研究目的

To obtain high-definition digital images of different varieties of rice using low-altitude drones, and to analyze the feasibility of estimating the LAI of rice canopy by empirical model method and porosity model method, and analyze the difference and estimation process between them.

研究成果

The study proves that it is feasible to use the UAV to mount the rice canopy image with the hyperspectral digital camera and combine the digital image processing technology to quickly obtain the rice LAI information. According to the difference of spectral characteristics of green vegetation and background, the threshold value was established, the LAI extracted by the gap rate model had a good correlation with the vegetation index, in addition, the fitting effect of LAI and the vegetation index of rice during jointing stage was greatly influenced by the variety and density.

研究不足

The study is based on noncontact test measurement of limited samples. The main research is on the correlation of the two LAI estimation methods. If we further obtain the field measured LAI and perform correlation analysis, the conclusion will be more persuasive. This test is aimed at a specific species, time, and development period. It is considered that each sampling area is composed of canopy and noncanopy pixels, but there are actually mixed pixels, so the accuracy of vegetation coverage and threshold selection will be Directly affect the results of the study, so whether the model has universality needs further verification.