The present invention uses a Bayesian approach based particle filter to predict the lung contour of the next slice using the results of the lung contour of successive previous slices, predict the lung contour of the next slice using the predicted contour again, And a lung contour extracting method and an extracting apparatus using a particle filter capable of detecting lung nodules and lung tumors located on the outer wall.In the present invention, the particles of the lung contour obtained from the n-1 frame of the DICOM image obtained through the Chest-CT in the lung contour extracting apparatus are denoted by (KN-1), The number of particles of the lung contour obtained from the n frame image through the expansion and contraction model (NPar) In the lung contour extracting device, the lung image of n frame is directly used to extract pulmonary form (ONA primary extraction step of a lung type for extracting a primary lung In the lung contour extracting apparatus, ON, An n frame image prediction function (FNA particle weight calculation and normalization step of calculating a particle weight of the particle and performing normalization In the lung contour extracting apparatus, ONAnd the standard deviation () using the Gaussian functionQ) And each ONA weight calculation step of calculating a weight of particles of the particle And removing a particle having a value of a weight ( ) Less than a predetermined value in the particles extracted primarily from the lung form in the lung contour extracting apparatus to form the contour shape of the final lung.본 발명은 Bayesian approach 기반의 파티클 필터를 이용하여 연속된 이전 슬라이스의 폐 윤곽선 결과를 이용하여, 다음 슬라이스의 폐 윤곽선을 예측하고, 다시 예측된 윤곽선을 이용하여 다음 슬라이스의 폐윤곽선을 예측하고, 폐 외벽에 자리 잡고 있는 폐 결절 및 폐 종양을 검출할 수 있는 파티클 필터를 이용한 폐 윤곽선 추출방법 및 추출장치를 제공하는 데 있다.본 발명은, 폐 윤곽선 추출장치에서 Chest-CT를 통해 얻은 DICOM영상의 n-1 frame에서 얻은 폐 윤곽선의 파티클을(Kn-1)을 이용하여, 확장 및 축소모델을 통한 n frame 이미지에서 얻은 폐 윤곽선의 파티클 수(Npar)를 예측하는 파티클 생성단계와 폐 윤곽선 추출장치에서 n frame 의 폐영상을 이용하여 직접 폐 형태(On)를 1차적으로 추출하는 폐형태의 1차적
