为实现藻类的自动计数，以小球藻为对象，提出一种基于荧光效应和调色板均衡化的小球藻自动计数方法。首先对小球藻荧光显微图像进行预处理，将真彩色小球藻荧光图像转换为256色图像，再采用调色板均衡化算法增强小球藻荧光图像对比度，并利用基于调色板的阈值分割算法将目标藻类与背景进行分割，同时进行计数。结果表明，该计数方法能够避免杂质影响，直接对藻类彩色图像进行图像增强和阈值分割，可以增强图像对比度，避免彩色图像灰度化过程中图像信息损失；采用连通区域标记方法结合连通区域的大小识别低重叠度藻类，可降低重叠藻类对计数结果的影响，从而提高小球藻的计算准确度。研究表明，该方法实现了针对藻类彩色图像的自动计数，且对小球藻的计数精度准确率达到95％以上，高于只使用灰度图像计数的方法。%In order to realize the automatic counting of algae, taking chlorella ( Chlorella vulgaris ) as the target, an automatic counting method based on fluorescence effect and palette equalization was proposed in this paper. Firstly, the fluorescent micrograph of chlorella was pretreated, by transforming it into 256-color images and enhancing its contrast by palette equalization algorithm. Then the target algae and the background were divided using threshold segmentation algorithm of the palette, and simultaneously counted. The results showed that this counting method could avoid the interference of impurities, directly realize the algae color image enhancement and threshold segmentation, increase the image contrast and avoid loss of image information during the process of color image graying;and by using the connected area marking method combined with the size of the connected region to recognize the algae of low overlapping degree, it reduced the influence of overlapping algae on the counting results, and thus could improve the calculation accuracy of chlorella. The experimental results show that this method can achieve automatic counting for algae color images, and the accuracy of chlorella counting accuracy is higher than 95%, which is higher than that of those methods using only gray scale image.
展开▼
