Integrated imaging technology can complete reproduction of the real scene, however small angle has been one of the main bottlenecks of the technology. This project aims to focus on the development of research on the viewing angle of integral imaging based on the time-multiplexed principle. Firstly, RGB data and multidimensional data are obtained by using color depth camera and sparse depth camera array. Then, the concerned area is extracted by the image feature with matching point. After that, the unified coordinate system and depth image inversion are needed. Secondly, in accordance with the mapping relationship between the object pixels and the pixels of the image elements, the rapid synthesis method for elemental image array are studied by using RGB data and multidimensional depth data after preprocessing. Then, according to the required time-multiplexed angle, the mapping relation on the rapid synthesis of directional elements image array are constructed. Then, according to the directional elements image array mentioned above, weighted time-multiplexed display with the understanding of image content is finished for the better display with different target feature scenes, which are either the main contour or the level of detail .Through the research of the project, the perspective of integral imaging is expected to expand twice as much as the traditional view, and the rapid synthesis method for elemental image array and the weighted time-multiplexed mechanism also have important academic value and application value.
集成成像技术能够完成对真实三维场景的再现,然而视角过小一直是该技术的主要瓶颈之一。本项目拟重点开展基于时分复用原理的集成成像视角研究。首先,利用彩色相机和稀疏深度相机阵列获取观测目标RGB数据和多维深度数据,通过图像特征点匹配获取关注区域,之后统一坐标系并做深度图像反转。其次,采用预处理后的RGB数据和多维深度数据,依照物体像素点和元素图像像素之间的映射关系,研究元素图像阵列快速合成方法。接着,根据时分复用所需的角度重新构建偏离映射关系,快速合成定向元素图像阵列。然后,基于前面合成的定向元素图像阵列,结合对图像内容的理解,实现基于图像内容的加权时分复用显示,以更好显示以轮廓为主或以多细节层次为主的不同目标特征的场景。通过本项目研究,集成成像视角有望拓展为传统视角的两倍,涉及的元素图像阵列快速合成算法和加权时分复用机理研究具有重要的学术价值和应用价值。
本项目围绕基于时分复用原理的集成成像相关问题开展研究。首先,利用相机获取观测目标RGB数据和多维深度数据,通过图像特征点匹配获取关注区域。其次,采用预处理后的RGB数据和多维深度数据,依照物体像素点和元素图像像素之间的映射关系,合成元素图像阵列。然后,将两组不同视角的元素图像整列,采用时分复用的方式,通过实时视频传输系统输出。通过本项目的开展,搭建的数据采集平台可采集被测物体的深度信息和三维点云信息,通过不同角度的数据拼接实现了该物体的三维重建;提供了一种物体像素点与元素图像像素间光学映射关系,为计算集成成像提供重要依据;开发的实时视频传输系统可实现双图像源的时分复用显示,即可应用于左右拼接的裸眼立体显示,也可应用于图像元阵列的显示。本项目的开展为集成立体成像技术的进一步研究提供较为重要的理论和应用依据。
{{i.achievement_title}}
数据更新时间:2023-05-31
基于 Kronecker 压缩感知的宽带 MIMO 雷达高分辨三维成像
基于混合优化方法的大口径主镜设计
混采地震数据高效高精度分离处理方法研究进展
自组装短肽SciobioⅡ对关节软骨损伤修复过程的探究
不同分子分型乳腺癌的多模态超声特征和临床病理对照研究
正交时分复用技术研究
高速光时分复用网络中分插复用技术的研究
基于连续反转剥层方法的集成式光纤光栅时分/波分混合复用阵列时域重构技术研究
高速时分复用光通信技术