1. This study proposes a new type of deep convolutional neural network based on Landsat-8 Operational Land Imager (OLI) imagery to extract land cover information.
2. The proposed deep convolutional neural network (DCNN) model can automatically construct the training dataset and classify images, improving the generalization ability of the model and simplifying its application.
3. The proposed DCNN model provides the best classification results in the Qinhuangdao area, with an overall accuracy of 82.0% and a kappa coefficient of 0.76, which is improved by 5% and 14% compared to other methods.
The article is generally reliable and trustworthy as it provides detailed information about the research conducted, including the methodology used, results obtained, and comparison with other methods. The authors have also provided sufficient evidence for their claims by citing relevant literature in support of their arguments. Furthermore, they have discussed potential risks associated with their method such as difficulty in automating the construction of a training database and weak generalization ability of the model.
However, there are some points that could be improved upon in terms of trustworthiness and reliability. For example, while the authors have discussed potential risks associated with their method, they do not provide any solutions or suggestions for mitigating these risks. Additionally, while they have compared their results to those obtained by traditional methods such as support vector machine method and maximum likelihood classification method, they do not discuss any unexplored counterarguments or present both sides equally when making comparisons between different methods. Finally, there is no mention of promotional content or partiality in this article which could be addressed if included in future versions of this article.