采取平场校正的方法消除薄型背照式电荷耦合器件(CCD)相机在对视网膜细胞成像时由于多光束干涉效应在输出图像上产生了干涉条纹。在分析CCD芯片多光束干涉效应产生机理的基础上，根据视网膜细胞成像的特点，提出将未进行像差补偿的视网膜图像作为平场校正的参考图像对像差补偿之后的图像进行校正，解决了标准反射率白板图像作为平场校正参考图像时残留有干涉条纹的问题。通过计算平场校正前后视网膜细胞图像的平均功率谱验证消除干涉条纹的效果，在空间频率70~90 lp/(°)范围内，校正之后图像的功率谱的平均值比校正之前提高了132.1%。实验结果表明，利用未进行像差补偿的眼底图像作为平场校正参考图像可以有效地消除视网膜细胞图像上的干涉条纹，提高视觉细胞的对比度，所获得的图像质量有了明显的改善。

A flat-field correction method is proposed to eliminate the interference fringes on the retinal photoreceptor images which origin from the etaloning effect in the thinned back illuminated CCD. First of all, the generation mechanism of etaloning effect is analyzed. Different from other flat-field method which uses the image of white balance reflectance target as reference image, the retinal image without adaptive optics compensation is used as reference image for flat-field correction method in this experiment, and the calibration result using the new reference image has a significant improvement compared with the result calibrated using the white balance reflectance target as reference image. As a means to quantify the calibrated result of flat-field method, the average power spectra of the photoreceptor image is calculated. The mean value of the average power spectra for the image after calibration is improvedly 132.1% compared with the image before calibration in the range of spatial frequency 70 and 90 lp/(°). Experimental result indicates that the image of retinal photoreceptor has a significant improvement using flat-field correction method.