Fig. 14From: Scene search based on the adapted triangular regions and soft clustering to improve the effectiveness of the visual-bag-of-words modelSample images associated with image benchmark of the Ground-truthBack to article page