说明：  提出了用提取轮廓分割识别目标图像的新方 法，即先对病害图像作反色运算，对得到的图像进行二值化后，再进行腐蚀运算，对其结果 作轮廓提取，可消除全部噪声。与传统轮廓提取方法相比，分割更为准确，几乎没有噪声。
(The proposed new method of identifying the target image segmentation by extracting contours, i.e. the first image on the disease as anti-color operation, after the obtained image is binarized, and then etching operation, and the results for contour extraction, can eliminate all noise. Compared with the conventional contour extraction method, segmentation is more accurate, almost no noise.)