水稻稻瘟病及水稻纹枯病病害程度图像检测

由于水稻稻瘟病及水稻纹枯病的病害程度是根据距离特征来检测，目前还没有定义数字图像距离的唯一方法。为了实时获取作物病虫害危害程度信息，该文研究了一种利用椭圆模型来检测水稻稻瘟病及水稻纹枯病的危害程度--用椭圆拟合水稻稻瘟病单株最大病斑，计算椭圆主轴长与人工检测的实际最大病斑的长度比较，分析两者相关性，结果具有极显著性；同理，用椭圆拟合水稻纹枯病单株高度及病斑最高处距基部的距离，计算两椭圆主轴长并求两者之比即可对水稻纹枯病危害程度进行检测，平均正确率在90%以上。该研究对大型精准喷雾植保机械实时获取植物病虫害危害程度信息有较大参考价值。

Detection of the Harm Degree of Rice Blast and Rice Sheath Blight

The harm degree of the rice blast and the rice sheath blight was detected according to the distance characteristic, at present there had not defined the only method of digital image distance.To gain harm degree information of crops disease in real time, this paper studied one method to examine harm degree of the rice blast and the rice sheath blight with ellipse model: used ellipse to fit the biggest sickness spot of the rice blast, calculated the length of principal axis of ellipse and the compared with manual examination the biggest sickness spot length, analysed both relevance, which had extremely the significance; Likewise, used ellipse to fit the length of paddy rice and the distance of disease apart from the base of the rice sheath blight, calculated two principal axis of ellipse length and asked the ratio, the average accuracy of measuring the harm degree of the rice sheath blight was upper to 90%. This study has greater reference value for the large-scale machine of plant protection to obtain the information of injury extent of plant diseases in real time.