Font Size: a A A

Study On Recognition Of QR Code Image Based On Grayscale Morphology

Posted on:2013-01-16Degree:MasterType:Thesis
Country:ChinaCandidate:X B ZhangFull Text:PDF
GTID:2218330362466813Subject:Measuring and Testing Technology and Instruments
Abstract/Summary:PDF Full Text Request
Because of the supervelocity and omni-directional recognition, the QR code (QuickResponse Code) has rapidly obtained the application promotion as soon as it appears on themarket. Its specific data compression pattern expressed the Chinese character and the Japanesecharacteristic promotes the use in our country.Because of various limitations, we cannot sample ideal QR code image, which includes onlythe QR code symbol and has no noise. Thus we need to process the sampled image previouslybefore recognizing. The main purpose of this paper is to solve the above problem using thetheory of morphology.Firstly, the basic theories, including the structure of the QR Code, coding, and mathematicalmorphology, are introduced. We analyze the code features and advantages and describe thecoding theory of the QR code in detail. After that, we discuss the basic theory of mathematicalmorphology, and extend the binary morphological concepts to gray-scale images. In addition,some applications, such as morphological filtering and morphological gradient, are introduceddetailedly.Then, we study the edge detection of the QR Code image polluted by noise. As there is acontradiction between enhancing edges and reducing noise in the classical edge detectionalgorithms, we propose an effective improvement scheme. We improve the morphologicalgradient operator to detect the edge information.Again, an approach based on grayscale morphological image processing is introduced for theQR code binaryzation. The image distortion caused by uniform illumination is solved well usingthis method. Given a QR code image's shape and size, we can acquire a suitable structuringfunction to deal with the image. Here our proposed approach is tested using some different typesof QR images. The testing results are compared with several well-known methods. Experimentsshow that our proposed method is not only less-time consuming, but also higher decodingaccuracy.Finally, we introduce the decoding process of QR code, explore the application of the QRcode in real life and make the prospect of the future work.
Keywords/Search Tags:Quick Response Code, mathematical morphology, edge detection, binarization
PDF Full Text Request
Related items