Recognition of banknotes is a prevalent problem for blind and visually impaired people. The bionic eyeglass is a mobile platform that integrates several common visual detection and recognition functions. Relevant shapes are extracted from the image flow of the banknote shown to the mobile camera using adaptive thresholding and morphological shape filters. In a two-level classification scheme different classifiers are used for different categories of patches (portrait, denomination, tactile marks), and the votes are combined through an ensemble decider. The confidence of the decision is established if the same class recurs through several frames. The system has been tested on blind subjects using a cell phone based prototype, and the results showed that they quickly became confident using it.