City-scale industrial tank detection using multi-source spatial data fusion

This paper focuses on the automatic detection of industrial storage tanks in urban areas using deep learning-based algorithms. Industrial storage tanks are critical for storing raw materials, finished products, and intermediate products in industries such as petroleum, chemical, and metallurgy. However, they can leak and cause environmental damage, making it important to monitor them in cities. The challenge lies in the large number and dispersed distribution of these tanks throughout the city. To address this, high-resolution remote sensing images and deep learning algorithms are used to improve the accuracy of industrial storage tank detection at the city scale. We construct a city scale industrial storage tank object detection dataset using high-resolution remote sensing images and explore the effect of deep learning object detection algorithm optimization on industrial storage tank detection. Techniques including ResNet50, FPN, and dilated convolution are utilized in this work for improving the model detection accuracy. Furthermore, we construct multiple city industrial storage tank datasets based on high-resolution remote sensing images enriched by integrating point-of-interest data and land use data of each city. The integration from multiple data sources provides an accurate and efficient solution for the identification of industrial storage tanks within urban areas.