Abstract: Objectives: The monitoring of salt marsh vegetation in the Yellow River Delta wetlands is the basis for the protection and restoration of the ecological functions of the Yellow River Delta wetlands. The vegetation situation in the Yellow River Delta is complex, and the detection and classification of vegetation is particularly important due to the spread of the invasive species Spartina alterniflora, which has a significant impact on wetlands. Methods: This study takes part of the Yellow River Delta wetland as the study area, takes high-resolution aerial images as the data source. Four feature variables are generated: spectral features, color features, index features and texture features, and six different classification schemes are constructed. Scheme 1 with only spectral features was regarded as the control group, and index features, texture features and color features were integrated into scheme 2, scheme 3 and scheme 4, respectively. Scheme 5 contains all features, and scheme 6 constructs a multi-feature optimization feature set for all features. The random forest was used to classify vegetation for each extraction scheme and their accuracy were verified, aiming to explore the influence and reasons of different feature variables on the classification. Then the best preferred features were selected to improve the effect of vegetation classification. Results: Based on the visible and near-infrared spectra, just adding different features to the experiment had different effects on the accuracy of vegetation classification: (1) Scheme 1, scheme 2 and scheme 3 have unsatisfactory extraction effect on Phragmites australis and Suaeda salsa. Scheme 4, scheme 5 and scheme 6 with the addition of color features can better distinguish between the two, probably because Suaeda salsa is dark red on the image, which is quite different from other vegetation, so the color features can be better distinguished. (2) Scheme 4 divides the tidal flat area with non-vegetation information into Phragmites australis at the edge of the water body, which are similar in color, but partially misdivided when color features are added. Schemes 5 and 6 were well classified, but due to the mixed nature of vegetation, all schemes had varying degrees of misclassification of Tamarix chinensis, Phragmites australis and Spartina alterniflora. (3) The index features have a positive effect on Suaeda salsa extraction, texture features will reduce the accuracy of vegetation classification, and the integration of color features for classification is the key to improve the overall accuracy. (4) Based on multi-feature optimization of random forest, the extraction effect was the best, with an overall accuracy of 88% and a Kappa coefficient of 0.85. Conclusions: The main advantages of this study are the acquisition of new data sources, the introduction of multiple feature variables, and the experimental evaluation and classification accuracy analysis of different feature variables. The importance of color features are verified and multi-feature optimization of random forest by integrating color features is a feasible method to classify vegetation information in the Yellow River Delta . The method can effectively distinguish vegetation from non-vegetation and extract each vegetation type at the same time. This study provides an effective technical route in feature selection and methodology for vegetation information extraction in the Yellow River Delta.