 This paper investigates the use of deep convolutional neural networks, CNNs, for classification of complex wetland classes using multispectral rapid eye optical imagery. Seven popular CNN architectures were evaluated, including DenseNet-121, Inception-V3, VGG-16, VGG-19, Inception-ResNet-50, and Inception-ResNet-V2. The results showed that the full training of Inception-ResNet-V2 provided the best performance, achieving a classification accuracy of 96.17%. This was significantly higher than the other CNN models tested, which ranged between 74.89% and 93.57%. Additionally, the results demonstrated that the integration of Inception and ResNet modules is an effective architecture for classifying complex remote sensing scenes such as wetlands. This article was authored by Massoud Muddy-Emprey, Baram Salehi, Muhammad Razi, and others. We are article.tv, links in the description below.