Astronomical Image Classification using Machine Learning for Point Source Image Detection

Universe is a vast area of research, every country in this world are competing in space research. There are many mysteries to reveal and they can give greater insights to astronomy studies. We used data from Lori Allen and colleagues' research on near-Earth objects using DECam on CTIO's Blanco 4m telescope. Data set consists of source and noise images for classification. There are 5 night photographs in this file. In this image we are looking for the slowest possible field (TNOS) from the orbital field. We collected several images considering these orbits. Our goal is to determine which elements are present in the composite image. Multiple classification algorithms are implied on training data and the best algorithm is chosen based on its accuracy. Training examples include images simulating TNOS (true quality; Stamps_sources. npz) and random trajectories of unknown origin (negative; Stamps_noise. npz). The quality of the signal-to-noise ratio is approximately 100 to 3.