基于GPU的多尺度Retinex图像增强算法实现

为提高多尺度Retinex算法的实时性,本文提出了基于GPU的多尺度Retinex图像增强算法,通过对算法进行数据分析和并行性挖掘,将高斯滤波、卷积和对数差分等计算量非常耗时的模块放到GPU中,利用大规模并行线程处理来提高效率。在GeForce GTX 480和CUDA 5.5中进行实验,结果表明该算法能显著提高计算速度,且随着图像分辨率的增加,最大加速比达160倍。

Realization of multi scale Retinex image enhancementalgorithm based on GPU

To improve the real time performance of the multi scale Retinex algorithm, a GPU based multi scale Retinex image enhancement algorithm was proposed in this paper. Through the data analysis and parallel mining of the algorithm, time consuming modules of the calculation, such as Gauss filter, convolution, and logarithm difference, were implemented in GPU, and the efficiency was improved by using massively parallel processing threads. Experiments were conducted in GeForce GTX 480 and CUDA5.5, and the results showed that the proposed algorithm could significantly improve the computing speed, and with the increasing of the image resolution, the maximum speed up ratio could reach 160 times.