二維條碼解碼與驗證應用於工業應用

• Main Authors: Yu-Jui Huang, 黃昱睿
• Other Authors: Jui-Feng Yeh
• Format: Others
• Language: zh-TW
• Online Access: http://ndltd.ncl.edu.tw/handle/02086214881615618983

碩士 === 國立嘉義大學 === 資訊工程學系研究所 === 104 === Today, among the management of computer automated production and automated control, optical inspection and quality management are both important industrial applications. Warehouse management and transportation of up and down stream industries need advanced recognition system as well. Though RFID can offer faster and more accurate checking results, however, electrical labels cost too high. After Wal-Mart abandons RFID due to the high cost, 2D barcode takes the first place in the market of automatic sensor and recognition technology. This research mainly focuses on 2D barcode decoding verification for industrial applications. First, we used CCD to take the 2D barcode pictures then used the features, such as symbol contrast, print growth, axial non-uniformity and unused error correction. These features have its representative symbol decoding level to classifier verification is pass or fail. In the experiment, we used the machine learning method J48 decision tree to training the classifier model, the classify accuracy is better then LibSVM and FURIA. This research here brings machine learning methods for 2D barcode quality verification. Keyword：2D Barcode、Barcode Decode、Barcode Verification、Machine Learning