The Design And Implementation Of The QR Code Recognition System Based On Android Embedded Platform

QR Code has many advantages such as high reliability, large capacity of information,, confidentiality of security, strong resistant to damage, comprehensive and ultra-high-speed reading. What’s more it can represent Chinese characters more efficient, so it is suitable for developing in China. With the development of information technology and two-dimensional bar code, bar code identification equipment is showing intelligence, networking and micro-shape development trend. It can be applied to the field of logistics, e-commerce, security, security, authentication, commodity circulation, advertising and so on. On the other hand, nowadays, Android has the largest market share in embedded field, especially in the handheld terminal field. Therefore, QR barcode recognition research based on android embedded platform has a very important practical significance.This paper first introduces the research background and basis of this subject and research on two-dimensional bar codes at home and abroad, studies the image pre-processing methods, uses median filter to eliminate noise, uses OTSU algorithm to select image threshold, and make image to binarization, uses HOUGH transform and bilinear interpolation methods to correct the images in geometric so as to solve image noise on the impact of bar code.After image pre-processing, the data of QR bar code with stains or defects may be incomplete or have some error data, so the bar code error-correcting ability is extremely important. One-dimensional bar code can check out whether there is dislocation, but the two-dimensional bar code can not only detect errors, but also can correct the error. Conventional RS decoding algorithm is based on solving Linear Equation Group who’s computing complexity is much high, so this article uses an improved Reed-Solomon error correction algorithm to the QR code. The improved algorithm use BM (Berlekamp-Massey) iterative algorithm to calculate the error location polynomial;Chein search algorithm to calculate the root of the error location polynomial, the reciprocal of the root is the wrong location; Forney algorithm to calculate the error value.Improved error correction decoding algorithm not only improves the success rate of the QR Code decoding, but also reduces the computational complexity and improved error correction decoding speed.Finally, due to the some limitation on the speed of bar code identification system based on Android SDK, also developers can use native code (C or C++) to accomplish application which not only can increase the reuse of code and can accelerate program running speed. The NDK applies to independent, less memory and CPU-intensive operations (eg, signal processing, physics simulation, etc.). Image pre-processing and decoding requires large amounts of data operations, so this article used Android NDK to accomplish the QR code image preprocessing and decoding part to improve the speed of the QR code recognition,then designed and developed embedded QR Coderecognition system on tiny6410with Android operating system, also the function of the system was tested.