讨论了一种对小样本光子图像的统计处理方法。在超微弱发光的研究中 (例如细胞的超微弱荧光 ) ,由于发光强度极弱 ,需要用像增强器对超微弱发光图像进行增强得到可视图像 ,超微弱发光图像不可避免地受到像增强系统暗噪声及背景噪声的影响 ,使光子图像湮没在噪声中。为从原始图像中检验出信号 ,根据信号光子和噪声光子的不同统计分布 ,运用信号检测与评估的方法判断光子是否属于信号光子 ,并得到一简明的判据 ,由此判据剔除图像中的噪声光子 ,得到信噪比改善的光子图像。并用此方法处理了人手掌的超微弱发光光子图像 A method of statistical processing of photon images in small samples is discussed. In ultra-weak luminescence studies (such as ultra-weak fluorescence of cells), ultra-weak luminescence images are inevitably affected by image enhancement system Dark noise and background noise, photon images annihilate in the noise. In order to check out the signal from the original image, according to the different statistical distributions of signal photons and noise photons, the method of signal detection and evaluation is used to determine whether the photon belongs to the signal photon and a concise criterion is obtained, Photon noise, to improve the signal to noise ratio of the photon image. And this method has been used to deal with ultra-weak photoluminescence images of the human palm