| 稻飞虱图像形状和颜色特征提取的研究 |
| |
| 引用本文： | 刘德营,丁为民,赵三琴.稻飞虱图像形状和颜色特征提取的研究[J].江西农业学报,2009,21(5). |
| |
| 作者姓名： | 刘德营 丁为民 赵三琴 |
| |
| 作者单位： | 南京农业大学,工学院,江苏,南京,210031 |
| |
| 基金项目： | 江苏省农机局资助项目 |
| |
| 摘 要： | 为实现灯下三种稻飞虱的自动识别和计数,对稻飞虱图像特征提取进行研究。通过对稻田飞虱人工识别特征的分析,确定采集稻飞虱的背部图像。在提取背部特征的研究中,首先对6张稻飞虱单样本图像由傅立叶系数提取形状特征不变量,采用色度百分率直方图法提取背部色度最大值。然后计算大量稻飞虱样本形状和颜色特征值的标准方差以确定各特征值的稳定性。结果表明,白背飞虱图像的各特征值最为稳定。
|
| 关 键 词： | 稻飞虱 形状 颜色 特征提取 |
Extracting Shape and Color Features of Rice Hopper Images |
| |
| LIU De-ying,DING Wei-min,ZHAO San-qin.Extracting Shape and Color Features of Rice Hopper Images[J].Acta Agriculturae Jiangxi,2009,21(5). |
| |
| Authors: | LIU De-ying DING Wei-min ZHAO San-qin |
| |
| Institution: | College of Engineering;Nanjing Agricultural University;Nanjing 210031;China |
| |
| Abstract: | In order to realize the automatic recognition and count of rice hopper,this paper studied the extration of shape and color features from rice hopper images.Through analyzing the artificial recognitive features,the back images of rice hopper were acquisited.During the study on extracting back features,the shape feature invariant of 6 single sample images was firstly extracted by Fourier coefficient,and color feature was extracted by chroma percentage histogram;then the stability of shape and color feature wa... |
| |
| Keywords: | Rice hopper Shape Color Features extraction |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
