基于CUDA的图像轮廓提取并行实现

Parallel Achieve of Image Contour Extraction Based on CUDA

对图像轮廓提取进行并行实现,介绍基于形态水平集的图像轮廓提取算法,对该算法的并行可行性进行分析,并采用CUDA技术并行实现。与串行的方式实现比较,采用CUDA技术实现可以节省更多的时间,得出CUDA并行技术在需要大量数据计算时加速的有效性。

Mainly realizes the image contour extraction parallelly.Introduces the image contour extraction algorithm based on the morphological level set,and analyses the parallel feasibility of this algorithm and realizes parallelly using CUDA technology.Compared to the serial implementation,we know that the parallel algorithm cost less time.Using CUDA parallel technology will accelerate more effective when requiring lots of data computing.