Abstract: Window detection is one of the important researches on building image understanding. We
proposed an approach of window detection under edges and glass attribute constraints. Firstly, windows
are basically located with edges in image. Next, color and texture features on the edge neighbor are
extracted, and the dissimilarities between the two features are applied for windows recognition. Finally,
the Mahalanobis distance classifier is first adopted to improve the local recognition containing window
area, then the image segmentation algorithm is used to extract the whole area of the window.
Experiments show that our proposed approach can efficiently detect windows in the case of unknowing
the shape and structure of windows.

Key words: window recognition, edge detection, color, texture, glass