Paper fingerprinting using alpha-masked image matching Tuan Q. Pham † Stuart W. Perry Peter A. Fletcher Canon Information Systems Research Australia (CiSRA) 1 Thomas Holt drive, North Ryde, NSW 2113, Australia. † corresponding author: tuan.pham@cisra.canon.com.au Abstract In this paper, we examine the problem of authenticating paper media using the unique fibre structure of each piece of paper (the so-called ”paper fingerprint”). In particular, we look at methods to authenticate paper media when text has been printed over the authentication zone. We show how alpha-masked correlation [8] can be applied to this problem and develop a modification to alpha-masked correlation that is more closely matched to the requirements of this problem and produces an improvement in performance. We also investigate two methods of pixel inpainting to remove printed text or marks from the authentication zone and allow ordinary correlation to be performed. We show that these methods can perform as well as alpha-masked correlation. Finally two methods of improving the robustness to forgery are investigated. 1. Introduction When a reasonable facsimile of an object can be created at a cost less than the value of the object, the object may be become a target for forgery. The concept of an object’s ”value” can be defined in many different ways. Paper currency is an example of an object with a cost to manufacture that is often small compared to the value society places on the object. Even plain paper documents whose cost is almost negligible to produce can have a very large value due to the importance of the information printed on the document. In recent times the availability of cheap printing technology has further reduced the cost of manufacture for paper objects, increasing the temptation of forgery and improving the quality of forgeries. There have been many different strategies employed to protect objects from forgery. One strategy is to make the equipment required for manufacture of the authentic object so expensive that forgery is discouraged. For example, the addition of holograms, special paper and/or inks or finely printed details. Another approach that has gained interest recently is security based on the inherent randomness present in many objects. An ordinary piece of office paper when viewed at the micrometer scale is a highly random arrangement of fibres and filler material. Every sheet of paper has a different arrangement of fibres and no two regions on the same sheet of paper have the same arrangement. The randomness of this arrangement can be exploited to create a unique signature for the sheet of paper. A recent article described the creation of unique signatures for a variety of common objects such as paper, coated cardboard packaging and matt-finish plastic cards using the diffuse reflection from a laser focussed onto a small region of the object [3]. A unique signature could be obtained even when the object had been soaked in water and then dried, baked in an oven, crumpled and creased, or written over with a ballpoint pen or thick marker. An object could be authenticated by cross-correlating the signature obtained from the object with those in a database of signatures of authentic objects. The basis of this idea has been around for a while. It can be traced back as least as far as US patents filed in 1981 [9] and 1984 [10]. In these patents the inventor discloses the idea of using information about inherent irreproducible randomness within an object imaged by an optical system to form a unique signature for the object. More recently it has been noted that a coherent light source is not required for a unique signature to be computed for paper [16]. Indeed, the earlier work [9], [10] does not assume a coherent light source. By simply imaging the paper surface illuminated by incoherent light at high resolution, sufficient information about the random structure of the paper surface can be obtained to create a highly unique signature for the paper [16]. The authors have found that the requirements for imaging a piece of common office paper to obtain a highly unique signature are easily met by many common consumer document scanners. In addition, experiments we have per- formed indicate that the signatures obtained are robust to damage to the paper such as light staining, wrinkling, and wetting followed by drying. The use of a desktop scanner for this application provides a number of benefits including removing the requirement for specialised equipment to read the signatures. In this work the signature we use is simply a greyscale 8-bit 256 by 256 pixel image of the paper surface captured by the scanner at 600dpi. In [3] and [16], cross- correlation was used to match an obtained signature with a previously stored signature, and this approach can be used to match the above image signatures from a desktop scanner. However, simply cross-correlating an image of the surface of the paper does not provide a signature authentication