• Proceedings of the Korea Information Processing Society Conference
• /
• 2021.05a
• /
• pp.335-337
• /
• 2021

안저 영상 촬영기술이 발달되며 진단에 사용되는 안저 영상에는 시각적으로 많은 변화가 일어났다. 새로운 촬영 기법인 초광각 안저 영상은 기존 영상에 비해 넓은 범위의 영상을 생성할 수 있다. 촬영 범위가 넓어짐에 따라 이미지에는 왜곡이 발생하고, 이로 인해 안저 영상을 통한 황반 부위 진단에 어려움을 야기하기도 한다. 본 논문에서는 이러한 왜곡을 보정하고 초광각 안저 영상을 기존 안저 영상의 영역으로 변환하는 시스템을 강화학습을 통해 구축한다. 제안하는 방법은 A3C 강화학습법을 사용하며 실험 결과는 제안 방법을 통해 안저 영상을 자동으로 변환할 수 있음을 보여준다.

• Journal of the Institute of Convergence Signal Processing
• /
• v.3 no.1
• /
• pp.41-47
• /
• 2002

Fractal image compression can reduce the size of image data by the contractive mapping that is affine transformation to find the block(called as range block) which is the most similar to the original image. Even though fractal image compression is regarded as an efficient way to reduce the data size, it has high distortion rate and requires long encoding time. In this paper, we presented a hybrid fractal image compression system with the modified SOFM Vector Quantizer which uses improved competitive learning method. The simulation results showed that the VQ hybrid fractal using improved competitive loaming SOFM has better distortion rate than the VQ hybrid fractal using normal SOFM.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.37 no.4
• /
• pp.267-277
• /
• 2019

Photogrammetry and computer vision are identical in determining the three-dimensional coordinates of images taken with a camera, but the two fields are not directly compatible with each other due to differences in camera lens distortion modeling methods and camera coordinate systems. In general, data processing of drone images is performed by bundle block adjustments using computer vision-based software, and then the plotting of the image is performed by photogrammetry-based software for mapping. In this case, we are faced with the problem of converting the model of camera lens distortions into the formula used in photogrammetry. Therefore, this study described the differences between the coordinate systems and lens distortion models used in photogrammetry and computer vision, and proposed a methodology for converting them. In order to verify the conversion formula of the camera lens distortion models, first, lens distortions were added to the virtual coordinates without lens distortions by using the computer vision-based lens distortion models. Then, the distortion coefficients were determined using photogrammetry-based lens distortion models, and the lens distortions were removed from the photo coordinates and compared with the virtual coordinates without the original distortions. The results showed that the root mean square distance was good within 0.5 pixels. In addition, epipolar images were generated to determine the accuracy by applying lens distortion coefficients for photogrammetry. The calculated root mean square error of y-parallax was found to be within 0.3 pixels.

• Journal of IKEEE
• /
• v.13 no.2
• /
• pp.224-231
• /
• 2009

Unsharp masking is a popular image enhancement technique used to sharpen an image appearance in gray images. However, the conventional unsharp making techniques amplify the noise and easily cause overshoot artifacts. Moreover, the unsharp masking tends to introduce color distortion when it is applied to the each color component independently. To solve these problems, we propose a novel unsharp masking technique based on human visual system and vector projection. The proposed algorithm consists of two steps. First, the proposed algorithm controls the level of sharpening by exploiting the characteristics of the human visual system and contrast region. Then the vector projection is applied to remove the color distortion. Experiment results show that our proposed algorithm successfully produces sharpened images that are free of noise and color distortion commonly found in the conventional unsharp masking algorithms.

• Ecology and Resilient Infrastructure
• /
• v.8 no.1
• /
• pp.1-8
• /
• 2021

A surface image velocimeter (SIV) measures the velocity of a particle group by calculating the intensity distribution of the particle group in two consecutive images of the water surface using a cross-correlation method. Therefore, to increase the accuracy of the flow velocity calculated by a SIV, it is important to accurately calculate the displacement of the particle group in the images. In other words, the change in the physical distance of the particle group in the two images to be analyzed must be accurately calculated. In the image of an actual river taken using a camera, camera lens distortion inevitably occurs, which affects the displacement calculation in the image. In this study, we analyzed the effect of camera lens distortion on the displacement calculation using a dense and uniformly spaced grid board. The results showed that the camera lens distortion gradually increased in the radial direction from the center of the image. The displacement calculation error reached 8.10% at the outer edge of the image and was within 5% at the center of the image. In the future, camera lens distortion correction can be applied to improve the accuracy of river surface flow rate measurements.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.49 no.4
• /
• pp.16-22
• /
• 2012

