The MR images also show the clear visibility of the complete set of the cervical vertebrae as well as the spinal cord. The acquired MRI results demonstrate the capability of the proposed RF coil to increase the receiver coverage in the longitudinal direction. Moreover, the B1 + efficiency, as well as receiver sensitivity in the brain, can be substantially improved with the use of multilayered capacitive plates of proper shape and size in conjunction with an RF coil.

Urban Road Extraction from Aerial Photo by Linking Method

Yang, Sung-Chul,Han, Dong-Yeo,Kim, Min-Suk,Kim, Yong-Il Korean Society of Surveying 2003 Korean journal of geomatics Vol.3 No.1

We have seen rapid changes in road systems and networks in urban areas due to fast urbanization and increased traffic demands. As a result, many researchers have put greater importance on extraction, correction and updating of information about road systems. Also, by using the various data on road systems and its condition, we can manage our road more efficiently and economically. Furthermore, such information can be used as input for digital map and GIS analysis. In this research, we used a high resolution aerial photo of the roads in Seongnam area. First, we applied the top-hat filter to the area of interest so that the road markings could be extracted in an efficient manner. The lane separation lines were selected, considering the shape similarity between the selected lane separation line and reference data. Next, we extracted the roads in the urban area using the aforementioned road marking. Using this technique, we could easily extract roads in urban area in semi-automatic way.

19 세기 한국의 갓과 영국의 탑햇에 관한 특징 및 디자인 사례연구

김호연(Kim Ho-Yeon),박규은(Park Kyu Eun) 한국조형디자인학회 2017 조형디자인연구 Vol.20 No.4

하루가 다르게 새로운 컬렉션들이 선보여지고 인터넷을 통해 빠르게 대중들에게 전달되는 현시 대에 패션은 중요한 산업이자 한 문화를 보여주는 요소이기도 하다. 빠르게 변화하고 무수한 제품 들이 쏟아져 나오는 가운데 브랜드의 개성을 표현하고, 소비자들의 관심을 유발하기 위해 다수의 패션 브랜드들은 사람의 얼굴에 가장 가까우면서 단번에 시선을 끌 수 있는 방법으로 모자 혹은 헤드피스라고 불리는 소품을 적극적으로 사용하고 있는 추세이다. 이러한 트렌드에 맞추어 본 연구는 한국과 영국 두 나라를 각각 상징하는 모자 연구를 통해 국내에서는 활성화되어있지 않은 모자 산업에 새로운 관점과 접근 가능성을 제시하고자 한다. 각 나라를 대표하는 모자로는 19세기의 남성 복식에서 중요한 위치를 차지하는 한국의 갓과 영국의 탑햇을 그 비교 대상으로 설정하였다. 두 모자의 발달과정과 특징을 연구하고 현대에 재해석되어 사용되는 디자인 사례연구를 통하여 모자 디자인을 위한 기초 자료로써의 역할을 기대한다. The purpose of this study is to compare the features and history of Korean traditional headwear, the gat, to the British top hat of 19th century, and to examine redesigned gats and top hats by contemporary designers. In the 21st century, fashion is an important, rapidly changing industry that reflects the wider culture and marks from where brand identity is often derived. Indeed, many high fashion brands are using headpieces to attract consumers and forge their identities. Hats have not been fully explored by the modern fashion industry in Korea yet, but there are historical precedents in the Joseon Dynasty. This paper suggests broad possibilities in the application of contemporary hat design inspired by traditional headwear. In addition, by examining the development process of the gat and top hat, this paper expects to contribute a framework by which other hats can be reimagined using traditional elements.

