Abstract: 【Objective】 To establish the image elevation data model for inversion of forest growing stock inversion of Picea schrenkiana var. tianshanica based on the trees in Qiaxi Forest Park in Gongliu County, western Tianshan, Xinjiang, UAV aerial image and sample ground per wooden scale as data source in the hope of providing a reference for the ecological restoration and scientific management of natural forest protection project. 【Methods】 Using point cloud classification and Kriging interpolation to extract UAV image elevation data to obtain the height of Picea schrenkiana var. tianshanica, and at the same time, the DBH tree height model was established according to the measured data of the sample land, and the forest volume of Picea schrenkiana var.tianshanica. 【Results】 There was a significant positive correlation between tree height and measured tree height extracted by UAV images. the average extraction accuracy was 88.42%. The correlation coefficient of DBH-crown width model of Picea Schrenkiana var. tianshanica was 0.696, while that of DBH-tree height model was 0.730. Verify the DBH-tree height model, the calculated RMSE value is 12.386, the fitting effect is significant. The accuracy of retrieving stand volume based on DBH-tree height model is 87.66%. Compared with the measured values, most of the residual values fall in the residual range of (- 2,+2). 【Conclusion】 The local maximum algorithm is effective in extracting tree height information of Picea Schrenkiana var. tianshanica forest. The establishment of DBH-tree height model makes up for the defect that UAV can not measure DBH directly, and then inverts the stand volume ofPicea Schrenkiana var. tianshanica forest.