Yan Junliang, Kong Lingsheng, Diao Zhihui, Liu Xiaofeng, Zhu Lilu, Jia Ping
Appl Opt. 2018 Jan 20;57(3):396-403. doi: 10.1364/AO.57.000396.
Panoramic stereo images, captured by distributed devices then mosaicking, are competent contents for virtual reality applications. Mosaicking raw images with different perspectives into satisfying final results is still not efficient enough, even if state-of-the-art algorithms are employed. For improving this efficiency in optical methods, we delve into the potential of the capturing system. Two parallax factors, peak parallax and deviation of parallaxes, are proposed to assess the mosaicking capability. By controlling variables and numerical computation, rules between parallax factors and design parameters have been revealed. Validation by simulations, large capturing distance, more cameras, compact arrangement, and moderate overlaps are suggested as the general design strategy. Benefiting from efficient mosaicking, systems based on our design strategy would have potential for real-time applications.
由分布式设备捕获然后拼接而成的全景立体图像，是虚拟现实应用的合适内容。即使采用了最先进的算法，将具有不同视角的原始图像拼接成令人满意的最终结果仍然不够高效。为了提高光学方法的这种效率，我们深入研究了捕获系统的潜力。提出了两个视差因子，即峰值视差和视差偏差，以评估拼接能力。通过控制变量和数值计算，揭示了视差因子与设计参数之间的规律。通过仿真验证，建议采用大捕获距离、更多相机、紧凑排列和适度重叠作为一般设计策略。受益于高效拼接，基于我们设计策略的系统将具有实时应用的潜力。
