컬러비와 거리비를 이용한 교통표지판 영역추출

곽현욱,이우범,김욱현,Kwak, Hyun-Wook,Lee, Woo-Beom,Kim, Wook-Hyun 한국정보처리학회 2002 정보처리학회논문지B Vol.9 No.5

본 논문에서는 컬러 정보인 컬러비와 형태학적 정보인 코너점 및 거리비를 이용한 교통표지판 영역추출 시스템을 제안한다. R G, B 성분들의 컬러비를 기반으로 생성한 이진영상에 대해서 불림 연산처리와 마스크를 이용한 코너점 및 중심점 검출에 의해 후보영역을 추출한다. 다양한 형태의 교통표지판 영역은 후보영역의 중심점을 기준으로 형태학적 정보인 거리비를 이용하여 추출한다. 실험결과, 제안한 시스템은 크기 및 위치에 불변하며, 삼각형, 원, 역삼각형, 사각형 등의 다양한 교통표지판과 밝기값의 차이가 큰 낮과 밤 영상에 대해서도 정확한 영역추출이 가능하였다. 또한, 높은 정확성과 빠른 처리속도를 보였다. This paper proposes a system detecting the area of traffic sign, which uses color rate as the information of colors, and corner point and distance rate as the information of morphology. In this system, a candidate area is extracted by performing dilation operation on the binary image made by the color rate of R, G, B components and by detecting corner point and center point through mask. The area of traffic sign with varied shapes is extracted by calculating the distance rate from center point, which is the information of morphology. The results of this experiment demonstrate that in this system which is invariable regardless of its size and location, it is possible to extract the exact area from varied traffic signs such as the shapes of triangle, circle, inverse triangle, and square as well as from the images at both day and night when brightness value is greatly different. Moreover, it demonstrates great accuracy and speed in processing.

실영상에서 형태 정보와 에지 영상을 이용한 교통 표지판 영역 추출과 인식

곽현욱,오준택,김욱현,Kwak, Hyun-Wook,Oh,Jun-Taek,Kim, Wook-Hyun 한국정보처리학회 2004 정보처리학회논문지B Vol.11 No.2

본 논문은 실영상에서 형태 정보와 에지 영상을 이용한 교통 표지판 영역 추출 및 인식 방법을 제안한다. 화소의 RGB 색상비를 이용하여 생성한 이진 영상에서 connected component 알고리듬에 의해 분할된 후보 영역들을 대상으로 형태 정보인 XY축 대칭성을 기반으로 교통 표지판 영역을 추출한다. 만약 후보 영역이 검출되지 않을 경우, 히스토그램 평활화에 의해 대비를 향상함으로써 영역 추출이 가능하다. 그리고 교통 표지판 영역의 에지 영상에서 추출한 수평-수직 투영(XY projection). 모멘트(moment), 동심원형 패턴 및 8 방향 광선과 에지와의 거리 및 교차점의 개수 등의 형태 정보를 기반으로, 사전에 구축한 데이터베이스와의 유사도 측정에 의해 인식을 수행한다. 다양한 실영상을 대상으로 실험한 결과, 본 방법이 빛이나 날씨 조건 등의 외부 환경에 강건하게 추출 및 인식함을 보인다. This study proposes a method for segmentation and recognition of traffic signs using shape information and edge image in real image. It first segments traffic sign candidate regions by connected component algorithm from binary images, obtained by utilizing the RGB color ratio of each pixel in the image, and then extracts actual traffic signs based on their symmetries on X- and Y-axes. Histogram equalization is performed for unsegmented candidate regions caused by low contrast in the image. In the recognition stage, it utilizes shape information including projection profiles on X- and Y-axes, moment, and the number of crossings and distance which concentric circular patterns and 8-directional rays from region center intersects with edges of traffic signs. It finally performs recognition by measuring similarity with the templates in the database. It will be shown from several experimental results that the system is robust to environmental factors, such as light and weather condition.

