无人机热红外遥感反演玉米根域土壤含水率方法研究

A Study on the Method of UAV Thermal Infrared Remote Sensing to Retrieve Soil Moisture Content in Corn Root Zone

为了减少土壤背景带来的干扰,更加准确、高效的获取无人机热红外图像中的玉米冠层温度,进而快速反演玉米地土壤含水率,以4种水分梯度处理的拔节期玉米为研究对象,借助无人机可见光和热红外图像,采用RGRI指数法、Otsu阈值法和不剔除土壤背景3种处理方法提取热红外图像中玉米冠层温度信息,计算作物水分胁迫指数(Crop water stress index,CWSI)并用于反演不同水分梯度处理下玉米地不同深度的土壤含水率,基于3种方法获得的CWSI分别记为CWSIRGRI、CWSIOtsu、CWSIsc。结果表明:(1)基于RGRI指数法获取的玉米冠层温度与实测冠层温度的相关性最高(R2均大于0.8;RMSE均小于1℃),Otsu方法次之,不剔除土壤背景方法效果最差。(2)在整个拔节期,CWSIRGRI反演土壤含水率效果最好(R2均大于0.5,P<0.01;效果显著),CWSIOtsu次之、CWSIsc反演效果最差。(3)选取CWSIRGRI为最优CWSI指标,其在玉米拔节期5个土壤深度内的R2呈现先上升后下降的趋势且都在0~30 cm深度内达到最大值。因此,基于RGRI指数法建立的CWSIRGRI可以作为反演玉米地土壤含水率的有效指标。

In order to reduce the interference caused by the soil background,obtain the corn canopy temperature in the UAV thermal infrared image more accurately and efficiently,and then quickly retrieve the soil moisture content of the corn field,this paper used four kinds of water gradient processing jointing stage corn as the research object and used the UAV visible light and thermal infrared images to extract the corn canopy temperature information in the thermal infrared image by using three processing methods,including RGRI index method,Otsu threshold method and no soil background removal,and calculated the crop water stress index(Crop water stress)index,CWSI)to retrieve the soil moisture content at different depths of the corn field under different water gradient treatments.The CWSI obtained based on the three methods were recorded as CWSIRGRI,CWSIOtsu,and CWSIsc.The results showed that:(1)The corn canopy temperature obtained by the RGRI index method had the highest correlation with the measured canopy temperature(R2 was greater than 0.8,and the RMSE was less than 1℃),followed by the Otsu method,and the method without soil background had the worst effect.(2)In the whole jointing period,CWSIRGRIhad the best inversion effect of soil moisture content(R2 was greater than 0.5,P<0.01,the effect was significant),CWSIOtsuwas the second,and CWSIschad the worst inversion effect.(3)CWSIRGRI was selected as the optimal CWSI index,and its R2 in the five soil depths of the corn jointing stage showed a trend of first rising and then falling,and all reached the maximum in the depth of 0~30 cm.Therefore,CWSIRGRIestablished based on the RGRI index method could be used as an effective index for retrieving soil moisture content of corn fields.