中文摘要：

黄瓜病害图像分割是其病害图像识别及后续工作的重要步骤。为此,基于二维最大熵原理,结合差分进化算法生成图像分割阈值,提出了一种黄瓜病害图像自动分割方法。为了避免实验中出现的偶然性误差,采用30次独立运行的差分进化优化结果平均值作为图像分割的阈值。自然条件下摄取的黄瓜炭疽病叶图像、灰霉病叶图像和霜霉病叶图像的实验测试表明,该方法具有良好的性能。

英文摘要：

Image segmentation of cucumber disease is an important step in its disease image recognition and subsequent works. With threshold value generated for image segmentation based on Two Dimensional Maximum Entropy principle and Differential Evolution algorithm, we propose a method of automatic segmentation of cucumber disease image. To obtain a relative fair threshold, we run Differential Evolution algorithm 30 times, and take average values of 30 times results as threshold, and use it to image segmentation. After experimental testing with Cucumber anthracnose leaf image, Botrytis leaf image and Downy mildew of leaf image which are acquired under natural conditions, results show high performance of our method.