一种改进的QR码图像二值化算法

摘要

All white or all black error areas would occur to QR code image with uneven illumination after global binariza-tion processing. Pseudo-boundary would appear in the processing of local binarization,and the calculation time would be long. Aiming at the nonuniform illumination QR code image,an improved binarization algorithm is proposed,which is based on back-ground gray-level. Firstly,doing sub-block according to the size of the Qrcode image,and gray-level estimation formula is used for computing the gray-level value of each block. Secondly,joint interpolation algorithm is adopted to create the background gray-level image which is used to take the place of original image to get the corrected image. Finally,the Ostu algorithm is used for the corrected image binarization. Experiment result shows that the algorithm can effectively correct the uneven illumination QR code image and obtain a good binary image.%对光照不均的QR码图像进行全局二值化处理后，会出现全白或全黑的误差区域，在局部二值化过程中会出现伪边界情况，并且计算时间也会变长。针对光照不均的QR码图像提出了一种改进的基于背景灰度的二值化算法。首先，根据二维码源图像大小进行分块处理，使用灰度估算公式对分块的灰度值进行计算。其次，使用联合插值算法产生背景灰度水平图像，然后用背景灰度水平图像替代源图像得到校正图像。最后，采用Ostu算法对校正图像进行二值化。实验结果表明，该算法能有效的校正光照不均的QR图像，并得到一个良好的二值化图像。