松杨栅锈菌担孢子显微图像识别方法

Microscopic Image Recognition Method of Melampsora larici-populina Basidiospore

针对松杨栅锈菌(Melampsora larici-populina)担孢子萌发数量计数问题,提出了利用图像处理技术实现对松杨栅锈菌担孢子显微图像的识别及自动计数。首先,运用形态学重建滤波算法去除松杨栅锈菌担孢子显微图像噪声;然后利用二维阈值化进行图像分割;再对分割后的松杨栅锈菌担孢子显微图像作距离变换;最后利用分水岭算法对松杨栅锈菌担孢子显微图像进行有效分割,实现了对分水岭算法过分割问题的改进。利用孔洞检测算法,实现了松杨栅锈菌担孢子的自动计数。

For counting the number of Melampsora larici-populina basidiospore germination,we presented a method to achieve the recognition and automatic count of M.larici-populina basidiospore microscopic image using image processing technology.First,morphological reconstruction filter was used to remove noise of microscopic image,and then two-dimensional thresholding was used for image segmentation,and distance transform on segmented microscopic image,finally watershed algorithm was used to segment the image.The above steps improved the segmentation of watershed algorithm.The automatic counting of M.larici-populina basidiospore was realized by the hole detection algorithm.