• Journal of Korea Multimedia Society
• /
• v.12 no.5
• /
• pp.633-643
• /
• 2009

We propose a color transfer method that used color contrast based templates to express the visual difference clearly between objects, while remaining the quality of the input image. Our algorithm employs colors of both the input image and template distributed on the $a^{\ast}b^{\ast}$chrominance plane of CIE $L^{\ast}a^{\ast}b^{\ast}$color space. The templates are made by considering the effect of color contrast and have the shape of either a line or a curve represented color distribution of the basic colors based gradation image. These tempates can be modeled on spline curves. We also generate simply new templates with the different basic colors by moving the control points of that curve. The color transfer method using the templates is done through a regressive analysis and color matching. We maintained color coherence of the input image by transforming similarly the color distribution of an input image to the one of templates.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.14 no.4
• /
• pp.663-670
• /
• 2019

The modified Canny edge detection algorithm that can detect the boundary between screen door and platform in the subway is proposed in this paper. Generally, in the subway, the boundary line between the platform and the screen door is darker than the surrounding area. Therefore, an edge image is using the modified bottom-hat transform by considering its characteristics. Double thresholded images with strong edge and weak edge through double thresholding are obtained. An algorithm that detects the boundary invasion between the platform and the screen door is proposed by calculating the length by applying the Hough transform to the double thresholded image and comparing the boundary line length between when there is an object such as a person and when there is no object. In this paper, the results of the proposed modified Canny edge detection algorithm using two different input images according to camera height position are shown by computer simulation.

• Journal of the Korean Society of Radiology
• /
• v.16 no.3
• /
• pp.295-302
• /
• 2022

Sparse-view computed tomography (CT) imaging technique is able to reduce radiation dose, ensure the uniformity of image characteristics among projections and suppress noise. However, the reconstructed images obtained by the sparse-view CT imaging technique suffer from severe artifacts, resulting in the distortion of image quality and internal structures. In this study, we proposed a convolutional neural network (CNN) with wavelet transformation and residual learning for reducing artifacts in sparse-view CT image, and the performance of the trained model was quantitatively analyzed. The CNN consisted of wavelet transformation, convolutional and inverse wavelet transformation layers, and input and output images were configured as sparse-view CT images and residual images, respectively. For training the CNN, the loss function was calculated by using mean squared error (MSE), and the Adam function was used as an optimizer. Result images were obtained by subtracting the residual images, which were predicted by the trained model, from sparse-view CT images. The quantitative accuracy of the result images were measured in terms of peak signal-to-noise ratio (PSNR) and structural similarity (SSIM). The results showed that the trained model is able to improve the spatial resolution of the result images as well as reduce artifacts in sparse-view CT images effectively. Also, the trained model increased the PSNR and SSIM by 8.18% and 19.71% in comparison to the imaging model trained without wavelet transformation and residual learning, respectively. Therefore, the imaging model proposed in this study can restore the image quality of sparse-view CT image by reducing artifacts, improving spatial resolution and quantitative accuracy.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• fall
• /
• pp.27-29
• /
• 2021

인공지능 기반 머신 비전 응용이 증가함에 따라 사람이 아닌 기계에서 소비되는 영상 정보를 전송하는 요구가 발생하고 있다. 일반적으로 영상 정보를 전송할 때는 전송 비용을 고려하여 정보를 압축하며 기존 영상 압축 방법은 사람의 시각 인지적 특성을 반영하여 설계되었다. 따라서 기존 영상 압축 방법은 기계에서 소비되는 영상 정보를 압축하는 방법으로 적절하지 않다고 판단하여 2019년 7월, 기계를 위한 영상 부호화 기술의 표준화가 시작되었다. 본 논문에서는 머신 비전 태스크 중, 객체 탐지를 수행하는 네트워크의 피처 맵을 압축하는 방법을 제안한다. 제안하는 방법은 피처 맵의 채널 간 중복성을 제거하기 위해 PCA 기반의 변환을 적용하여 피처 맵의 차원을 축소하며 특히 해상도 계층 구조를 갖는 네트워크의 피처 맵을 압축하기 위해 각 해상도 계층간 변환 기저를 예측하여 추가로 압축률을 높인다. 제안하는 방법을 적용하여 객체 탐지 결과의 큰 성능 하락 없이 약 92.3%에 데이터양 감소를 달성하였다.

