Unmanned aerial vehicles were used to monitor cotton growth by capturing hyperspectral images of the canopy. By analyzing the spectral reflectance and vegetation indices, a model was created to estimate the Leaf Area Index (LAI) of cotton plants. The study found that reducing spectral noise improved the accuracy of LAI estimation, with the best results achieved using a combination of spectral transformation and vegetation indices. The model based on this approach had a high accuracy in predicting cotton LAI throughout the growth period.