基于CNN的二维码识别算法研究

Research on QR Code Recognition Algorithm Based on CNN

针对生活中大量二维码需要准确识别的需要,提出了一种利用深度学习算法实现二维码识别的方案。该方案利用ROI与鼠标监听技术聚焦待识别区域,线性变换优化图片细节信息;采用CNN网络框架对模型进行训练,提取图片特征;基于OpenCV库设计二维码识别算法。结果表明,该算法对二维码识别率高达98.0%,具有一定的抗干扰能力。

In order to meet the need of accurate recognition of a large number of two-dimensional codes in our daily life,a scheme of two-dimensional code recognition using depth learning algorithm is proposed.In this scheme,ROI and mouse monitoring technology are used to focus on the area to be identified,and linear transformation is used to optimize the image details;CNN network framework is used to train the model and extract image features;Two dimensional code recognition algorithm is designed based on OpenCV library.The results show that the recognition rate of two-dimensional code is 98.0%,and the algorithm has certain anti-interference ability.