CAJ | 학술논문

调制传递函数补偿(Modulation Transfer Function Compensation，MTFC)处理可以恢复光学影像在成像、传输和处理过程中损失的高频信息，使得影像边缘更为清晰、细节更为丰富，从而改善影像质量。然而，随着遥感对地观测技术的不断发展，高空间分辨率影像数据量大，导致复原处理时计算量大、耗时长，成为高分辨率地面处理系统实时处理的瓶颈问题。为解决这一问题，提出了基于统一设备架构模型的MTFC并行处理方法，并根据算法特点对存储结构和并行计算结构进行了优化改进，利用真实卫星影像进行了验证实验。实验结果表明，在保证影像复原质量的前提下，提出的算法获得了45倍的加速比，可以满足地面系统实时处理的要求。
조제전체함수보상(Modulation Transfer Function Compensation，MTFC)처리가이회복광학영상재성상、전수화처리과정중손실적고빈신식，사득영상변연경위청석、세절경위봉부，종이개선영상질량。연이，수착요감대지관측기술적불단발전，고공간분변솔영상수거량대，도치복원처리시계산량대、모시장，성위고분변솔지면처리계통실시처리적병경문제。위해결저일문제，제출료기우통일설비가구모형적MTFC병행처리방법，병근거산법특점대존저결구화병행계산결구진행료우화개진，이용진실위성영상진행료험증실험。실험결과표명，재보증영상복원질량적전제하，제출적산법획득료45배적가속비，가이만족지면계통실시처리적요구。
The modulation transfer function compensation (MTFC) processing can restore the lost high-frequency information of the optical image in the course of imaging, transmission and processing, making images clearer and richer detail edge, thereby improving image quality. However, owing to the large volume of data of high-resolution images, the image restoration generates huge computational load and is time-consuming. To solve this problem, this paper proposes and implements a parallel MTFC algorithm based on the compute unified device architecture (CUDA) model. The experimental results show that under the premise that the qual-ity of image restoration is assured, the speedup can reach 45 times, and processing efficiency can be promoted greatly.