Abstract:

In order to accurately and quickly obtain the vegetation index characteristics and vegetation coverage information of rice， the UAV was used to collect the multispectral image data of rice tillering stage， heading stage and seed setting stage， select different types of commonly used vegetation indexes， and extract and explore the vegetation index characteristics of three growth stages of rice at the scale of plot and pixel by using the ideas of sample statistics method and vegetation index intersection method， Then the threshold segmentation method was used to extract and calculate the rice vegetation information and coverage information. The results showed that in the three growth periods of rice， there were obvious phenological characteristics at the pixel and plot scales， which had obvious phenological characteristics， and there were obvious differences between weeds and trees； the extraction accuracy of vegetation coverage of multispectral vegetation index was higher than that of visible vegetation index； （normalized difference vegetation index， NDVI） had the highest accuracy in extracting vegetation coverage in three stages of rice， with extraction errors of 0.40%， 0.43% and 0.81%， R^{2of 0.77， 0.92 and 0.98， and （root mean square error， RMSE） of 9.09%， 2.97% and 0.38%. The extraction accuracy of （visible-band difference vegetation index，VDVI） was higher than that of （excess green-red-blue difference index，EGRBDI） and （excess green-excess red index，ExG-EXR）， the extraction errors were 4.30%， 1.36% and 1.60%， respectively， R2 were 0.53， 0.77 and 0.80， respectively， and RMSE were 14.62%， 3.70% and 5.50%，respectively. It could be seen that the research results can provide technical support for crop growth monitoring and vegetation coverage extraction.}

Key words: rice, multi-phase, remote sensing index features, UAV multispectral imagery, vegetation coverage, vegetation index

摘要：

为准确快速获取水稻的植被指数特征和植被覆盖度信息，利用无人机采集水稻分蘖期、抽穗期和结实期的多光谱影像数据，选择不同类型的植被指数，利用样本统计法和植被指数交点法，提取并探究水稻3个生长期在地块和像元尺度下的植被指数特征，并运用阈值分割法提取水稻植被信息及覆盖度信息。结果表明，水稻3个生长期内，在像元和地块尺度下均表现出明显的物候特征，且与杂草和树木存在明显区别；多光谱植被指数的植被覆盖度提取精度整体高于可见光植被指数；归一化植被指数（normalized difference vegetation index，NDVI）对水稻3个时期植被覆盖度提取精度最高，提取误差分别为0.40%、0.43%和0.81%，R²为0.77、0.92和0.98，均方根误差（root mean square error，RMSE）为9.09%、2.97%和0.38%；可见光波段差异植被指数（visible-band difference vegetation index，VDVI）提取精度高于超绿红蓝差分指数（excess green-red-blue difference index，EGEBDI）和过绿减过红指数（excess green-excess red index，ExG-ExR），提取误差分别为4.30%、1.36%和1.60%，R²分别为0.53、0.77和0.80，RMSE分别为14.62%、3.70%和5.50%。该研究成果可为作物长势监测及其植被覆盖度提取提供技术支撑。

关键词: 水稻, 多时相, 植被指数特征, 无人机多光谱影像, 植被覆盖度, 植被指数