• 정보처리학회논문지A
• /
• 제11A권5호
• /
• pp.391-400
• /
• 2004

서로 다른 종류의 영상을 정확하게 연관시켜 복합적인 정보를 제공하는 다중모달리티 의료 영상정합기법 중 표면정보 기반 영상정합에서는 일반적으로 동일 대상에 대한 서로 다른 모달리티에서 추출된 표면 윤곽정보 사이의 거리를 최소화함으로써 매칭이 이루어진다. 그런데 동일대상에 대해 취득되는 서로 다른 두 모달리티는 관심 영역 상의 표면 특성이 서로 유사하다. 그러므로 다중모달리티 영상정합에서 표면거리와 함께 표면의 형태 특성을 고려하여 두 영상을 매칭하는 방법이 정합결과의 정확도를 향상시킬 수 있다. 본 연구에서는 동일 대상의 서로 다른 두 모달리티 뇌영상 간의 표면거리와 표면곡률을 최적화하는 정합기법을 제안한다. 영상정합은 참조영상과 테스트영상에 대한 표면정보 생성과 이 두 개의 표면정보를 최적화하는 단계로 구성된다. 표면정보 생성 단계에서는 두 모달리티로부터 관심영역의 윤곽선을 추출하고, 이 중 참조 볼륨의 윤곽선에 대해서는 표면거리맵과 표면곡률맵을 구성하게 된다. 최적화 단계에서는 표면거리맵과 표면곡률맵을 참조하는 최적화 평가함수(cost function)에 의해 두 객체의 표면거리 차이와 표면곡률 차이를 최소화하는 정합 변환 값이 결정되고, 이것이 테스트영상의 변환에 적용되어 결과적으로 두 영상이 정합 되게 된다. 제안된 최적화 평가함수는 표면거리 정보만을 사용하는 평가함수에 비해 보다 견고한 정합 정확도를 보였으며 또한 본 연구는 정합결과의 볼륨 가시화를 통해 효율적인 영상 분석 수단을 제공하고자 하였다.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2005년도 추계종합학술대회
• /
• pp.481-484
• /
• 2005

Inherent opportunities on research for restoring high resolution image from low resolution images are increasing in these days. Super resolution image reconstruction is the process of combining multiple low resolution images to form a higher resolution one. To achieve super resolution reconstruction, proper observation model which is based on subpixel shift information is required. In this context, the importance of the subpixel registration cannot be estimated because subpixel shift information cannot be obtained from original image. This paper presents a regularized adaptive super resolution reconstruction method based on phase correlated subpixel registration, where the Constrained Least Squares(CLS) Restoration is adopted as a post process.

• Journal of Korean Neurosurgical Society
• /
• 제29권1호
• /
• pp.72-77
• /
• 2000

The authors describe the experience with the interactive image-guided Zeiss SMN system, which has been applied to 20 patients with various intracranial lesions during one year. Preoperative radiologic evaluation was CT scan in 6 cases, MRI in 14 cases. In all except one case, average fiducial registration errors were less than 2mm. There was no statistical difference in registration error between CT and MR image. This system considered to be relatively stable with respect to soft and hardware. Also it was useful for the designing of the scalp incision and bone flap and assessing the extent of resection in tumors, especially in gliomas. Moreover, it was helpful to evaluate complex surgical anatomy in skull base surgery.

• 융합신호처리학회논문지
• /
• 제16권1호
• /
• pp.9-14
• /
• 2015

Outlier detection and removal is a crucial step needed for various image processing applications such as image registration. Random Sample Consensus (RANSAC) is known to be the best algorithm so far for the outlier detection and removal. However RANSAC requires a cosiderable computation time. To drastically reduce the computation time while preserving the comparable quality, a outlier detection and removal method based on modified K-means is proposed. The original K-means was conducted first for matching point pairs and then cluster merging and member exclusion step are performed in the modification step. We applied the methods to various images with highly repetitive patterns under several geometric distortions and obtained successful results. We compared the proposed method with RANSAC and showed that the proposed method runs 3~10 times faster than RANSAC.

• 대한전자공학회논문지
• /
• 제24권1호
• /
• pp.35-45
• /
• 1987

For the registration problem in which the matching of two images is made, a new algorithm using an 1-D LPC model was proposed. The proposed algorithm employed LPC coefficients as feature vector of an image. The similarity of two images was measured using an LPC distance, proposed by Itakura, between each image's feature vector. The comparision of performance with normalized correlation method and template matching method was made by a computer simulation with several real images. The results of simulation showed that the proposed algorithm was more robust to image intensity variation and computationall efficient.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2003년도 학술회의 논문집 정보 및 제어부문 A
• /
• pp.191-194
• /
• 2003

In this paper, we propose a new accurate and robust image mosaic technique of the mural underdrawing taken from the infra-red camera, which is based on multiple image registration and adaptive blending technique. The image mosaicing methods which have been developed so far have the following deficits. It is hard to generate a high resolution image when there are regions that do not have features or intensity gradients, and there is a trade-off in overlapping region site in view of registration and blending. We consider these issues as follows. First, in order to mosaic Images with neither noticeable features nor intensity gradients, we use a Projected supplementary pattern and pseudo color image for features in the image Pieces which are registered. Second, we search the overlapping region size with minimum blending error between two adjacent images and then apply blending technique to minimum error overlapping region. Finally, we could find our proposed method is more effective and efficient for image mosaicing than conventional mosaic techniques and also is more adequate for the application of infra-red mural underdrawing mosaicing. Experimental results show the accuracy and robustness of the algorithm.

