| With the rapid development of modern information technology, barcode has become an effective tool in information storage and recognition. 2D barcode contains information in both horizontal and vertical directions. It holds the features of high capacity, high density, high reliability, strong ablity of error correction and representation obout many kinds of information. At present, Data Matrix code has the smallest sizes, the largest information density and the most widely application.In actual barcode image, there are various kinds of problems such as uneven illumination, diverse perspective, complex background and defaced barcodes. How to recognize barcode accurately and quickly has become an important research area. Based on this idea, the main content and results include the following:1. Study and analysis the development status, structural characteristics and coding principle of Data Matrix code. Design a large number of algorithms and implement a recognition system including architecture design on hardware platform, algorithm design and transplantation on software platform.2. Design and implementation of Data Matrix code recognition algorithms, including image pre-processing, region position, line segment detection, rotation correction, accurate extraction, sampling and decoding. Compare and analyse a variety of classic algorithms, put forward a series of improved algorithms and verify the algorithms by MATLAB.3. Implement Data Matrix code recognition system on SEED-VPM642 hardware system which has TMS320DM642 DSP chip as the core and CCD, SDRAM, FLASH, UART as external modules based on CCS IDE and algorithm design. Finally, test and evaluation on the recognition system.The result shows that the algorithms designed are feasible and the recognition system based on these algorithms can recognize Data Matrix code availably and has higher speed, higher recognition rate and broad development space. |
PDF Full Text Request