• Proceedings of the Korean Information Science Society Conference
• /
• 1999.10b
• /
• pp.509-511
• /
• 1999

본 논문에서는 일반적으로 영역 기반형 분할방법보다 우수한 분할결과와 계산의 효율성을 가지는 경계선 기반형 방법의 하나인 scan line approximation 방법을 응용함으로써 경계선의 기하학적 해석이 가능하도록 하는 경계선 강도(edge intensity) 정보를 제공한다. 따라서 면 특성과 국부적인 면 특성인 면 법선과 면 곡률정보 없이 잡음에 강건하고 계산의 효율성에서 우수한 거리영상분할 방법을 제안한다. 합성 거리영상을 대상으로 scan line approximation 방법을 응용하여 얻어진 경계선을 경계선 그룹화의 영역 레이블링을 거쳐서 면 특징을 추출하였다.

• Korean Journal of Radiology
• /
• v.22 no.10
• /
• pp.1597-1608
• /
• 2021

Cardiovascular computed tomography (CT) is among the most active fields with ongoing technical innovation related to image acquisition and analysis. Artificial intelligence can be incorporated into various clinical applications of cardiovascular CT, including imaging of the heart valves and coronary arteries, as well as imaging to evaluate myocardial function and congenital heart disease. This review summarizes the latest research on the application of deep learning to cardiovascular CT. The areas covered range from image quality improvement to automatic analysis of CT images, including methods such as calcium scoring, image segmentation, and coronary artery evaluation.

• Journal of KIISE:Software and Applications
• /
• v.33 no.1
• /
• pp.95-104
• /
• 2006

The Gradient Vector Flow (GVF) snake or active contour model offers the best performance for image segmentation. However, there are problems in classical snake models such as the limited capture range and the slow progress into concavity. This paper presents a new method for enhancing the performance of the GVF snake model by extending the external force fields from the neighboring fields and using a modified smoothing method to regularize them. The results on a simulated U-shaped image showed that the proposed method has larger capture range and makes it possible for the contour to progress into concavity more quickly compared with the conventional GVF snake model.

• Journal of the Korean Geophysical Society
• /
• v.3 no.1
• /
• pp.49-56
• /
• 2000

As a part of geophysical studies on segmentation of the Ulsan fault, walkaway refraction seismic data were measured at 17 stations near National Road 7 between Kyungju and Ulsan. Seismic anisotropy was analyzed in the offset range of 1-48 m. The average refraction velocity of 1787 m/s indicates the refractor is the upper boundary of weathered basement. P-wave anisotropy is computed to be 0.056 in average, which may serve as a weak evidence that the strike of major geologic structure coincide with the inferred fault direction. In the south of the province boundary between Kyungsangnam-do and Kyungsangbuk-do, the velocity anisotropy is normal in that P-wave velocity in the strike direction is faster than the one measured in the dip direction. On the contrary, it appears that the fault strikes in many directions or that fractures may be developed better in the dip direction in the northern par. Such a difference in anisotropic pattern is believed to be a seismic evidence indicating that a segmentation boundary of the Ulsan fault locates near the province boundary.

• The Journal of the Acoustical Society of Korea
• /
• v.19 no.1
• /
• pp.4-10
• /
• 2000

The work presented in this paper is about a postprocessor, which improves the performance of automatic speech segmentation system by correcting the phoneme boundary errors. We propose a postprocessor that reduces the range of errors in the auto labeled results that are ready to be used directly as synthesis unit. Starting from a baseline automatic segmentation system, our proposed postprocessor trains the features of hand labeled results using multi-layer perceptron(MLP) algorithm. Then, the auto labeled result combined with MLP postprocessor determines the new phoneme boundary. The details are as following. First, we select the feature sets of speech, based on the acoustic phonetic knowledge. And then we have adopted the MLP as pattern classifier because of its excellent nonlinear discrimination capability. Moreover, it is easy for MLP to reflect fully the various types of acoustic features appearing at the phoneme boundaries within a short time. At the last procedure, an appropriate feature set analyzed about each phonetic event is applied to our proposed postprocessor to compensate the phoneme boundary error. For phonetically rich sentences data, we have achieved 19.9 % improvement for the frame accuracy, comparing with the performance of plain automatic labeling system. Also, we could reduce the absolute error rate about 28.6%.

