• Journal of the Korean Society of Radiology
• /
• v.17 no.3
• /
• pp.305-314
• /
• 2023

With the development of medical technology and radiation treatment equipment, the frequency of high-precision radiation therapy such as intensity modulation radiation therapy has increased. Image-guided radiation therapy has become essential for radiation therapy in precise and complex treatment plans. In particular, with the introduction of imaging equipment for diagnosis in a linear accelerator, CBCT scanning became possible, which made it possible to calibrate and correct the patient's posture through 3D images. Although more precise reproduction of the patient's posture has become possible, the exposure dose delivered to the patient during the image acquisition process cannot be ignored. Radiation optimization is necessary in the field of radiation therapy, and efforts to reduce exposure are necessary. However, when acquiring 3D CBCT images by changing the imaging conditions to reduce exposure, there should be no image quality or artefacts that would make it impossible to align the patient's position. In this study, Rando phantom was used to scan and evaluate images for each shooting condition. The highest SNR was obtained at 100 kV 80 mA 25 ms F1 filter 180°. As the tube voltage and tube current increased, the noise decreased, and the bowtie filter showed the optimal effect at high tube current. Based on the actual scanned images, it was confirmed that patient alignment was possible under all imaging conditions, and that image-guided radiation therapy for patient alignment was possible under the condition of 70 kV 10 mA 20 ms F0 filter 180°, which showed the lowest SNR. In this study, image evaluation was conducted according to the imaging conditions, and low tube voltage, tube current, and small rotation angle scan are expected to be effective in reducing radiation exposure. Based on this, the patient's exposure dose should be kept as low as possible during CBCT imaging.

• Journal of Internet Computing and Services
• /
• v.19 no.5
• /
• pp.43-54
• /
• 2018

Images taken in haze weather are characteristic of low contrast and poor visibility. The process of reconstructing clear-weather image from a hazy image is called dehazing. The main challenge of image dehazing is to estimate the transmission map or depth map for an input hazy image. In this paper, we propose a single image dehazing method by utilizing the Generative Adversarial Network(GAN) for accurate depth map estimation. The proposed GAN model is trained to learn a nonlinear mapping between the input hazy image and corresponding depth map. With the trained model, first the depth map of the input hazy image is estimated and used to compute the transmission map. Then a guided filter is utilized to preserve the important edge information of the hazy image, thus obtaining a refined transmission map. Finally, the haze-free image is recovered via atmospheric scattering model. Although the proposed GAN model is trained on synthetic indoor images, it can be applied to real hazy images. The experimental results demonstrate that the proposed method achieves superior dehazing results against the state-of-the-art algorithms on both the real hazy images and the synthetic hazy images, in terms of quantitative performance and visual performance.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.24 no.4
• /
• pp.372-378
• /
• 2014

Noise in biomedical X-ray image degrades the quality so that it might causes to decrease the accuracy of diagnosis. Especially the noise reduction techniques is quite essential for low-dose biomedical X-ray images obtained from low radiation power in order to protect patients, because their noise level is usually high to well discriminate objects. This paper proposes an efficient method to remove the noise in low-dose X-ray images while preserving the edges with diverse resolutions. In the proposed method, a noisy image is at first decomposed into several images with different resolutions in pyramidal representation, then the stable map of edge confidence is obtained from each of analyzed image using a fuzzy logic-based edge detector. This map is used to adaptively determine the parameter for guided filters, which eliminate the noise while preserving edges in the corresponding image. The filtered images in the pyramid are extended and synthesized into a resulted image using interpolation technique. The superiority of proposed method compared to the median, bilateral, and guided filters has been experimentally shown in terms of noise removal and edge preserving properties.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2018.11a
• /
• pp.104-107
• /
• 2018