뇌혈관용 Guiding catheter와 Penumbra를 이용하여 시행한 흉·복부 및 다리혈관에 대한 기계적 혈전제거술

곽현욱(Hyun-Wook Kwak),최대연(Dae-Yeon Choi),임동욱(Dong-Wook Lim) 대한인터벤션영상기술학회 2010 대한인터벤션영상기술학회지 Vol.13 No.1

Purpose More than 12 sorts of medical tools for mechanical thrombolysis have been created since 1989. medical tools for thrombolysis, AngioJet, Oasis & Arrow-Trerotola device etc are currently used in domestic hospitals. manual aspiration for thrombolysis have been performed with 50cc syringe in the angio department of PNUYH. Mechanical thrombolysis for abdomen & lower extremity by guiding catheter & penumbra for cerebrovascular only was on trials. the trials are going to be presented as the case of successful study. Materials and methods Case1 60 years old female diagnosed as Arteriosclerosis obliterans received mechanical thrombolysis in March, 2010. 8 hundred thousand Urokinase had been infused to the patient using Multi-sideport catheter (Cook, Bloomington, USA) in advance. Blocked lesion ranged from abdominal aorta at lumbar level to iliac artery. The abdominal aorta was aspirated by Penumbra pump (Penumbra, Inc. Alameda, USA) connected with 6Fr Envoy guiding catheter (Cordis, Florida, USA) in the method of mechanical thrombolysis. On the other hand, iliac artery was aspirated by either guiding catheter with a pump or guiding catheter with 0.014in * 0.035in (cone) * 200㎝ Penumbra separator 041 (Penumbra, Inc. Alameda, USA). Thrombus in lower extremity flowing from abdominal aorta thrombolysis was also aspirated by guiding catheter with separator 041. Total blood loss was 370cc. 8 hundred thousand Urokinase was infused into Popliteal artery through Multi-sideport catheter. Case2 33years old male, diagnosed as thrombosis in Internal jugular vein and Innominate vein took the procedure in April, 2010. 1million Urokinase was infused at Internal Medicine. The blocked lesion ranged from subclavian vein via brachiocephalic vein to SVC. The thrombus was eliminated in the way of guiding pumping and guiding & separator pumping side by side using 6Fr Envoy guiding catheter connected with Penumbra pump. The total blood loss was 370㏄. Percutaneous Transluminal Angioplasty was performed in the stenosis lesion after the thrombolysis. Result Case1 It took approximately 1~2min for aspiration from abdominal aorta to iliac artery. During the aspiration however, some departed thrombus flowed to lower extremity vessel and then aspiration for femoral artery and thrombolytic therapy for popliteal artery were carried out to remove the thrombus. Next day, opening from abdominal aorta to dorsal artery of foot could be confirmed on angiography. Moreover complications such as blood spill out of vessels and recurrence of thrombus were not observed. Case2 Mechanical thrombolysis by guiding catheter and penumbra was successful. It took about 15min for aspiration. The opening vessel was confirmed and complications were not found. the procedure was finished and then Percutaneous Transluminal Angioplasty for remaining occlusion lesion would be followed up due to recoil. Discussion Through the medical trial above there are several value-judgements bellow that can be studied and verified continuously. First, the matter of usefulness for guiding catheter and penumbra pump or separator & pump. Second, the matter of possibility to solve vessel occlusion caused from DVT by guiding catheter and penumbra pump only Third, the study of possibility to eliminate thrombus occurred in femorofemoral bypass graft. Forth, the study of possibility to remove thrombus caused by Arteriovenous anastomosis site as the case of using general 4Fr or 5Fr catheter. Fifth, the study of using Urokinase which has severe complication of bleeding Sixth, the study for removing thrombus using Penumbra Reperfusion catheter and Separation instead of percutaneous transluminal angioplasty when the case of acute thrombus located in the area of Knee joint occurs. Nevertheless it has different problems that Penumbra system is for cerebrovascular only. Conclusion There are several problems in this procedure. that the mechanical thr

