Enhancing the mental representations of space used by blind pedestrians, based on an image schemata model