• Journal of Korean Society of Transportation
• /
• v.36 no.2
• /
• pp.129-140
• /
• 2018

In this study, we analyzed the electric vehicle (EV) purchasing behavior with market segmentation in Seoul using the binary logit model. For the model estimation, the experimental design of SP survey generated 24 scenarios with purchase price, charging station availability, and driving range of EVs. The results of market segmentation analysis indicated that the owners of subcompact and compact size cars were primarily affected by the purchasing price while those of mid and full size cars were more sensitive to the charging station availability. By housing type, the charging station availability was the most important factor for the residents of apartment while the purchase price was the most important factor to choose the EVs. These results presented that the EV supply strategy of the automobile manufacturer should be diversified according to the marketing target and the expansion of the public charging infrastructure should be the top priority in the government policy.

• Journal of Broadcast Engineering
• /
• v.5 no.1
• /
• pp.102-112
• /
• 2000

The previous video structuring techniques are mainly limited to shot or shot group level. However, the shot level structure couldn't provide semantics within a video. So, researches on high level structuring are going on for getting over the drawbacks of shot level structure, recently. To overcome the drawbacks of shot level structure, we propose video scene segmentation technique based on color and motion features. For considering various color distribution, each shot is divided into sub-shots based on color feature. A key frame is extracted from each sub-shot. The motion feature in a shot is extracted from MPEG-1 video's motion vector. Moreover adaptive weights based on motion's property in search range are applied to color and motion features. The experiment results of proposed technique show the excellence in view of the over-segmentation and the reflection of semantics, comparing with those of previous techniques. The proposed technique decomposes video into meaningful hierarchical structure and provides video browsing or retrieval based on scene.

• Journal of Biomedical Engineering Research
• /
• v.44 no.5
• /
• pp.346-353
• /
• 2023

In this paper, we propose an innovative approach that leverages deep learning to find optimal reference points for achieving precise tooth segmentation in three-dimensional tooth point cloud data. A dataset consisting of 350 aligned maxillary and mandibular cloud data was used as input, and both end coordinates of individual teeth were used as correct answers. A two-dimensional image was created by projecting the rendered point cloud data along the Z-axis, where an image of individual teeth was created using an object detection algorithm. The proposed algorithm is designed by adding various modules to the Unet model that allow effective learning of a narrow range, and detects both end points of the tooth using the generated tooth image. In the evaluation using DSC, Euclid distance, and MAE as indicators, we achieved superior performance compared to other Unet-based models. In future research, we will develop an algorithm to find the reference point of the point cloud by back-projecting the reference point detected in the image in three dimensions, and based on this, we will develop an algorithm to divide the teeth individually in the point cloud through image processing techniques.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.30 no.1
• /
• pp.49-58
• /
• 2012

Buildings become complex and diverse with time. It is difficult to extract individual buildings using only an optical image, because they have similar spectral characteristics to objects such as vegetation and roads. In this study, we propose a method to extract building area and boundary through integrating airborne Light Detection and Ranging(LiDAR) data and aerial images. Firstly, a binary edge map was generated using Edison edge detector after applying Adaptive dynamic range linear stretching radiometric enhancement algorithm to the aerial image. Secondly, building objects on airborne LiDAR data were extracted from normalized Digital Surface Model and aerial image. Then, a temporary building areas were extracted by overlaying the binary edge map and building objects extracted from LiDAR data. Finally, some building boundaries were additionally refined considering positional accuracy between LiDAR data and aerial image. The proposed method was applied to two experimental sites for validation. Through error matrix, F-measure, Jaccard coefficient, Yule coefficient, and Overall accuracy were calculated, and the values had a higher accuracy than 0.85.

• Proceedings of the Korea Contents Association Conference
• /
• 2006.05a
• /
• pp.401-404
• /
• 2006

Mean shift procedure is applied for the data points in the joint spatial-range domain and achieves a high quality. However, a color image is segmented differently according to the inputted spatial parameter or range parameter and the demerit is that the image is broken into many small regions in case of the small parameter. In this paper, to improve this demerit, we propose the method that groups similar regions using region merging method for over-segmented images. The proposed method converts a over-segmented image in RGB color space into in HSI color space and merges similar regions by hue information. Here, to preserve edge information, the proposed method use by merging constraints to decide whether regions is merged or not. After then, we merge the regions in RGB color space for non-processed regions in HSI color space. Experimental results show the superiority in region's segmentation results.