Inland water sources like lakes, rivers, and streams are important for the environment and human well-being. Monitoring these water sources is essential to ensure that they remain healthy and productive. This paper presents a study of deep learning-based inland water image classification using neural networks through satellite. The objective of the study is to develop VGG-16 neural network architecture that can be used to accurately distinguish normal images from water images. To assess the performance of the proposed network, several performance metrics are employed. The performance of the neural network is compared to existing methods to ascertain the efficacy of the proposed network. The results of the study show that the proposed neural network architecture is capable of accurately distinguishing normal images from water images, thus demonstrating its potential for successful implementation in real-world applications.