Abstract: In order to extract the shared flavour fingerprint information from Guangxi fragrant rice and to rapidly realize the discrimination of fragrant rice between Guangxi and Wuchang regions, a total of 53 fragrant rice samples were collected and gas chromatography-ion mobility spectrometry (GC-IMS) technology was used to analyze the volatile flavor components. Based on the image pretreatment and automatic threshold segmentation algorithm, the shared flavour components of Guangxi fragrant rice were extracted in the two-dimensional map, and quadratic discriminant analysis (QDA) combined with principal component analysis (PCA) was applied to establish the qualitative model for distinguishing two different regions of fragrant rice. The results showed that 285 shared characteristic variables were extracted from fragrant rice by using the automatic threshold segmentation method. The first two principal component scores showed that each region of fragrant rice had their own corresponding areas. The discriminant model indicated that the recognition rate of two regions of fragrant rice could reach 100%, which could be used to estimate the discrimination between fragrant rice samples. In conclusion, digital image processing technology combined with chemometrics could effectively extract characteristic flavour variables from the two-dimensional map, which could achieve the rapid, non-destructive and accurate analysis in fragrant rice quality based on GC-IMS technology, and also provide an analysis method for multi-dimensional information for other similar instruments.