Revision history [back]

I cannot give you some code but maybe an idea how to do it:

Well for face identification - the face is aligned before extracting the feature vector. This is done by using facial landmarks of person (eyes, nose, mouth, etc). Heres how its done

I think you could do the same for buildings. Extract landmark (windows, door, roof, etc) and use an algorithm to measure the angle and align the image.

How to extract building landmarks? Try this

Hope this helps, Good luck.

I cannot give you some code but maybe an idea how to do it:

Well for face identification - the face is aligned before extracting the feature vector. This is done by using facial landmarks of person (eyes, nose, mouth, etc). Heres how its done

I think you could do the same for buildings. Extract landmark (windows, door, roof, etc) and use an algorithm to measure the angle and align the image.building.

How to extract building landmarks? Try this

Hope this helps, Good luck.

I cannot give you some code but maybe an idea: Its just an idea how to do it:discuss!

Well for face identification - the face is aligned before extracting the feature vector. This is done by using facial landmarks of person (eyes, nose, mouth, etc). Heres how its done

I think you could do the same for buildings. Extract landmark (windows, door, roof, etc) and use an algorithm to measure the angle and align the building.

How to extract building landmarks? Try this

Hope this helps, Good luck.