关键词: 小麦, 无人机, RGB图像, LAI, 估算模型

Abstract: Crop LAI is one of the important indicators to characterize the growth status of crops. In order to monitor wheat LAI rapidly and nondestructively, 3 densities and 4 nitrogen fertilizer treatments were set in this study to form different wheat growth groups, and field images were acquired by RGB camera mounted by UAV, and LAI was measured by synchronous sampling. The 9 color characteristic indexes constructed by R, G and B were selected to conduct correlation analysis with the measured LAI during the overwintering stage, returning green stage, jointing stage, flowering stage and grouting stage of wheat growth. The results showed that, at the early stage of wheat growth (overwintering stage, and returning green stage), the correlation between color index and LAI was weak, while at the late growth stage (jointing stage, flowering stage, and grouting stage), the correlation between all color indexes and LAI reached extremely significant level. The model of maximal R2 in the commonly used exponential, linear, logarithmic, polynomial and power function models was selected as the final estimation model to estimate wheat LAI in three periods: jointing, flowering and grouting stages. The estimation models were verified to be reliable and accurate by the measured LAI. These results provided a new means for rapid measurement of LAI in crop fields.

Key words: wheat, UAV, RGB image, LAI, estimation model