AI Based Airplane Air Pollution Identification Architecture Using Satellite Imagery

Air Pollution has become an important problem for governments, researchers, and environmentalists over the last few decades. There are many primary transportation sources of air pollution including airplanes. Automatic airplane recognition in high-resolution satellite images has many applications. One of the applications using artificial intelligence and satellite imagery to design lean smart cities and work on primary sources of transportation air pollutions detection using high-resolution satellite imagery. With the help of satellite imagery and artificial intelligence model, airplane count and detection can be done with accuracy. This paper aims to analyze satellite images in order to help cities to have an idea about the number of planes in the city region. This paper presents web-based end to end aircraft identification framework based on F-RCNN. Utilizing artificial intelligence using deep learning is the state-of-the-art technique to identify the number of planes in a given region with the help of satellite images. The results of the self-made dataset show that the improved F-RCNN has better precision, detection accuracy and masking accuracy which improves the overall efficiency of pollution source identification project in the smart city. The proposed method tested on an image dataset including several airports and non-airports regions. The detection rate could reach approximately 92% accuracy and reduced computation time.