U-NET and RCNN Ensembled Satellite Object Detection and Segmentation

Humans have always been curious to learn and explore new things and have achieved great things with that curiosity. "CURIOSITY IS THE WICK IN THE CANDLE OF LEARNING," said William Arthur Ward. With the advancing technologies and advancing human needs, automation has become a leading technology in the IT sector. In the year of 1959, the first ever man-made satellite, which could take pictures was launched for the purpose of observing our planet. This led to the widely used technology in the year 2022, The Satellite images play a very crucial role in the n day-to-day activities of humans. It’s because of satellite imagery, the weather forecasting capabilities of humans have improved drastically. People can sleep without worrying about the next upcoming storm as it can be detected with the help of satellite imagery and machine learning algorithms. Thus, precautions for the same can be taken even before the storm is formed. GPS system is another widely used technology used for security and exploration purposes. Satellite imagery with GPS technology gave another revolutionary use for humans, such as live location tracking in a graphical map, detected and captured by satellites. An example of such an application is Google Maps. Google Maps use various type of satellite images and machine learning algorithms to automate the various operations for public use. With the help of these technologies, this article will give an overview that how satellite imagery helps us to detect objects from satellite images and further segment images from satellite images using machine learning algorithms like RCNN with RPN and U-NET.