摘要
在进行图像超分辨率重建时,使用多幅图像比使用一幅图像能够得到更好的效果;但是目前基于多幅图像的超分辨率重建算法普遍存在重建速度慢、重建质量不够理想的问题。为此,构造了一种基于块对称对叠(PsyCo)的多幅图像超分辨率重建算法。首先对低分辨率图像序列进行基于ORB的图像配准,再对配准后的图像进行PSyCo重建,最后对重建的图像进行像素灰度最大值融合。实验结果表明,提出的重建算法具有更好的重建效果;并且具有较快的重建速度。
Super resolution reconstruction based on multiple images is the process of reconstructing a high- resolution image through multiple low-quality and low-resolution image sequences. However,there are many problems in the reconstruction algorithm based on multi-image resolution reconstruction. Aiming at these problems, a multiimage super-resolution reconstruction algorithm based on PSyCo was constructed. Firstly, the image registration based on ORB was performed on the low-resolution image sequence, and then formed on the image after registration. Finally, the reconstructed image was fused with the maximum gray value ofthe pixel.The experimental validation of algorithm shows the reconstruction quality is better than traditional algorithm and has a faster reconstruction speed.
作者
杨飚
邸苗
YANG Biao;DI Miao(Beijing Key Laboratory of Intelligent Control Technology of Urban Road Traffic,North China University of Technology,Beijing 100144, China)
出处
《科学技术与工程》
北大核心
2018年第12期225-228,共4页
Science Technology and Engineering
基金
国家自然科学基金(61374191)
北京市教育委员会科技计划(KM201710009001)资助