• 대한전자공학회논문지SP
• /
• 제42권4호
• /
• pp.1-12
• /
• 2005

영상정합은 동일한 장면에 대해서 서로 다른 시간 혹은 서로 다른 특성의 센서로부터 서로 다른 위치에서 얻은 영상들의 위치적 대응관계를 찾는 기법이다. 이 논문에서는 특성이 다른 적외선 센서와 광학 센서로부터 얻은 영상의 정합을 위한 새로운 알고리즘을 제안한다. 지금까지 제안된 서로 다른 특성의 영상을 위한 정합기법은 크게 특징점 기반 영상정합기법과 밝기값 기반 영상정합기법으로 구분될 수 있다. 특징점 기반의 영상정합기법은 정확하게 대응하는 특징점을 선택하는 것이 성능에 결정적인 영향을 준다 그러나 적외선 영상과 가시광선 영상에서는 특징점이 서로 같지 않은 경우가 많기 때문에 강인하지 못하다 그리고 밝기 값 기반의 정합기법에서는 정규상호정보를 유사성 척도로 사용한 영상정합기법이 가장 좋은 성능을 제공하는 것으로 알려져 있다. 그러나 정규상호정보 기반의 영상정합기법은 두 영상의 통계적 상관성이 전역적이어야 한다는 가정을 전제하는데, 적외선 영상과 가시광선 영상에서는 이를 보장하지 못하는 경우가 많아 정규상호정보를 유사성 척도로 사용하는 영상정합기법에서도 좋은 성능을 기대하기 힐들다. 따라서 이 논문에서는 적외선 영상과 가시광선 영상의 통계적 상관성의 해석에 기반한 두 단계 영상정합기법을 제안한다. 정확하고 강인한 정합을 위해서 첫 단계에서는 두 영상에서 통계적 상관성이 높은 부분을 추출하는 ESCR기법과 두 영상을 통계적 상관성이 높도록 필터링하는 ESCF기법을 수행한다. 그리고 두 번째 단계에서는 첫 단계에서의 결과 영상에 대해서 정규상호정보를 유사성 척도로 한 영상정합을 수행한다. 다양한 적외선 영상과 가시광선 영상을 이용한 실험으로부터 제안하는 두 단계 영상정합기법이 기존의 정규상호정보 기반의 영상정합기법에 비해 정확도와 강인함, 그리고 실행 속도의 측면에서 더욱 향상된 성능을 제공함을 확인하였다.

• IEIE Transactions on Smart Processing and Computing
• /
• 제4권5호
• /
• pp.311-317
• /
• 2015

The aim of this study is to develop a 3D registration algorithm for positron emission tomography/computed tomography (PET/CT) and magnetic resonance (MR) images acquired from independent PET/CT and MR imaging systems. Combined PET/CT images provide anatomic and functional information, and MR images have high resolution for soft tissue. With the registration technique, the strengths of each modality image can be combined to achieve higher performance in diagnosis and radiotherapy planning. The proposed method consists of two stages: normalized mutual information (NMI)-based global matching and independent component analysis (ICA)-based refinement. In global matching, the field of view of the CT and MR images are adjusted to the same size in the preprocessing step. Then, the target image is geometrically transformed, and the similarities between the two images are measured with NMI. The optimization step updates the transformation parameters to efficiently find the best matched parameter set. In the refinement stage, ICA planes from the windowed image slices are extracted and the similarity between the images is measured to determine the transformation parameters of the control points. B-spline. based freeform deformation is performed for the geometric transformation. The results show good agreement between PET/CT and MR images.

• 한국측량학회:학술대회논문집
• /
• 한국측량학회 2006년도 춘계학술발표회 논문집
• /
• pp.181-185
• /
• 2006

Image co-registration is the process of overlaying two images of the same scene, one of which represents a reference image, while the other is geometrically transformed to the one. In order to improve efficiency and effectiveness of the co-registration approach, the author proposed a pre-qualified area matching algorithm which is composed of feature extraction with canny operator and area matching algorithm with cross correlation coefficient. For refining matching points, outlier detection using studentized residual was used and iteratively removes outliers at the level of three standard deviation. Throughout the pre-qualification and the refining processes, the computation time was significantly improved and the registration accuracy is enhanced. A prototype of the proposed algorithm was implemented and the performance test of 3 Landsat images of Korea showed: (1) average RMSE error of the approach was 0.436 Pixel (2) the average number of matching points was over 38,475 (3) the average processing time was 489 seconds per image with a regular workstation equipped with a 3 GHz Intel Pentium 4 CPU and 1 Gbytes Ram. The proposed approach achieved robustness, full automation, and time efficiency.

• 대한의용생체공학회:의공학회지
• /
• 제31권2호
• /
• pp.134-140
• /
• 2010

Readily available high resolution brain MRI scans allow detailed visualization of the brain structures. Researchers have focused on developing methods to quantify shape differences specific to diseased scans. We have developed a novel method to quantify shape information for a specific population based on Multidimensional scaling(MDS). MDS is a well known tool in statistics and here we apply this classical tool to quantify shape change. Distance measures are required in MDS which are computed from pair-wise image registrations of the training set. Registration step establishes spatial correspondence among scans so that they can be compared in the same spatial framework. One benefit of our method is that it is quite robust to errors in registrations. Applying our method to 13 brain MRI showed clear separation between normal and diseased (Cushing's syndrome). Intentionally perturbing the image registration results did not significantly affect the separability of two clusters. We have developed a novel method to quantify shape based on MDS, which is robust to image mis-registration.