摘要
真三维显示技术将图像每个三维体素空间位置、亮度和色彩进行真实还原,这样就可以建立起三维空间图像。但进行智能交互过程中会存在图像畸变问题。文章针对该问题提出一种新型的计算机视觉算法,通过ARM芯片构建起计算机视觉系统,利用高速摄像机、定向散射屏来进行高速投影,再结合CNN模型来进行深度生产,对投影时产生的畸形图像进行矫正处理,可以达到较高分辨率和较宽的视角,具有很高的应用价值。
The true three-dimensional display technology truly restores the spatial position,brightness and color of each three-dimensional voxel of the image,so that a three-dimensional image can be established.However,there will be image distortion problems in the process of intelligent interaction.This paper proposes a new type of computer vision algorithm to solve this problem.A computer vision system is built using ARM chips.It uses high-speed cameras and directional scattering screens to perform highspeed projection,and then combines the CNN model to perform deep production and correct the deformities generated during projection.Image correction processing can achieve higher resolution and wider viewing angle,which has high application value.
作者
叶一帆
YE Yifan(Hunan Institute of Technology,Hengyang Hunan,421002)
出处
《长江信息通信》
2021年第10期73-75,共3页
Changjiang Information & Communications
关键词
计算机技术
视觉算法
图像处理技术
computer technology
Visual algorithm
Image processing technology
作者简介
叶一帆(1993-),湖南湘乡人,硕士,计算机学院专任教师,研究领域:计算机视觉及人工智能。
