快速响应码识别技术的研究

发布时间：2018-06-12 23:09

本文选题：QR码 + 失真　；参考：《暨南大学》2015年硕士论文

【摘要】：快速响应码(Quick Response Code)也称QR码,属于二维码的一种。近几年QR码在生活中应用越来越广泛,特别是在移动互联网发展起来之后,QR码到处便利着我们的生活。QR码是由日本的丰田公司旗下的电装公司Denso wave在1994年研发出来的,QR码和其他条码比较有很明显的优势,包括存储容量大、数据类型丰富、安全稳定性高等;所以QR码被广泛应用在很多行业,包括物流行业、仓储行业、图书行业、邮政系统、交通领域等。在移动互联网逐渐发展起来之后,QR码又在移动互联网中占据了重要地位,如今很多被大众熟知的移动应用都带有条码扫描的功能,QR码极大地方便了移动应用中查询、支付、下载等功能,这一个小小的符号逐渐深入影响着我们的生活。当然QR码在逐渐发展和推广的同时也会有一系列的相关问题。实际应用中,QR码所处的环境多变:电子屏幕、印刷品上、户外广告上等等,自然对于采集到QR码图像会有诸多影响,可能会出现倾斜失真、几何失真、光照不均等问题,可能会导致QR码无法准确识别。所以在对QR码解码前,需要对QR码进行一系列图像预处理过程消除失真问题,得到一个标准规格的QR码。本文围绕QR码在采集过程中可能会出现的问题进行了研究。关于光照不均匀、倾斜和几何失真,需要对图像进行灰度化、去噪、二值化、倾斜校正、几何校正和取样,研究每个预处理过程的已有算法。着重对QR码的几何失真和取样进行了研究,在倾斜校正的基础上图像的几何校正中的关键点是对QR码的四个顶点进行定位,本文根据QR码的自身特点提出一种新的顶点定位方法,相对于传统方法的边缘检测和Hough变换更简单方便易于操作;几何校正后,有一些几何失真较严重的QR码仍带有模块不均匀的问题,传统的取样方法无法准确地对其进行准确取样,所以在此基础上,本文提出一种自适应匹配的取样方法,可以准确地对几何校正后仍然带有失真的QR码进行取样并且提高了效率,在实际应用中具有一定的研究价值。最后对全文做了总结,以及对往后的工作进行了展望。
[Abstract]:Quick response Code (QR code), also called QR code, belongs to two dimensional code. In recent years, QR codes have been used more and more widely in daily life. Especially after the development of the mobile Internet, QR codes are everywhere to facilitate our lives. QR codes, developed in 1994 by Denso wave, an electric equipment company owned by Toyota Corporation in Japan, have obvious advantages over other bar codes. QR code is widely used in many industries, including logistics industry, storage industry, book industry, postal system, transportation and so on. After the gradual development of mobile Internet, QR code occupies an important position in mobile Internet. Nowadays, many well-known mobile applications have the function of bar code scanning, QR code greatly facilitates query and payment in mobile applications. Download and other functions, this small symbol is gradually affecting our lives. Of course, QR codes in the gradual development and promotion will also have a series of related problems. In practical applications, QR codes are in a changeable environment: electronic screens, printed materials, outdoor advertisements, etc., naturally have a lot of effects on the acquisition of QR code images, which may lead to skew distortion, geometric distortion, and uneven illumination. QR codes may not be recognized accurately. Therefore, before decoding QR codes, a series of image preprocessing processes are needed to eliminate distortion problems and to obtain a standard QR code. In this paper, the possible problems in the acquisition of QR codes are studied. For illumination inhomogeneity, tilt and geometric distortion, gray scale, denoising, binarization, skew correction, geometric correction and sampling are needed to study the existing algorithms for each preprocessing process. The geometric distortion and sampling of QR codes are studied emphatically. The key point of image geometric correction based on tilt correction is to locate the four vertices of QR codes. According to the characteristics of QR codes, a new vertex location method is proposed in this paper, which is simpler and easier to operate than the traditional methods such as edge detection and Hough transform. Some QR codes with serious geometric distortion still have the problem of module inhomogeneity, which can not be accurately sampled by traditional sampling methods. Therefore, an adaptive matching sampling method is proposed in this paper. The QR codes with distortion after geometric correction can be sampled accurately and the efficiency is improved. It has certain research value in practical application. At last, the paper summarizes the whole paper and looks forward to the future work.
【学位授予单位】：暨南大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP391.41

【参考文献】