三波段彩色夜视分谱优化研究

Spectrum matching optimization study on tri-band color night vision technology

为了得到更符合人眼视觉特性的彩色夜视图像,依据RGB彩色理论模型,提出一种基于滤光片分谱作用的三波段彩色夜视方法。根据微光夜视系统成像理论,计算微光夜视系统的视距一般表达式。以典型场景作为探测目标,研究引入滤光片前后微光夜视系统的视距变化率,其中蓝光波段的滤光片引入使视距下降到引入前的15%。基于视距理论分析及色度学原理,提出了提高滤光片的透过率、拓宽滤光片的透过波段、融合图像加入全波图像的三个改进方法。以此制备了三片滤光片,并进行了彩色夜视实验。实验表明,提出的改进方法得到的彩色夜视图像在熵和标准方差两个指标上均优于原始微光图像和传统融合方法得到的结果,且与人眼主观感受一致。

In order to obtain color night vision images, which are more in line with thecharacteristics of the human eye, tri-band color night vision technology is put forward basedon RBG color model. The technique is the use of sub-spectral filter effect. Visual range iscalculated based on night vision imaging systems theory. The change rate of visual range,before and after the introduction of filter, is calculated in a typical scenario. The result showsthat visual range which is after the introduction of filter falls to 15% compared to before.Considering analysis above and color theory, three methods have been proposed to improve：Increasing the transmittance of the filter in the band needed; Broaden the band of the filter;Full-wave image fusion. Three filters are produced according to the improved methods.Experiment is taken by these three filters. Fused images which are experimentally obtained,are better than original low-light images by the objective evaluation, on the two indicators thatare entropy and average variance. The results are also consistent with the observed effect ofthe human eye. The experimental results show the methods are effective.