论文部分内容阅读
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。
Modulation Transfer Function Compensation (MTFC) processing can recover the high frequency information lost in the imaging, transmission and processing of optical images, so as to make the edges of the image clearer and the details more abundant so as to improve the image quality. However, with the continuous development of remote sensing earth observation technology, the large amount of data of high spatial resolution imagery leads to the computational complexity and time-consuming of restoration processing, and becomes the bottleneck of real-time processing of high-resolution ground processing system. In order to solve this problem, MTFC parallel processing method based on unified device architecture model is proposed. Based on the algorithm characteristics, the storage structure and parallel computing structure are optimized and improved, and the real satellite image is used to verify the experiment. The experimental results show that the proposed algorithm achieves a 45-fold speedup under the premise of ensuring the quality of image restoration, which can meet the real-time processing requirements of ground systems.