偏振光谱图像去雾技术研究

发布时间：2018-01-09 20:03

本文关键词：偏振光谱图像去雾技术研究　出处：《光谱学与光谱分析》2017年08期 　论文类型：期刊论文

【摘要】：空气污染严重影响了成像质量,雾霾天气下成像的对比度和分辨率下降。利用不同波段偏振光谱数据的图像去雾研究国内外尚未见报道;运用偏振光谱去雾技术对光谱数据立方体研究具有重要意义。在中重度雾霾天气下,利用光谱仪,获取了380~1 000nm的可见近红外偏振光谱数据,研究了450,550,650,750和850nm五个波长下偏振图像和去雾结果。结果表明,波长对图像目标与背景的灰度差值影响显著,去雾后图像的灰度动态范围和平滑度都有显著提升;去雾后图像远景对比度平均提升了5倍,其中450nm图像对比度最低,但对比度提升量最大,提高了7.13倍;850nm图像对比度最高,但对比度提升量最小,提高了3.86倍;去雾后不同波长图像的灰度动态范围均有明显展宽,展宽范围为13.5%~28.6%,而图像近景动态范围展宽为33.3%~44.0%。分析了图像复原过程中可能出现的估计误差,根据其产生机理提出了对大气散射光的偏振度和无穷远处的大气散射光强度的校正方法,并总结出了最佳校正系数的选取规律,对未知条件下的图像去雾提出指导。通过斯托克斯参数从原始数据中提取偏振信息,基于大气散射光与景物透射光偏振特性差异对图像去雾,偏振光谱图像去雾技术为光谱图像去雾研究提供一种新手段,也为图像去雾技术的发展提供一种新思路。
[Abstract]:Air pollution has seriously affected the imaging quality, and the contrast and resolution of the imaging are decreased in haze weather. The study of image defog using polarizing spectrum data of different bands has not been reported at home and abroad. It is very important to study the spectral data cube by using the polarizing spectrum de-fogging technique. In the moderate and severe haze weather, the visible near infrared polarized spectral data of 380 ~ 1 000 nm are obtained by using the spectrometer. The results of polarization image and defogging at 450,550,650,750 and 850 nm are studied. The results show that the wavelength has a significant effect on the gray difference between the target and the background. The dynamic range of gray level and smoothness of the image are improved significantly after fog removal. After fog removal, the contrast of the image is increased by an average of 5 times, and the contrast of 450 nm image is the lowest, but the contrast enhancement is the largest, which is 7.13 times. The contrast of 850nm image is the highest, but the contrast enhancement is the smallest, which increases 3.86 times. After fog removal, the dynamic range of gray level of different wavelength images is obviously broadened, and the range is 13.5- 28.6%. And the dynamic range of image close-range is 33.3 / 44.0. the possible estimation errors in the process of image restoration are analyzed. According to its generation mechanism, a correction method for the degree of polarization of atmospheric scattered light and the intensity of atmospheric scattering light at infinity is proposed, and the rule of selecting the best correction coefficient is summarized. The polarization information is extracted from the original data by Stokes parameters, and the image is de-fogged based on the difference of polarization characteristics between the atmospheric scattering light and the transmitted light of the scene. Polarizing spectral image de-fogging technology provides a new method for the research of spectral image de-fog, and also provides a new way of thinking for the development of image de-fogging technology.
【作者单位】：中国科学院西安光学精密机械研究所中国科学院光谱成像技术重点实验室;中国科学院大学;
【基金】：国家自然科学基金项目(11327303,61405239,61501456)资助
【分类号】：O436.3;TP391.41
【正文快照】： 2.中国科学院大学,北京100049引言目前严重的大气污染导致雾霾问题不断加重[1-2]。雾霾天气下,大气溶胶浓度显著提升[3],气溶胶粒子的散射会使景物光发生退化,严重影响了成像的对比度和分辨率,普通成像手段无法在雾霾天气下获取清晰图像,给信息提取和解析带来很大困难。偏振是

【相似文献】