论文部分内容阅读
稀疏分解在图像处理中应用的关键障碍之一是图像稀疏分解速度十分缓慢。针对这一问题,提出了一种新的图像稀疏分解的匹配追踪快速算法。研究了图像稀疏分解中使用的原子的能量分布特性,根据原子能量的分布特性,图像匹配追踪中的绝大部分的计算可以省略,因而极大地提高了图像匹配追踪的计算速度。实验结果表明,新的算法比传统的图像匹配追踪算法速度提高了许多倍,而恢复图像的质量没有任何的降低。
Sparse Decomposition One of the key obstacles in the application of image processing is the very slow rate of image sparse decomposition. Aiming at this problem, a new fast matching pursuit algorithm for image sparse decomposition is proposed. The energy distribution characteristics of atoms used in image sparse decomposition are studied. According to the distribution of atomic energy, the majority of image matching pursuit can be omitted, which greatly improves the computational speed of image matching pursuit. The experimental results show that the new algorithm has many times faster speed than the traditional image matching pursuit algorithm, and the quality of the restored image has not been reduced.