基于特征光线选择的多光场相机采样优化方法

Approach to optimize the sampling of multi-light field camera system based on feature rays selection

为满足多光场相机测量系统快速、准确地重建火焰三维温度场的需求,针对火焰辐射光场采样具有低秩性和方向冗余性的问题,本文提出了一种基于特征光线选择的光场采样优化方法。该方法根据辐射光线穿行火焰网格的顺序以及入射角度的特点,通过选取特征光线来表征原始光场采样,以降低火焰辐射光场采样间的方向冗余性。采用数值计算对该方法的抗噪性和适用性进行了分析,进一步对乙烯层流双峰扩散火焰开展了实验验证研究。结果表明,该采样优化方法能够良好地适用于多光场相机系统,当向光线辐射强度添加不同水平的噪声时,与运用原始采样进行重建相比,利用优化采样不仅能够准确地重建火焰三维温度场,而且能够减小部分网格的重建误差,重建时间也缩短为利用原始采样重建时间的1/14。同时运用该方法可以有效地重建乙烯层流双峰扩散火焰三维温度场,重建的时间分辨率和空间分辨率均有大幅度提升。

It is necessary to quickly and accurately reconstruct the three-dimensional flame temperature field by the multi-light field camera system.However,the sampling information collected by the light field camera is usually low-rank and most of them are redundant in direction.In this paper,a novel approach was proposed to optimize the sampling of the multi-light field camera system based on the feature rays selection.According to the distribution of ray and angle characteristics,the proposed approach performed rays selection to obtain representative feature rays,which reduced the light field sampling redundancy.The anti-noise ability and applicability of the proposed approach were evaluated through numerical simulations,and the ethylene diffusion bimodal flame temperature was reconstructed experimentally.Results showed that the proposed approach can be well applied in the multi-light field camera system.The optimized sampling not only reconstructed the flame temperature successfully even with different noise levels,but reduced reconstruction errors in partial voxels.The reconstruction time was shorter for the optimized sampling which was 1/14 of the original sampling.Meanwhile,the proposed approach also could reconstruct the ethylene diffusion bimodal flame temperature with high spatial and temporal resolution.