• Proceedings of the IEEK Conference
• /
• 2003.07a
• /
• pp.426-429
• /
• 2003

본 논문에서는 네크？ traffic이 원활하지 못해 전체 영상 데이터의 전송이 불가능하더라도 일부 중요한 영상 정보만으로도 원 영상에 근접한 영상 복원을 할 수 있는 스트림의 구성을 연구 하였다. 서버측에서 획득한 영상을 웨이블릿 변환을 거쳐 프레임이 우선 순위에 따라 여러 패킷으로 나누어 UDP를 통해 전송하고 PDA에서는 일부 중요 데이터만을 이용하여 원 영상에 근접하는 영상을 얻을 수 있다.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2015.07a
• /
• pp.268-271
• /
• 2015

본 논문에서는 고해상도 비디오를 위한 프레임 레이트 변환 방법을 제시한다. 서로 다른 방식이나 포맷을 가진 영상 신호를 변환할 때에는 시각적으로 거슬리는 움직임 지터(motion jitter)나 영상의 모서리에서의 블러링(blurring) 현상이 일어날 수 있다. 제시하는 방법에서는 이러한 문제점을 해결하기 위해 양방향 움직임 추정/보상을 이용하여 프레임 레이트를 증가 변환한다. 또한 보다 정확한 움직임 벡터를 찾기 위해 다중 프레임을 통한 양방향 탐색 방법을 제안한다. 먼저, 픽셀 값을 알고 있는 현재 프레임과 이전 프레임간의 움직임 벡터를 이용하여 양방향 움직임 벡터를 얻는다. 얻은 움직임 벡터는 벡터 스무딩 블록을 거친다. 그 후 OBMC(Overlapped Block Motion Compensation)를 이용하여 보간 프레임을 구성한다. 실험 결과는 기존 방법들보다 제안 방법이 주관적 화질 면에서 우수함을 나타낸다.

• Journal of Digital Convergence
• /
• v.12 no.10
• /
• pp.329-335
• /
• 2014

With the spread of mobile phones and wireless networks, medical imaging needs to be accessible outside the hospital for patients. We develop the mobile PACS platform that allows doctors or patients to store, search, query and transfer medical images for their convenience. This system is a mobile based platform that patients DICOM images can be viewed from the hospital PACS server, then they can be converted according to smart standardization. For this, we designed the platform to transfer the data between PACS, mPACS and users. Also we suggest to use DICOM transformation processing to display DICOM images in PACS server.

• Journal of the Korea Society of Computer and Information
• /
• v.4 no.3
• /
• pp.73-77
• /
• 1999

In this paper. wavelet transform coding for image conference is studied. Original video frames are transformed into hierarchical pyramidal images with multiresolution using the band property of wavelet transform coefficients. Moving information between neighboring frames is obtained from the low-resolution band. Also, to control the video coding procedure. a new picture set filter is proposed. This filter controls the compression ratio of each frame depending on the correlation to the reference frame by selectively eliminating less important high-resolution areas. Consequently. video quality can be preserved and bit rate can be controlled adaptively In the simulation, to test the performance of the proposed coding method, comparisons with the full search block matching algorithm and the differential image coding algorithm are made. Consequently. the proposed method shows a reasonably good performance over existing ones.

• Journal of KIISE:Software and Applications
• /
• v.31 no.7
• /
• pp.939-948
• /
• 2004

In this paper. we Propose a robust and fast image registration technique for motion correction in brain CT-CT angiography obtained from same patient to be taken at different time. First, the feature points of two images are respectively extracted by 3D edge detection technique, and they are converted to locally weighted 3D distance map in reference image. Second, we search the optimal location whore the cross-correlation of two edges is maximized while floating image is transformed rigidly to reference image. This optimal location is determined when the maximum value of cross-correlation does't change any more and iterates over constant number. Finally, two images are registered at optimal location by transforming floating image. In the experiment, we evaluate an accuracy and robustness using artificial image and give a visual inspection using clinical brain CT-CT angiography dataset. Our proposed method shows that two images can be registered at optimal location without converging at local maximum location robustly and rapidly by using locally weighted 3D distance map, even though we use a few number of feature points in those images.

• Proceedings of the Korean Information Science Society Conference
• /
• 2010.06c
• /
• pp.469-473
• /
• 2010

본 논문에서는 조각 병리 영상을 강체 정합을 통해 하나의 영상으로 자동 스티칭하는 방법을 제안한다. 제안 방법은 영상의 위치 초기화, 위치 보정, 강체 정합의 세 단계로 이루어진다. 첫째, 영상의 위치 초기화 단계에서는 순서 없이 흩어진 부분 영상을 탬플릿 매칭 기법을 사용한 영상 내 문자 인식을 통해 위치를 초기화한다. 둘째, 강체 정합의 정확성을 높이기 위해 코너점을 이용해 부분 영상의 위치를 보정한다. 셋째, 조각 영상 간 거리를 최소화하는 강체 정합을 수행한다. 실험 결과, 부분 영상 간 간격이 최소화되어 하나의 영상으로 스티칭되는 것을 확인하였고, 최적화 반복 횟수와 변환 벡터에 따른 정확성, 견고성 평가를 통해 거리 차의 제곱 합이 최소화되어 수렴됨을 알 수 있었다. 본 논문의 제안 방법은 조각 영상을 하나의 영상으로 스티칭함으로써 병리 조직의 전체적인 구조 파악과 이를 이용한 전립선암 확진에 사용될 수 있다.