基于红外数字全息技术的透水雾全息成像研究

【作者】 王刚；

【导师】 张永安；

【作者基本信息】 昆明理工大学 ， 光学， 2024， 硕士

【摘要】 浓雾天气是造成道路交通事故的主要原因之一,每年造成大量的人员与财产损失,目前对于浓雾天气形成的低能见度现象尚未有良好的克服方法。红外全息技术具有较可见光更长的波长,可以在衍射过程中绕过空气中的小颗粒,使其不易受到吸收和散射,具有穿透雾气的能力、检测大视场物体和更强的抗干扰特性,对于浓雾天气的克服体现出其独特的优势。相较于红外光数字全息,一般数字全息技术同样具有三维成像、高精度和强实时性的特点,但由于其通常使用可见光作为相干光源,在水雾和烟雾等恶劣环境中,可见光的成像效果可能会受到严重影响,使得难以有效监控被水雾阻挡的物体。因此与可见光数字全息相比,红外数字全息技术更适用于浓雾天气的水雾环境。由此提出了一种结合红外光可以穿透雾气的特性与全息技术的水雾环境全息成像方法,该方法通过模拟雾气环境,探索不同雾浓度对红外光和可见光全息成像的影响,证明了红外数字全息技术透水雾全息成像的可行性与优越性。同时为提高透水雾全息重建像的质量,扩大该技术适用场景,还研究了透水雾红外全息重建像的图像增强技术。主要的工作内容与创新如下:(1)通过实验方法验证红外光对水雾具有衍射透过的能力,根据数字全息的技术特点,选择合适的实验器材并设计实验方案,搭建透射型与反射型全息实验光路,进行改进与优化达到条件允许范围下的最好状态,并采集透水雾的全息图。(2)通过实验探究红外与可见光全息重建像与水雾之间的浓度关系,使用无水雾环境下的全息重建图像作为比较的参考,观察相位变化,并计算不同浓度下的图像评估参数,证明在相同的水雾浓度下,红外光恢复图像的强度衰减仍然较小,能够更好地保持图像的原始结构。(3)进行透水雾全息重建像的图像增强,首先对重建像进行频域滤波方法的处理,使得重建图像的其他级次被有效消除;再对图像背景中的散斑噪声进行增强Lee滤波,与其他去噪算法相比,增强Lee滤波去除了大部分散斑噪声同时保留了相当多的细节;最后对重建像边缘进行改进的Canny算法提取,效果较传统算法得到有效提升。更多还原

【Abstract】 Dense fog weather is one of the main causes of road traffic accidents,leading to significant personnel and property losses annually.Currently,there is no effective method to overcome the low visibility caused by dense fog weather.Infrared holography,with its longer wavelength compared to visible light,can bypass small particles in the air during diffraction,making it less susceptible to absorption and scattering.This technology possesses the ability to penetrate fog,detect large field-of-view objects,and exhibits stronger anti-interference characteristics,demonstrating its unique advantages in overcoming dense fog weather conditions.Compared to infrared digital holography,traditional digital holography also possesses the characteristics of three-dimensional imaging,high precision,and strong real-time performance.However,due to its reliance on visible light as a coherent light source,its imaging effect can be severely affected in harsh environments such as water mist and smoke,making it difficult to effectively monitor objects obstructed by water mist.Therefore,infrared digital holography is more suitable for water mist environments encountered during dense fog weather.Based on this,a holographic imaging method for water mist environments is proposed,combining the ability of infrared light to penetrate fog with holography.This method explores the influence of different fog concentrations on infrared and visible light holographic imaging through simulated fog environments,demonstrating the feasibility and superiority of infrared digital holography for holographic imaging through water mist.To further improve the quality of holographic reconstruction images through water mist and expand the applicability of this technology,image enhancement techniques for infrared holographic reconstruction images through water mist are also studied.The main work content and innovations are as follows:(1)Experimental methods are used to verify the ability of infrared light to diffract and penetrate water mist.Based on the technical characteristics of digital holography,suitable experimental equipment is selected,and experimental schemes are designed.Transmission-type and reflection-type holographic experimental optical paths are set up,optimized,and adjusted to achieve the best state within the conditions allowed.Holograms through water mist are then collected.(2)The relationship between infrared and visible light holographic reconstruction images and water mist concentration is explored experimentally.The holographic reconstruction images in a non-foggy environment are used as a reference for comparison.Phase changes are observed,and image evaluation parameters under different concentrations are calculated.It is demonstrated that under the same water mist concentration,the intensity attenuation of images recovered by infrared light is still relatively small,better preserving the original structure of the image.(3)Image enhancement is performed on holographic reconstruction images through water mist.First,the reconstructed images are processed using frequency domain filtering methods,effectively eliminating other orders of the reconstructed images.Then,speckle noise in the image background is enhanced using Lee filtering.Compared with other denoising algorithms,enhanced Lee filtering removes most of the speckle noise while preserving considerable detail.Finally,an improved Canny algorithm is used to extract the edges of the reconstructed image,achieving better results than traditional algorithms.更多还原