최근 촬영 기기의 기술발전으로 인해 디지털 영상의 해상도가 증가함에 따라 선명한 디지털 영상에 대한 요구가 증가하고 있다. 이러한 요구에도 불구하고 디지털 영상 내 가우시안 잡음 (gaussian noise)은 촬영기기를 통해 영상 획득 및 처리 과정에서 발생하여 화질을 열화 시킨다. 디지털 이미지에서 발생하는 가우시안 잡음을 제거하기 위해서 기존의 저대역 통과 필터 (low-pass filter: LPF)를 사용하면 잡음은 제거되지만, 블러링 현상 (blurring phenomenon)이 나타난다. 이러한 문제점을 개선하기 위해 소벨 연산자 (sobel operator)를 사용하여 영상 내 에지 맵 (edge-map)을 생성하여 에지 영역과 동질 영역을 구분한다. 에지영역에서는 약한 저역 필터 (weak low-pass filter)를 사용하고, 그 외의 이미지 영역에서는 강한 저역 필터 (strong low-pass filter)를 사용하는 알고리듬을 제안하였다. 그리고 다양한 이미지에 대하여 기존 알고리듬과 제안한 알고리듬의 적용한 결과를 통해 주관적 화질 비교하였고 객관적 지표로 최대 신호 대 잡음비 (peak signal-to noise ratio: PSNR)와 구조 유사성 (structural similarity: SSIM)을 사용하여 성능을 평가하였다. 실험결과를 통해 제안된 알고리듬이 잡음 제거 및 외곽선 보존의 우수함을 확인하였다.

• Progress in Medical Physics
• /
• v.19 no.4
• /
• pp.209-218
• /
• 2008

According to improved radiation therapy technology such as IMRT and proton therapy, the accuracy of patient alignment system is more emphasized and IGRT is dominated research field in radiation oncology. We proposed to study the feasibility of cone-beam CT system using simple x-ray imaging systems for image guided proton therapy at National Cancer Center. 180 projection views ($2,304{\times}3,200$, 14 bit with 127 ${\mu}m$ pixel pitch) for the geometrical calibration phantom and humanoid phantoms (skull, abdomen) were acquired with $2^{\circ}$ step angle using x-ray imaging system of proton therapy gantry room ($360^{\circ}$ for 1 rotation). The geometrical calibration was performed for misalignments between the x-ray source and the flat-panel detector, such as distances and slanted angle using available algorithm. With the geometrically calibrated projection view, Feldkamp cone-beam algorithm using Ram-Lak filter was implemented for CBCT reconstruction images for skull and abdomen phantom. The distance from x-ray source to the gantry isocenter, the distance from the flat panel to the isocenter were calculated as 1,517.5 mm, 591.12 mm and the rotated angle of flat panel detector around x-ray beam axis was considered as $0.25^{\circ}$. It was observed that the blurring artifacts, originated from the rotation of the detector, in the reconstructed toomographs were significantly reduced after the geometrical calibration. The demonstrated CBCT images for the skull and abdomen phantoms are very promising. We performed the geometrical calibration of the large gantry rotation system with simple x-ray imaging devices for CBCT reconstruction. The CBCT system for proton therapy will be used as a main patient alignment system for image guided proton therapy.

• 한국지형공간정보학회:학술대회논문집
• /
• 1994.10a
• /
• pp.42-46
• /
• 1994

In this paper, we studied on an efficient detection and removal of the periodic scanner interference noise in LANDSAT images. Firstly, noise models and their characteristics are discussed. And we proposed a new scheme of noise detection in Fourier domain. Then, an dfficient noise filter can be designed based on the detected noise components. To verifythe effectiveness of our scheme, some experiments guided by our proposed scheme are performed using a real LANDSAT image.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2018.11a
• /
• pp.55-58
• /
• 2018

