基于嵌入式平台的结构光检测算法的实现与优化

以采煤工作面运输皮带上的结构光检测算法为例,介绍基于海思Hi3516A嵌入式平台的视频分析算法的实现与优化方案.实验结果表明,基于PC实现的结构光检测算法移植到海思平台上,经过有针对性地优化,检测一帧高清视频所需时间从1000多ms减少到少于500 ms,提高了1倍多,为实现嵌入式平台上利用结构光中心线形变特征进行物体特征识别提供了有力的前提条件.

Realization and Optimization of Structure Light Detection Algorithm Based on Embedded Platform

Taking the structure light detection algorithm of the belt in the mine as the example,the realization and optimization of the intelligent video algorithm is introduced in the paper,which is based on the Hisilicon Hi3516A embedded platform.The experiment results show that the structure light detection algorithm which is based on PC is transplanted to Hisilicon platform,the detection time for a HD video frame is reduced from more than 1000 milliseconds to less than 500 milliseconds after the optimization.It provides powerful premise condition for the realization of the use of the center line of the structured light deformation features for object recognition on the embedded platform.