基于Hu不变矩和BP网络的条形码图像识别方法

摘要

To improve the accuracy in the recognition of one-dimension and two-dimension bar code images with geometric distor tion, a novel bar code recognition method based on invariable moment and BP network classifier is presented. The invariable mo ment eigenvector of the bar code images is extracted as eigenvalue inputting into BP network in order to train and test the BP net work. The deformed bar code images are recognized by using the trained BP network. The method realizes correct identification the bar code images with geometric distortion about rotation, translation and scaling. The experimental results show that this method is simple, fast training and high recognition rate in image's translation, scaling and rotation.%针对目前比较流行的一维条形码和二维条形码识别算法存在对几何失真图像的识别准确率较低的问题,提出了一种新的基于不变矩和BP网络的条形码识别方法,提取不变矩特征向量作为特征值输入BP网络,对其进行训练与测试,利用训练好的BP网络对形变条形码图像进行识别,实现了对存在旋转、平移和缩放等几何失真的条形码图像的正确识别.实验结果表明,基于Hu不变矩和BP网络的条形码识别方法具有很强的抗图像平移、拉伸和旋转识别能力,并且具有实现简单、训练速度快、识别率高等特点.