基于显著性物体检测和S-PCG算法的彩色全息显示系统  

Color holographic display system based on significant object detection and S-PCG algorithm

作　　者：赵宇 卜静雯 陈舒涵[1] 张涛 ZHAO Yu;BU Jingwen;CHEN Shuhan;ZHANG Tao(School of Information Engineering,Yangzhou University,Yangzhou 225127,China)

机构地区：[1]扬州大学信息工程学院,江苏扬州225127

出　　处：《扬州大学学报（自然科学版）》2023年第6期57-64,共8页Journal of Yangzhou University：Natural Science Edition

基　　金：国家自然科学基金资助项目(62205283);江苏省自然科学青年基金资助项目(BK20200921);江苏省高等学校自然科学基金资助项目(20KJB510024,20KJB520007).

摘　　要：针对全息三维显示系统图像生成速度慢、重建精度低以及波前调制器件调制性能受限等问题,设计了一种面向真实物体的彩色全息三维实时显示系统.首先,利用深度相机采集真实物体的三维信息;其次,在采集端构建三维显著物体检测模型,采用显著性检测滤除冗余背景信息,获得准确的点云信息;最后,利用分割型点云网格化(segmented point cloud gridding,S-PCG)算法快速生成全息图,完成彩色全息三维实时显示系统的构建.实验结果表明,与传统方法相比,所提方法可有效降低全息系统的计算复杂度,提升采集端实时图像处理速度和重建精度.To solve the problems of slow image generation,low reconstruction accuracy and limited modulation performance of wavefront modulator,a color holographic 3D real-time display system for real objects is designed.Firstly,the depth camera is used to collect three-dimensional information from real objects.Secondly,a 3D significant object detection model is constructed at the acquisition end,and the redundant background information is filtered by significance detection to obtain accurate point cloud information.Finally,holograms are quickly generated using segmented point cloud gridding algorithm to complete the construction of color holographic 3D real-time display system.The experimental results show that compared with the traditional methods,the proposed method can effectively reduce the computational complexity and improve the real-time image processing speed and reconstruction accuracy.

关 键 词：计算全息 彩色全息术 三维图像处理 实际物体 全息系统 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] TN26[自动化与计算机技术—计算机科学与技术]