Extraction of maize canopy temperature and variation factor analysis based on multi-source unmanned aerial vehicle remote sensing data

Canopy temperature is an important parameter for monitoring the physiological status of maize. The study of its change law can help to improve the growth environment and increase the yield and quality. Based on the RGB, thermal infrared, and multispectral images obtained by an Unmanned Aerial Vehicle (UAV), two peak threshold methods were used in this study to eliminate the soil background and obtain the maize canopy information in the study area. Then, by using the radiation transmission model and Planck’s law, the maize canopy temperature was extracted, and the correlation of canopy temperature change with the regional coverage and soil drought degree was analysed. Results indicated that the accuracy of canopy temperature extraction can be effectively enhanced by eliminating the influence of soil background. At jointing and filling stages, the Pearson correlation coefficients r of canopy temperature and coverage were 0.731 and 0.722, respectively, and the slope of the linear fitting model was less than 0, indicating that the coverage has a negative impact on canopy temperature, i.e., the higher the coverage, the lower the canopy temperature, and vice versa. Additionally, statistical analysis of the canopy temperature and temperature vegetation drought index (TVDI) clustering results at the jointing and filling stages demonstrated that the total overlapping rates of the categories were 72.5