Entropy 기반의 Weighted FCM 알고리즘을 이용한 컬러 영상 Multi-level thresholding

오준택,곽현욱,김욱현,Oh, Jun-Taek,Kwak, Hyun-Wook,Kim, Wook-Hyun 대한전자공학회 2005 電子工學會論文誌-SP (Signal processing) Vol.42 No.6

This paper proposes a multi-level thresholding method using weighted FCM(Fuzzy C-Means) algorithm in color image. FCM algerian determines a more optimal thresholding value than the existing methods and can extend to multi-level thresholding. But FCM algerian is sensitive to noise because it doesn't include spatial information. To solve the problem, we can remove noise by applying a weight based on entropy that is obtained from neighboring pixels to FCM algerian. And we determine the optimal cluster number by using within-class distance in code image based on the clustered pixels of each color component. In the experiments, we show that the proposed method is more tolerant to noise and is more superior than the existing methods. 본 논문은 weighted FCM(Fuzzy C-Means) 알고리즘을 적용한 컬러 영상 multi-level thresholding을 제안한다. FCM 알고리즘은 기존의 thresholding 방법들과 달리 최적의 임계치를 결정할 수 있으며 multi-level thresholding으로의 확장이 가능하다. 그러나 공간정보를 포함하고 있지 않기 때문에 잡음 등에 민감하다는 단점을 가진다. 본 논문은 이러한 단점을 해결하기 위해서 이웃 화소들로부터 얻은 entropy 기반의 가중치(weight)를 FCM 알고리즘에 적용함으로써 잡음의 제거가 가능하다. 그리고 각 색상별 성분의 군집 화소들을 기반으로 생성한 코드 영상에 대해서 군집 내부의 거리값을 이용하여 최적의 군집수를 결정한다. 실험에서 제안한 방법이 기존의 방법들보다 잡음에 대해서 강건하며 우수한 분할 성능을 보였다.

Texture-based Segmentation and Classification for Document Image

Bo-Ram Kim(김보람),Jun-Taek Oh(오준택),Hyun-Wook Kwak(곽현욱),Young-Ho Sohn(손영호),Wook-Hyun Kim(김욱현) 한국정보과학회 영남지부 2005 한국정보과학회 영남지부 학술발표논문집 Vol.13 No.1

In this paper, we propose a novel segmentation and classification method using texture features for the document image. First, the local entropy is extracted and the document image is segmented to separate the background and the foreground using the Otsu’s method. Then we classify the segmented regions into each components using principle component analysis(PCA) algorithm base on the texture features that extracted with the co-occurrence matrix in the entropy image. The entropy based segmentation process robust to not only noise and change of light, but also skew and rotation. Contrary to the gray value, texture features don't restrict the form of the document and is superior to the discrimination of components. And contrary to the neural network, the PCA algorithm that use for the classifier can classify components robustly.

실영상에서 형태 정보와 에지 영상을 이용한 교통 표지판 인식

오준택(Jun-Taek Oh),김보람(Bo-Ram Kim),곽현욱(Hyun-Wook Kwak),김욱현(Wook-Hyun Kim) 한국정보과학회 영남지부 2004 한국정보과학회 영남지부 학술발표논문집 Vol.12 No.1

본 논문은 실영상에서 형태 정보와 에지 영상을 이용한 교통 표지판 영역 추출 및 인식 방법을 제안한다. 화소의 RGB 색상비를 이용하여 생성한 이진 영상에서 connected component 알고리듬에 의해 분할된 후보 영역들을 대상으로 형태 정보인 XY축 대칭성을 기반으로 교통 표지판 영역을 추출한다. 그리고 추출된 교통 표지판 영역의 에지 영상에서 추출한 수평-수직 투영(XY projection), 모멘트(moment), 동심 원형 패턴 및 8방향 광선과 에지와의 거리 및 교차점의 개수 등의 형태 정보를 기반으로, 사전에 구축한 데이터베이스와의 유사도 측정에 의해 인식을 수행한다. 다양한 실영상을 대상으로 실험한 결과, 본 방법이 빛이나 날씨 조건 등의 외부 환경에 강건하게 추출 및 인식함을 보인다.