본 논문에서는 가이디드 영상 필터를 (guided image filter) 이용하여 컨볼루션 신경망 (convolutional neural network) 을 이용한 역 톤 매핑 (inver tone - mapping; iTMO) 기법의 결과를 향상 시킬 수 있는 알고리듬을 제안한다. 기존 low dynamic range (LDR ) 영상을 high dynamic range (HDR ) 디스플레이에서 표현할 수 있는 역 톤 매핑 기법이 과거부터 계속 제안되어 왔다. 최근에 컨볼루션 신경망을 이용하여 단일 LDR 영상만으로 넓은 동적 범위 (dynamic range) 를 가진 HDR 영상으로 변환하는 알고리듬이 많이 연구되었다. 기존의 알고리듬 중 포화 영역 (saturated region) 으로 인해 잃어버린 화소 정보를 학습된 컨볼루션 신경망을 이용해서 복원하는 알고리듬은 그 효과가 좋지만 포화 영역이 아닌 부분의 잡음을 제거하지 못하며 포화 영역의 디테일을 복원하지 못한다. 제안한 알고리듬은 입력 영상에 가중치 기반 가이디드 영상 필터를 사용해서 비포화 영역의 잡음을 제거하고 포화 영역의 디테일을 복원시킨 다음 컨볼루션 신경망에 인가하여 결과 영상의 품질을 개선하였다. 제안하는 알고리듬은 실험을 통해서 기존의 알고리듬에 비해 높은 정량적 화질 평가 지수를 나타내었고, 기존의 알고리듬에 비해 세부 사항을 효과적으로 복원할 수 있음을 확인할 수 있었다.

• Journal of Broadcast Engineering
• /
• v.21 no.5
• /
• pp.770-781
• /
• 2016

As a result of reducing color information and edge information, object distinction in haze image occurs with difficulty. One of the famous defogging algorithm is haze removal by using 'Dark Channel Prior(DCP)', which is used to predict for transmission rate using color information of an image and eliminates haze from the image. But, In case that haze rate is estimated under color information, there is a miscalculated issue which is posed by haze rate and transmission in area with high brightness such as a white object or a light source. In this paper, We deal with a miscalculated issue by correcting from around haze rate, after application of color normalization used by main white part of image haze. Moreover, We calculation improved transmission based on the result of improved haze rate estimation. And then haze image quality is developed through refining transmission.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.52 no.11
• /
• pp.105-114
• /
• 2015

This paper proposes methods to remove background area and segment region of interest (ROI) in chest X-ray images. Conventional algorithms to improve detail or contrast of images normally utilize brightness and frequency information. If we apply such algorithms to the entire images, we cannot obtain reliable visual quality due to unnecessary information such as background area. So, we propose two effective algorithms to remove background and segment ROI from the input X-ray images. First, the background removal algorithm analyzes the histogram distribution of the input X-ray image. Next, the initial background is estimated by a proper thresholding on histogram domain, and it is removed. Finally, the body contour or background area is refined by using a popular guided filter. On the other hand, the ROI, i.e., lung segmentation algorithm first determines an initial bounding box using the lung's inherent location information. Next, the main intensity value of the lung is computed by vertical cumulative sum within the initial bounding box. Then, probable outliers are removed by using a specific labeling and the pre-determined background information. Finally, a bounding box including lung is obtained. Simulation results show that the proposed background removal and ROI segmentation algorithms outperform the previous works.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.17 no.5
• /
• pp.48-54
• /
• 2018

With the 4th industry era, smart factories are emerging. In the era of multi-product small scale production, unmanned transportation vehicles are rapidly increasing in utilization of unmanned transportation vehicles that carry and arrange goods in the work space. The conventional unmanned vehicle detected its position by using the guided line method and the position based method for indoor location recognition and movement. This method has disadvantages of initial high cost and maintenance / maintenance. In this paper, to solve the disadvantages, the method of predicting the direct path of the unmanned vehicle through the Kalman filter is verified using the white LED marker of the warehouse and the position data and the image data of the white LED marker recognition image. Through this, the reliability of the linear movement which occupies the most part in the lattice structure is secured. It is also expected that the reliance on additional position sensors will also be reduced.