햇(Hat) 패턴 개발을 위한 시판 제품 분석 및 패턴 선호도 평가

김나영(Na Young Kim),장정아(Jeong Ah Jang) 한국디자인문화학회 2013 한국디자인문화학회지 Vol.19 No.4

본 연구는 다양한 야외 레포츠를 즐기는 현대인들의 모자 수요 기호에 맞춰 야외 활동 시 착용하는 모자들 중 가장 대표적인 햇 패턴 개발을 위한 기초 지표를 얻기 위해 야외활동이 많은 20대 성인여자를 대상으로 국내 스포츠·레져 브랜드에서 시판 되고 있는 햇의 크라운 형태, 색상 분포, 원단의 종류, 사이즈표기방법, 기능성 등의 제품생산 실태를 조사하였으며, 적절한 크라운 및 브림의 크기와 형태 등을 파악하기 위해 모자 패턴북과 시판제품 패턴의 착의실험을 통하여 각 패턴들의 여유량과 맞음새를 확인하고, 실무패턴과 모자 패턴북의 패턴 차이를 알아보고 이를 바탕으로 보다 이상적인 햇 패턴을 개발하는 기초자료를 제시하는데 그 목적이 있다. 자료의 분석은 PASW statistics 18.0 통계 프로그램을 사용하여 기술통계와 분산분석을 실시하였으며, 연구 결과 다음과 같다. 시판되고 있는 햇의 크라운 형태는 원통형, 세로절개형, 카우보이형, 가로절개형, 반구형, 테이프형 등 6개의 유형으로 나눌 수 있었으며, 원통형이 48.6%로가장 많았다. 색상 분포를 조사한 결과, 검정 13.1%, 네이비 11.6%, 베이지10.9% 순으로 많이 생산되고 있었으며, 원단의 종류를 조사한 결과 폴리에스테르100%의 고어텍스 원단과 나일론 100%의 서플렉스 원단이 전체의 80%이상을 차지하였다. 시판 햇의 대표 기능성을 조사한 결과 통기성과 사용의 편의, 햇볕의 효과적인 차단, 보온성의 향상 기능으로 분류되었다. 시판 햇 패턴 2종과 모자 패턴북의 햇 패턴 2종의 외관에 대한 착의 평가 결과 탑크라운의 크기 기준은 세로17.4㎝, 가로 15.8㎝의 타원형이었고, 사이드크라운의 높이는 9㎝, 브림의 길이는 6㎝인 B패턴이 전체 평점3.61로가장 좋게 평가 되었으며, 착용감에 대한 착의 평가 결과도 전체 평점 3.35로 가장 좋은 평가를 얻었다. The purpose of this study was to suggest base linedata for development of ideal Hat pattern based on thefollowing facts. In this study, we investigated actualsituation of Hat crown figure, color distribution, typesof fabric, designation method of size, product productionwith functionality for obtaining base line data, whichis essential for development of Hat pattern, targetingwomen in their twenties who enjoy outdoor activities. For the grasp of appropriate size and figure of crownand brim, through cap patternbook and wearingevalution in the market product, we verified tolerancesand fitness of pattern and the difference of workingpattern and cap pattern book. Descriptive statistics and variance analysis in dataanalysis were carried out with PASW statistics 18.0, and the results were as following. Crown type and way of incision in Hat in themarket were divided into 6 crown figures, which arecylinder, cowboy, trans-incision, hemisphere andtape-type, and hemisphere type occupied 48.6% of all, the most. In the analysis of color distribution, black, navy, and beige were produced as much as 13.1%, 11.6% and10.9% in order, respectively. In the analysis of fabrictype and distribution, Goretex with 100% polyester andSupplex with 100% nylon occupied more than 80% ofall fabric in outdoor and golf brand. Representative functionality of Hat were dividedwith air permeability, consumer convenience, sunlightblock and thermokeeping improvement. In the analysis of wearing evaluation about exteriorof two type of Hat pattern and cap pattern in the market, size standard of top crown was ovoid, 17.4cm lengthand 15.8cm width, and B pattern whose height of sidecrown was 9cm, and length of brim was 6cm, wasevaluated with the highest score, 3.61 in all as well aswith the highest score of 3.35 in wearing evaluation forsensation.

캠브리지 대학 고고인류학 박물관 소장 한복유물에 관한 연구

김순영(Soon Young Kim) 한국복식학회 2011 服飾 Vol.61 No.9

This paper investigates the formative features and the historical meaning of the Korean dress collection held in the Museum of Archaeology and Anthropology in Cambridge. The Museum holds a total of eleven items of Koran dresses that were donated by two anthropologists in the first half of the twentieth century. Male dresses consist of a white cotton jacket (Jeogori), a under- vest (Deungbaeja) made of light wisteria rings, a headband (Manggeon) made of horsehair, a broad brimmed top hat (Gat) made of black horsehair gauze, an oilskin cover (Galmo) drawn over the hat in wet weather, and a hemispherical hat box (Gatjib). Female dresses comprise a pink silk jacket (Jeogori), a blue silk skirt (Chima) with pleats, a pair of woman`s white cotton trousers (Sokgot), a black silk cap (Jobawi) decorated with pink tassels and imitation pearls, and a pair of green and magenta silk shoes (Danghye) with leather soles and metal rivets. These Korean dresses show what the western anthropologists had interests in. When collectors collect the folk objects, they thought much of the specificity of shape and material, the esthetic appreciation, and the representation of daily life. In terms of the value as the historical materials in the history of Korean dress, the under-vest of wisteria, the hat box, and the female dresses are worth paying attention to. The under-vest is one that was produced in earlier time among the remaining under-vests. The hat box represents that the hat belonged to the merchant classes. The female dress items show daily dresses worn by women of higher classes of the society in the 1920s.

형태학적 필터 및 사영왜곡 보정을 이용한 컨테이너 영상의 앞/뒷면 판별 방법

서정현 ( Jung Hyun Seo ),이종극 ( Jong Keuk Lee ),최재각 ( Jae Gark Choi ) 한국화상학회 2010 한국화상학회지 Vol.16 No.3

본 논문에서는 컨테이너 영상의 앞/뒷면을 판별하는 알고리즘을 제안한다. 제안 방법에서는 컨테이너 뒷면 손잡이 부분의 존재 유무를 앞/뒷면 영상의 판별기준으로 정하고, 손잡이 영역이 컨테이너 표면 배경보다 밝다는 가정 하에 형태학적 필터를 사용하여 손잡이 영역만을 추출한다. 그리고 컨테이너 영상의 손잡이 영역의 밝기를 수직으로 누적하여 피크를 찾고 피크의 크기와 피크 간의 거리를 이용하여 컨테이너 영상의 앞/뒷면을 판별하였다. 많은 다양한 컨테이너 영상에 대한 실험결과, 제안된 방법이 우수한 판별 성능을 나타내었다. This paper proposes a decision algorithm of front/back side container images. The proposed method uses the handle bar in the back side of container as decision criterion and extracts the handle bar area using morphological filters under the assumption which the handle bar is brighter than the background of container. Then the intensity peaks are calculated by accumulating the intensity of the handle bar vertically. The algorithm decides the front/back side by the height of peaks and the distances between the peaks. As shown in experiments results, the proposed method shows the excellent decision performance.