基于改进逆滤波的红外图像目标信息快速复原研究

Research on rapidrestoration of infrared image target information based on improved inverse filtering

针对当前复原红外图像目标信息时,在去噪过程中无法保留红外图像的边缘信息,导致复原后的红外图像QI值低、光谱信息不完整和均方误差大,存在复原效果差等问题。提出一种基于改进逆滤波的红外图像目标信息快速复原研究,建立红外图像退化模型,利用退化模型建立红外图像目标信息,提高红外图像质量;采用保边缘规整化技术对红外图像进行处理,在保留红外图像完整的同时有效去噪;利用小波变化高频去噪低频去模糊的原理对红外图像进行处理,利用逆滤波法得到红外图像各频段的正则化参数,利用获取的正则化参数对红外图像目标信息进行修复,以此来复原红外图像。实验结果表明,保边缘规整化方法QI值可达到0.06,光谱信息完整和均方误差在0.01,复原后的图像质量提高,更贴近原始图像。

When the target information of the infrared image is restored,the edge information of the infrared image cannot be retained in the process of de-noising,resulting in the low QI value of the restored infrared image,incomplete spectral information,large mean square error and poor restoration effect.In this paper,a rapid restoration of infrared image target information based on improved inverse filtering is proposed,and a degradation model of infrared image is established.The degradation model is used to build infrared image target information and improve infrared image quality.The infrared image is processed by using edge preserving structuring technology,which can effectively de-noise while preserving the integrity of the infrared image.The infrared image is processed by using the principle of wavelet change,high frequency and de-noising,low frequency and de-blurring.The regularization parameters of each frequency band of the infrared image are obtained by using the inverse filtering method.The regularization parameters obtained are used to repair the target information of the infrared image so as to reconstruct the original infrared image.The experimental results show that the method of edge preserving normalization has high QI value,complete spectral information and small mean square error,and the quality of the restored image is improved and closer to the original image.