Small wetlands are widely distributed in urban and rural areas, serving as important water resources and water environment control units, as well as migratory habitats for flora and fauna and sites for biodiversity conservation. However, they are facing threats from climate change and the transformation of the interface between urban and rural spatial dynamics. Based on Gaofen remote sensing images, this study extracted and validated the extraction accuracy of small wetlands and other land use types around Chaohu Lake from 2015 to 2021 using three techniques, namely, random forest (RF), support vector machine (SVM), and maximum likelihood (MLE). Changes in the number of areas of small wetlands and the main driving factors during the period of 6 years were computed using ArcGIS. The results are as follows: (1) The overall classification accuracy and Kappa coefficient trends for 2015, 2018, and 2021 were all RFt > SVM > MLE, and the RF classification effect was the best. (2) The area of small wetlands around Chaohu Lake increased from 9114.42 hm2 in 2015 to 10706.84 hm2 in 2021, but the number decreased from 22279 to 21338. (3) The interaction between two factors has a greater impact on the area of small wetlands than a single factor. The interaction intensity of construction land, annual average precipitation, and altitude is relatively strong with others, which has the strongest impact on the dynamic changes of wetlands. The results emphasize that the accuracy of basic research data on small wetlands can be improved by using high-resolution remote sensing images and selecting classifiers, and that reducing disturbance from anthropogenic construction is a prerequisite for protecting and maintaining the ecological functions of small wetlands, and provide decision-making basis for the sustainable development of small wetlands.