• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2011.07a
• /
• pp.14-15
• /
• 2011

본 논문은 TFT-LCD 디스플레이를 위한 저전력 화질 개선 기법을 제안한다. 제안하는 기법에서는 TFT-LCD 디스플레이의 어두워진 백라이트를 보상하기 위한 기법을 히스토그램 균등화에 기반하여 유도하며, 밝기 보상으로 인하여 손실되는 정보량이 최소가 되게 하는 변환 함수를 구한다. 컴퓨터 모의실험을 통해 제안하는 알고리듬이 전력 소비를 줄이는 동시에 영상의 화질을 개선하는 것을 확인한다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2014.11a
• /
• pp.158-159
• /
• 2014

본 논문에서는 대표 히스토그램을 이용한 효율적인 저전력 화질 개선 기법을 제안한다. 제안하는 기법은 우선 히스토그램으로 이미지를 분류하고, 각 분류된 이미지를 대표하는 히스토그램들을 저전력 화질 개선 기법이 적용된 전달 함수(transformation function)를 통해 미리 구한다. 따라서 매번 이미지마다 처리되어야 하는 복잡한 계산과정을 대표 히스토그램을 이용하여 미리 구함으로써, 기존 기법 대비 빠르고 효율적인 시스템을 구현할 수 있다. 실험을 통하여 제안 기법이 기존 기법에 비하여 낮은 전력 소모와 빠른 처리 속도를 보임을 확인한다.

• Radiation Oncology Journal
• /
• v.18 no.4
• /
• pp.345-354
• /
• 2000

Purpose : Conventional radiation therapy Portal images gives low contrast images. The purpose of this study was to enhance image contrast of a linacgram by developing a low-cost image processing method. Materials and Methods : Chest linacgram was obtained by irradiating humanoid Phantom and scanned using Diagnostic-Pro scanner for image processing. Several types of scan method were used in scanning. These include optical density scan, histogram equalized scan, linear histogram based scan, linear histogram independent scan, linear optical density scan, logarithmic scan, and power square root scan. The histogram distribution of the scanned images were plotted and the ranges of the gray scale were compared among various scan types. The scanned images were then transformed to the gray window by pallette fitting method and the contrast of the reprocessed portal images were evaluated for image improvement. Portal images of patients were also taken at various anatomic sites and the images were processed by Gray Scale Expansion (GSE) method. The patient images were analyzed to examine the feasibility of using the GSE technique in clinic. Results :The histogram distribution showed that minimum and maximum gray scale ranges of 3192 and 21940 were obtained when the image was scanned using logarithmic method and square root method, respectively. Out of 256 gray scale, only 7 to 30$\%$ of the steps were used. After expanding the gray scale to full range, contrast of the portal images were improved. Experiment peformed with patient image showed that improved identification of organs were achieved by GSE in portal images of knee joint, head and neck, lung, and pelvis. Conclusion :Phantom study demonstrated that the GSE technique improved image contrast of a linacgram. This indicates that the decrease in image quality resulting from the dual exposure, could be improved by expanding the gray scale. As a result, the improved technique will make it possible to compare the digitally reconstructed radiographs (DRR) and simulation image for evaluating the patient positioning error.