I have folder named with a lot of location and each of them have 360 degree pics of each of those location i give it a random picture and it has to guess where am i, I tried ORB and it gave me good results and i made it better also but the issue is that lighting conditions might give me a hard time like sunny morning or cloudy morning as per some papers these parameters can cause a issue and i wanted to find is there some way that i can handle this issue is there any way of finding key points irrespective of the weather outside.
I tried to use ORB and getting good results for that movement of time but when lighting changes my dataset is not performing well so i need a solution for that
