I have a number of building polygons which are extracted from the aerial images, along with their postal address. (The aerial images are not available in my case.) For example, the polygon of a square shape bungalow is represented as:
[(1,1), (2,1), (1,2), (2,2)] There are several types of building such as house, flat, commercial and bungalow.
What I am currently doing is running shape recognition using Python. However, since most of the building polygons are square or rectangular, the recognition accuracy is very low. So I'm wondering if there is any open source packages that I can use? Or if there is any external data source that can lookup the building type with a given address?
أكثر...
[(1,1), (2,1), (1,2), (2,2)] There are several types of building such as house, flat, commercial and bungalow.
What I am currently doing is running shape recognition using Python. However, since most of the building polygons are square or rectangular, the recognition accuracy is very low. So I'm wondering if there is any open source packages that I can use? Or if there is any external data source that can lookup the building type with a given address?
أكثر...