The information acquisition of tree height, individual tree crown and diameter at breast height is an important part of forest resources survey. In this study, the UAV images of artificial Robinia pseudoacacia forest in Zhangjiatan Town of Yan River Basin were obtained by DJI Phantom 4 UAV platform. The crown width and tree height were obtained by neural network clustering algorithm, visual interpretation and spatial measurement. The relative error between the extracted value and the measured value of individual tree crown width was 5. 98%, and the relative error of tree height was 7. 53%. We used 80% of the data to extract, crown, tree height and diameter at breast height to establish single regression models, exponential function models, logarithmic function models, power function models and binary regression models. Among those established DBH inversion models, the model with the highest fitting accuracy was y = 4.255a + 0.044b² − 0.135a² − 2.111, with a determination coefficient of 0.85, and the average relative error was 4.3% after being verified by using the remaining 20% of the data. The stand factors of the study area can be quickly obtained by using UAV images, and the DBH can be obtained with high precision, which can provide a reference for the realization of precision forestry.