In this paper, we present spiral pattern which suits for optical simulator to calibrate fisheye lens and compensate geometric distortion. Using spiral pattern, we present calibration without mathematical modeling in advance. Proposed spiral pattern used to input image of optical simulator. Using fisheye lens image, we calibrate a fisheye lens by matching geometrically moved dots to corresponding original dots which leads not to need mathematical modeling. Proposed algorithm calibrates using dot matching which matches spiral pattern image dot to distorted image dot. And this algorithm does not need modeling in advance so it is effective. Proposed algorithm is enabled at processing of pattern recognition which has to get the exact information using fisheye lens for digital zooming. And this makes possible at compensation of geometric distortion and calibration of fisheye lens image applying in various image processing.

• Proceedings of the Korean Association of Geographic Inforamtion Studies Conference
• /
• 2003.04a
• /
• pp.575-580
• /
• 2003

KOMPSAT-1 위성의 EOC영상은 위성에서 지구를 촬영하는 동안 발생하는 영상 왜곡을 포함하고 있다. 본 연구는 EOC영상의 영상왜곡을 보정하기 위하여 수치지도를 이용하는 정밀기하보정에 대하여 연구한다. 정밀기하보정 과정은 수치지도와 EOC영상의 좌표계를 통합하는 과정을 거쳐 오버레이를 만들어 수치지도의 삼각점을 기준으로 위성영상에서 GCP를 선택하고, 이 GCP를 이용하여 위성 영상을 딜로니 삼각형들의 Mesh형태로 변환하여 모든 딜로니 삼각형을 리샘플링하는 과정을 거쳐 보정된 EOC영상을 얻는다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2008.06b
• /
• pp.236-239
• /
• 2008

본 논문에서는 영상의 중요한 정보를 유지하는 동시에 영상의 왜곡을 막으며, 영상의 크기를 줄이는 방법을 제시한다. 영상의 각 픽셀의 중요도를 측정 후, 가장 중요하지 않은 픽셀들을 제거하는데, 이 때 단순히 중요하지 않은 픽셀들을 제거하면, 이로 인해 영상의 왜곡이 생길 뿐 아니라 중요하지 않은 픽셀들이 가지고 있던 정황에 대한 정보도 잃게 된다. 이를 막기 위해 제시된 방법은 중요하지 않은 픽셀들을 제거한 후에, 그 픽셀들의 중요도를 주변으로 확산시키는 방법을 이용하여 영상의 왜곡을 막고, 영상의 정황을 보다 효과적으로 전달하며 영상 리타겟팅을 수행한다. 실험 결과는 제시된 방법의 효율성을 보여준다.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.25 no.1B
• /
• pp.156-166
• /
• 2000

A rate-distortion based segmentation using recursive merging is presented, which considers texture as a homogeneity by adopting the procedure of a generalized texture approximation. The texture in a region is approximated by SA-DCT and a set of two uniform quantizers with fixed step sizes, one for DC and another for AC. Using the approximated texture, we calculated the rate-distortion based cost. The segmentation using recursive merging is performed by using the rate-distortion based cost. Experimental results for 256$\times$256 Lena show that the region-based coding using the proposed segmentation yields the PSNR improvements of 0.8~ 1.0 dB and 1.2~1.5 dB over that using the rate-distortion based segmentation with DC approximation only and JPEG, respectively.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.26 no.3B
• /
• pp.295-301
• /
• 2001

전송과정에서 발생하는 왜곡량에 따라 전송률의 하한이 결정되는 율-왜곡 이론은 시간적으로 민감한 부분을 왜곡없이 부호화하여 전송하는 영상 시스템에서 기본이 되는 중요한 요소이다. 율-왜곡 이론은 정보량의 개념으로부터 시작되어 원 신호의 확률분포와 왜곡의 측정기준에 의해 결정되는데, 이 논문에서는 가우시안과 라플라시안 분포함수에 절대치 오차기준과 자승 오차기준을 적용하여 율-왜곡 함수를 각각 구하였다. 나아가서 저전송률 부호기로 개발된 H.263 부호기에 이 함수를 적용하여 분석하였다. 비교를 위해 자승 오차기준에 위한 이론치와 실제 측정치를 제시하였다. H.263 부호기는 엔트로피 부호화, 부호화를 블록 패턴 등 다양한 기법을 사용하여 율-왜곡 함수에 의한 이론치보다 주어진 MSE에서 정규화 비트율이 최대 0.55만큼 더 낮은 전송률을 얻을 수 있었다.