Leaf area index is an important indicator of crop growth evaluation, so it is crucial to estimate LAI quickly and accurately. The imaging data of the three growth stages of winter wheat was obtained through the imaging hyperspectrum carried by the UAV, and the plant height (Hcsm) was extracted from it. Firstly, the correlation between vegetation indices, Hcsm and LAI was analyzed, and the optimal vegetation index was selected; then the LAI linear estimation model of a single parameter was constructed separately; finally, taking the vegetation indices and vegetation indices combined with Hcsm as the model input factor, the partial least squares regression method was used to construct LAI estimation model. The results showed that the height of the plant height Hcsm extracted from the UAV hyperspectral remote sensing image was highly accurate (R2=0.95); the correlation between most vegetation indices and Hcsm at different growth stages and LAI was at 0.01 significant level; the accuracy of estimating the LAI based on the optimal vegetation index combined with Hcsm was better than that based on the optimal vegetation index or Hcsm only; taking vegetation indices and vegetation indices combined with Hcsm as input variables, the LAI estimation model constructed by partial least square regression achieved the highest accuracy during flowering stage, so partial least squares regression can improve the estimation effect, and the ability to estimate the LAI with the vegetation indices combined with Hcsm as the independent variable was better (modeling R2=0.73, RMSE was 0.64). The research was based on the Hcsm extracted from the UAV hyperspectral remote sensing image combined with the vegetation indices, which can improve the accuracy of estimating LAI and provide a reference for agricultural managers. 