基于CUDA的直升机旋翼桨叶挥舞角快速测量方法

针对基于立体视觉的直升机旋翼桨叶挥舞角测量CPU串行算法耗时多、效率不高的问题,利用图像处理单元(GPU)并行计算的优势,提出一种基于CUDA统一计算设备构架的并行处理快速算法.首先,对算法中最耗时的图像去噪、阈值分割、连通域标记三部分进行并行化设计;然后,采用多层次并行策略将大量密集运算分配到不同的图像处理单元上并行执行,利用共享内存和共享寄存器加速数据访问;最后,进行多次测量实验,结果表明该方法执行效率明显高于CPU串行方法,可满足旋翼桨叶挥舞角快速测量的要求.

Fast Measurement Method of Helicopter Rotor Blade Flapping Angle

In order to solve the time-consuming and low efficiency problem of helicopter rotor blade flapping angle measurement based on CPU algorithm,a fast parallel method is proposed by using the advantage of the GPU(graphics proceesing unit) parallel computing based on CUDA(compute unified device architecture).Firstly,the parallel algorithms of the three most time-consuming parts in the flapping angle measurement about image denoising,threshold segmentation,connected component labeling are studied.Secondly,a large number of intensive operations are parallelly implemented to different image processing units by using multi-level parallel strategies,using the shared memory and shared registers to accelerate access to data.Finally,The results of simulation tests show that the operative efficiency of the method is obviously improved compared with the CPU algorithm.It can meet the requirements of fast measurement for helicopter rotor flapping angle.