Two-dimensional Bar Code Identification, In The Smart Phone

In this thesis, it introduced a solution method of PDF417 barcode recognition by use of the camera function in a Smartphone according to a practical project work. The algorithm of Barcode image recognition was appropriately designed by the software and hardware structure of a cell phone and has been finally realized in a Linux Smartphone.In the first session, it analyzed the development trend of two-dimension barcode application, introduced the structure of PDF417, compared Smartphone's structure with the specified barcode readers', analyzed the CPU compute ability of a Smartphone and the condition features of the embedded Linux. As cell phones are not designed for image recognition, their optical characteristic and compute ability are much inferior to the specified barcode readers. It means the classical algorithm of barcode recognition is not quite suitable for cell phones, while using the cell phone as a barcode reader, a lot of peculiar problems have occurred. These peculiar problems are what have been mainly analyzed in this part.In the following session, it presented an algorithm of PDF417 two-dimension barcode recognition customized for a Linux model according to the analysis conclusion. The special point of the algorithm is: it solves the optical degradation problem ubiquitous in the camera pattern assembly of cell phones and presents the image restoration solution by use of the very limited processor resource; it is a small calculated amount and low memory assumption barcode recognition algorithm by the application processor characteristic of cell phones. It resolved the conflict between"the large calculated amount of image recognition and the high memory assumption"with"the small compute ability of the embedded process and the limited memory space".