Abstract：Planting area of fruit trees in Xinjiang accounts for about 13% of the national planting area, which is the main fruit trees producing area in China. With the help of suitable climate condition and resource advantages, the four prefectures of Southern Xinjiang around the Tarim River Basin have become the main producing area of Xinjiang specialty fruit (e.g. walnut, jujube, apricot, fragrant pear and apple). The fruit planting area here accounts for more than 80% of the total fruit planting area in Xinjiang. Real-time and accurate acquisition of fruit tree type and area information under the pattern of forest and crops interplanting is significance to improve the quality and efficiency of specialty fruit industry in southern Xinjiang. It will be conducive to improvement local farmers’ income, stabilize the achievement of poverty alleviation and promote rural revitalization. This study is to take the continuous area of forest and crops interplanting in Hotan Oasis of southern Xinjiang as an example. It proposes a method for extracting the structure information of fruit trees that integrates high-resolution remote sensing image data with abundant texture and spectral characteristics and medium-resolution sensing image data with multi-temporal characteristics. Firstly, it used object-oriented methods to extract high-precision boundary of fruit trees parcel based on GF-2 (PMS) image data. The classification rules of GF-2 image data are divided into winter (February) and summer-autumn (July-September). By analyzing Normalized Difference Vegetation Index (NDVI), spectral characteristics and texture feature information between target objects and other ground objects, it identified the correspondence between feature information and ground objects, and the classification of four plots was obtained by gradually eliminating non-target ground categories. Then, it constructed NDVI time series products based on multi-temporal Sentinel-2 image data, and established a decision tree model based on the characteristics of phenology to extract interplanting walnut orchard, pure walnut orchard, jujube orchard and grape orchard. The NDVI time series of fruit trees was analyzed, and it found that NDVI time series curve had many peaks and troughs in one year. The peak would represent the flourishing period of fruit trees growth, and the trough reflected the orchard management (such as irrigation and pruning branches). Although the NDVI timing series of pure walnut, interplanting walnut, jujubes and grapes are rarely staggered and overlapping, there are still obvious and different time windows. These differences contribute to the fruit trees classification. Finally, the multi-phase orchard classification results were overlay the high-resolution of fruit trees parcel to obtain the distribution of fruit crops in Hotan Oasis. The research results show that the area of major fruit was 4.29×105 hm2 here, with 3.31×105 hm2 of walnut orchard (including interplanting and pure walnut), 8.29×104 hm2 of jujube and 1.40×104 hm2 of grapes.The area of interplanting walnuts accounts for 63.8% of total fruit area,followed by jujube (19.38%) and grapes (3.3%). The user accuracy and overall classification accuracy are both exceed 90%, and the Kappa coefficient is 0.95 which would meet the accuracy requirements of agroforestry classification at the county and city level. Compared with the forestry survey results in 2019, the relative accuracy of walnuts jujube and grapes based on remote sensing extraction results was 62.1%, 97.8%, and 85.2%, respectively. The results show that the area of jujube and grape based on remote sensing extraction is close to the forestry survey datas. The walnut planting areas in Hotan area are mainly distributed in the upper reaches of Yurunkax River and Karakax River with suitable soil and water conditions. The jujube trees are mostly distributed in the downstream oasis desert ecotone and the grapes are mostly distributed in the sandy desert of lower reaches of Karakax River,. This method could provide valuable reference for the research of fruit tree type extraction under the pattern of forest and crops intercropping.