[Python] Breaking a high res image into chunks for image recognition using computer vision

Im a noob to computer vision so please go easy on me.

I have a 15000x15000 png image of the night sky. I want to train a CV model based on certain features in the image that I identify as galaxies. My plan is to break it up into say 700x700 pixels. Classify each chunk, and use that to train my model. The image I have is pretty high res so when I view it in an image view fully zoomed out you cant really see the galaxies. Its mostly a black sky with faint streaks of stars. But if you zoom in deep enough, you can see the stars and constellations with higher clarity.

My question is:

1. 
   What methods would you recommend to go about achieve this? I'm planning on using python. (I don't need the code, just the best approach to go about this)
   
   
2. 
   If I break the image into chunks using pillow or OpenCV, will it retain the resolution of the zoomed up image?
   

Thanks for the guidance!

Continue reading...