| 基于量子神经网络和组合特征参数的玉米叶部病害识别 |
| |
| 引用本文： | 张飞云.基于量子神经网络和组合特征参数的玉米叶部病害识别[J].广西农业科学,2013(8):1286-1290. |
| |
| 作者姓名： | 张飞云 |
| |
| 作者单位： | 浒昌学院电气信息工程学院,河南许昌461000 |
| |
| 基金项目： | 河南省教育厅科学技术研究重点项目（12A510021）;河南省许昌市科技计划项目（1101060） |
| |
| 摘 要： | 【目的】探索一种基于量子神经网络和组合特征参数的玉米叶部病害识别方法，以提高玉米叶部病害识别的准确率和效率。【方法】应用K_means硬聚类算法对玉米叶部病害图像进行彩色图像分割，得到彩色分割图像，分别利用提升小波变换和灰度共生矩阵从彩色分割图像中提取颜色和纹理特征参数，利用多重分形分析从灰度图像中提取病害的形状特征参数。【结果】根据提取的组合特征参数，利用量子神经网络进行玉米病害分类识别，对玉米灰斑病、玉米普通锈病和玉米小斑病的识别率分别达到92．5％、97．5％和92．5％，高于误差反向传播神经网络法的识别率（分别为90．0％、90．0％和92．5％）。【结论】设计的方法可用于玉米叶部病害识别，并为其他农作物病害的智能识别提供借鉴。
|
| 关 键 词： | 玉米病害 组合特征参数 量子神经网络 病害识别率 识别 |
Recognition of corn leaf disease based on quantum neural network and combination characteristic parameter |
| |
| ZHANG Fei-yun.Recognition of corn leaf disease based on quantum neural network and combination characteristic parameter[J].Guangxi Agricultural Sciences,2013(8):1286-1290. |
| |
| Authors: | ZHANG Fei-yun |
| |
| Institution: | ZHANG Fei-yun (College of Electrical and Information Engineering, Xuchang University, Xuchang, Henan 461000, China) |
| |
| Abstract: | Objective ]A new recognition method of corn leaf disease based on quantum neural network（QNN） and combination features was proposed to improve recognition and efficiency. Method]The diseased color image region was segmented by K_means hard clustering algorithm to attain color segmentation image. The color and texture features were extracted by lifting wavelet transform and using gray level co-occurrence matrix （GLCM） from the color segmentation image. The disease shape features were extracted using multi-fractal analysis from gray image. Result ]Combination features were used to index QNN to recognize the corn leaf disease. The corn leaf disease recognition rate of corn gray leaf spot, rust disease and southern leaf blight was 92.5%, 97.5% and 92.5%, respectively, which was higher than that of error back propagation neural network method （90.0%, 90.0%, 92.5%）. Conclusion]Experimental results showed that the method was suitable for corn diseases recognition, and could provide references for intelligent recognition of other crop diseases. |
| |
| Keywords: | corn disease combination characteristic parameter quantum neural network |
| 本文献已被 维普 等数据库收录！ |
