一种高通量dPCR荧光图像自适应增强算法

目的 为了改善荧光图像背景光照不均匀和对比度低的问题，提出一种荧光图像自适应亮度校正和低对比度增强算法。方法 根据光照成像原理，利用引导滤波提取出荧光图像的光照分量，通过改进的二维Gamma函数动态校正背景光照，利用Top-hat变换分离出校正后的前景和背景，对前景进行自适应直方图均衡化，以实现荧光图像自适应增强的目的。结果 对比传统算法，文中算法处理后的图像背景光照均匀，对比度增强效果明显，其中标准差平均提高了9.4倍，平均梯度平均提高了1.2倍，信息熵平均提高了0.2倍。结论 文中算法可以改善高通量dPCR荧光图像背景光照不均匀性，提高图像对比度，突出图像中隐藏的细节，对其他荧光图像处理也具有参考价值。

An Adaptive Enhancement Algorithm for High-throughput dPCR Fluorescence Image

The paper aims to propose an adaptive brightness correction and low contrast enhancement algorithm for fluorescence images to improve the background illumination unevenness and low contrast of fluorescent images. According to the principle of illumination imaging, the illumination component of the fluorescence image was extracted by guided filtering. The background illumination was dynamically corrected by the improved two-dimensional Gamma function. The corrected foreground and background were separated by Top-hat transformation. And the adaptive histogram equalization was performed on the foreground, to have adaptive enhancement of fluorescent images. Compared with the traditional algorithm, the background illumination of the image processed by the algorithm was uniform and the contrast enhancement effect was obvious. The standard deviation was improved by 9.4 times; the average gradient was increased by 1.2 times; and the information entropy was increased by 0.2 times. The algorithm can improve the background illumination unevenness of high-throughput dPCR fluorescence image, improve the image contrast, highlight hidden details in the image, and has reference value for processing of other fluorescence images.