同态滤波的一种GPU提速实现方法
A Method of Using GPU to Accelerate Homomorphic Filter
-
摘要: 为了提高光照不均图像的增强速率,提出了基于GPU平台的同态滤波并行算法.根据同态滤波算法的并行性,利用CUDA软硬件体系架构,实现了同态滤波算法向GPU上的移植.利用多幅不同分辨率图像作为测试数据,对比CPU和GPU方案的计算效率.实验结果表明,GPU实现方案大幅度提升了计算效率.Abstract: In order to speed up the enhancement of non-uniform illumination image, a parallel homomorphic filter method based on GPU is presented.According to the parallelity of the algorithm, employing the CUDA programming framework, homomorphic filter based on GPU was implemented.Using various resolution images as test dataset, compared the computational efficiency of CPU and GPU.The experiment result shows that GPU can effectively implement the fast algorithm.
下载:
