【作者】 韩春花；

【导师】 史大光； 武洪恩；

【作者基本信息】 山东科技大学 ， 精密仪器及机械， 2008， 硕士

【摘要】 一维条码的广泛应用提高了数据采集和信息处理的速度,但因受信息容量的限制,一维条码仅仅是对“物品”的编码,而不能直接包含物品的描述性信息或字符信息,在没有数据库或没有联网的地方,不能使用。二维条码正是为了弥补一维条码的这一缺陷出现的。二维条码的种类很多,目前使用最广泛的是PDF417条码,我国已经发布了PDF417条码国家标准。二维条码阅读技术一般是基于图像的。因此结合PDF417条码的图像处理技术在其推广应用过程中是十分重要的支持技术。基于图像方式的条码识别是将摄像头或扫描仪采集的条码图像通过一定的算法进行处理并译码的过程。本文主要对条码识别系统中条码图像预处理算法做了研究。由于实际应用及图像采集过程中,不可避免的会引入各种噪声,还可能出现光照不均匀、条码污损、原理性的几何形状畸变等情况。所以,在基于图像方式的二维条码图像识别系统中,需要对采集到的PDF417条码图像数据进行适当的图像预处理。本文对条码图像的去噪、二值化问题,应用通用方法进行了算法适用性研究。通过比较一般图像处理方法应用于条码图像的效果,本文对滤波和二值化处理中的算法和参数的选取做了分析与对比。本文的核心工作是在分析条码图像特征的基础上,研究了基于条码基线特征识别的条码图像倾斜角度测量原理,继而给出了倾斜旋转校正的算法,同时也研究了基于条码特征的条码有效区域提取算法。最后,本文还初步探讨了基于条码特征的条码污染部分的恢复算法等问题。本论文在MATLAB6.5平台上对所研究算法进行验证,取得了较满意的效果。这些算法的研究和验证,为基于图像方式的二维条码识读器的研制打下了坚实的基础。更多还原

【Abstract】 Although one-dimension barcode finding its widely application, it is only an encoder of the goods due to its limited information-capacity. One-dimension barcode can’t involve the more information for describing its feature in text mode, and it dose not work where the database and network aren’t available. Two-dimension barcode is just invented to supplement this defect. There are many standards of Two-dimension barcode, and PDF417 standard is the most popular one, which has been published as China national standard.For the technology of reading 2-dimention barcode is image-based, image processing technology adopted in PDF417- barcode processing is the foundation for the application of the barcode. Barcode-processing of image-based is the course of two steps: the processing of the image obtained by video-camera or scanner with certain kind of algorithm and decoding. The study in this paper focused on the pre-processing algorithm of PDF417-barcode.Due to the inevitable noise introduced when obtaining the image, the non-uniform illumination, barcode contaminating, and geometry distortion caused by camera-lens, it is necessary to pre-process the barcode-image in application case. In this thesis, the noise-ellimination and the binaryzation of barcode-image are discussed with general algorithms existed in the textbook in the field of image-processing. The effects from applying different image-proceesing algorithms for barcode-image are compared, e.g. we studied some noise-filtering, image-rotating and image-binarization algorithms with different algorithms and different parameters.The main jobs of this paper are to study the measuring principle of image-skew of barcode based on the base-line characteristic of the barcode-image and to provide the algorithm for rotating the image. Extracting method of valid barcode region is also under discussed. Furthermore, preliminary analysis of contaminated barcode-image recovering technology based on the characteristic of PDF417 barcode is also studied.All algorithms studied in the paper are simulated and verified on MATLAB6.5 platform, and the results are satisfactory. The works can be the foundation of developing a better barcode-image reader for PDF417 application system.更多还原