基于星点亚像元误差补偿的大视场红外相机在轨检焦方法

On-Orbit Refocusing Method for Wide-Angle Infrared Camera Based on Star Point Sub-Pixel Error Compensation

针对大视场红外相机在轨调焦时受大气条件、地物丰富度、月相角等因素影响导致对地调焦和对月调焦周期长、效率低的问题,文章提出一种基于星点亚像元误差补偿的红外相机在轨检焦方法,通过对未经大气退化的星点图像进行精准质心定位提取清晰度评价指标,有效规避外界因素的影响,提升在轨检焦效率,保障在轨检焦精度。首先对星点图像分别采用自适应阈值质心法和高斯拟合法提取质心估计值;依据图像信噪比、星点目标能量集中度确定最佳因子,构建误差补偿模型进行精确质心定位;再提取图像点扩散函数的波形半高宽(Full Width at Half Maximum,FWHM)作为在轨检焦清晰度评价指标;设置不同谱段、不同能量集中度的星点红外图像作为在轨检焦图像,分别采用质心法、高斯拟合法及星点亚像元误差补偿方法进行目标质心提取并估计FWHM,实验结果表明:三种方法的质心提取平均误差分别为0.1195、0.0107、0.0027,均方根误差分别为0.1210、0.0124、0.0085,星点亚像元误差补偿方法质心提取误差最小,稳定性最好。能量集中度为0.4~0.8之间时,采用星点亚像元误差补偿方法质心提取平均误差均小于0.01,优于其他方法。在此基础上,采用该方法提取的FWHM平均精度提升了三倍以上,且不受星点质心位置随机性的影响,对于实现基于恒星的在轨检焦具有较高的可靠性和稳定性,满足在轨检焦要求。

A method for wide-angle infrared camera on-orbit refocusing based on sub-pixel error compensation of star points is proposed to solve the problems of long refocusing cycles and low efficiency caused by atmospheric conditions,land richness,moon phase angle,and other factors.By accurately locating the centroid of star point images without atmospheric degradation and extracting clarity evaluation indicators,the method effectively avoids the influence of external factors,improves the refocusing efficiency,and ensures the refocusing accuracy.Firstly,adaptive threshold centroid method and Gaussian fitting method are used to extract centroid estimation values for star point images;According to image signal-to-noise ratio and energy concentration of star point targets to determine the optimal factor,and construct an error compensation model for accurate centroid localization;Extract the waveform full width at half maximum(FWHM)of the point spread function from the image as an evaluation index for clarity;Set up infrared images of star points with different spectral ranges and energy concentrations as on-orbit focus detection images,and use centroid method,Gaussian fitting method,and the method proposed in this paper to extract target centroids and estimate FWHM.The experimental results show that the average errors of centroid extraction for the three methods are 0.1195,0.0107,and 0.0027,and the root mean square errors(RMSE)are 0.1210,0.0124,and 0.0085,respectively.The method proposed in this paper has the smallest centroid extraction error and the best stability.When the energy concentration is between 0.4-0.8,the average error of centroid extraction using this method is less than 0.01,which is superior to other methods.On this basis,the FWHM average accuracy extracted by the method proposed in this article has been improved by more than three times,and is not affected by the randomness of the centroid position of star points.It has high reliability and stability for refocusing,meeting the requirements of on-orbit refocusing.