Research And Implementation Of Two-Dimensional Code Recognition Algorithms In Complex Dynamic Environment

With the increasing usage of QR code in mobile payment,information distribution,device interconnection,smart home and other fields,people put forward higher requirements for the recognition speed,accuracy and robustness of two-dimensional code recognition equipment.Currently,the technology of identifying single two-dimensional codes in ideal environment is very mature.The main challenge of two-dimensional codes in practical application scenarios comes from the recognition of multiple two-dimensional codes in dynamic and complex environments.This thesis firstly analyzes the context of use of such recognition application in current business environment,and focuses on solving the problem of fast and effective identification of QR code in complex and dynamic scenes.Then,after comprehensively comparing various advanced processors,STM32 with high performance-price ratio is selected as the hardware testing platform.This thesis also study QR two-dimensional code image processing pipeline,optimize two-dimensional code recognition processing algorithm,and carry out relevant preliminary testing and research.In this thesis,for complex environment,the task of two-dimensional code recognition without prior location information,multi-scale and illumination changes is analyzed,and the algorithm is designed and implemented.Then,for the recognition of two-dimensional codes in dynamic scenes,the algorithm is designed by pre-estimating the two-dimensional codes and camera pose and restoring the blurred image.This paper tries to solve the problem of fast and effective recognition of two-dimensional codes in complex and dynamic environments.The research shows that the QR two-dimensional code recognition algorithm in complex and dynamic scenarios studied in this thesis has good feasibility and effectiveness,and has practical application prospects after further improvement and optimization.