基于图像处理的温室大棚中番茄的病害识别

为了准确有效判定温室大棚中番茄病害，利用图像处理和模式识别技术对其（早疫病、睨疫病、叶霉病）进行识别。经过刚像预处理后将叶片病害部位颜色及形状特征提取出来，并通过实验的方法，选取确定了5种显著性较大的特征用于研究，根据最后选取的特征值（颜色特征u，v；形状特征：圆度、复杂度、伸长度）采用贝叶斯判别法对番茄痫害进行识别。取每种病害各40组数据进行实验，结果表明早疫病、晚疫病识别率达到92％，叶霉病识别率达到96％。研究表明该方法能对番茄病害进行有效的识别，并有较高的识别率。

Recognition of Greenhouse Tomato Disease Based on Image Processing Technology

In order to accurate determine, greenhouse shelter tomato diseases, the use of image processing and pattern recognition technology to its （early blight, late blight, leaf mildew） for identification. After image preprocessing after blade disease site color and shape features extracted, and through the experimental methods, five significant larger characteristics for research are determined, according to the characteristics of the selected value （color features u, v, shape features： roundness, complexity, elongation） using bayes discriminant method of tomato diseases identification. Take each of the disease and the set of data, and the results show that early, late blight disease recognition rate reached 92%, leaf mildew recognition rate reached 96%. Research shows that this method can effectively identity tomato diseases, and has high recognition rate.