大津法联合信息熵迭代算法在彩码图像恢复中的应用

An Application in Image Restoration of Color Barcode Based on Otsu Method and Iterative Algorithm of Information Entropy

针对彩色条码受异常光照影响色彩偏移的问题,根据其具有双色、四色或八色的特点,提出一种大津法(Nobuyuki Otsu)联合信息熵迭代算法确定阈值的自适应彩色条码图像恢复方法。利用Otsu算法确定初始阈值,结合信息熵迭代算法在该阈值上下范围内进一步精确定位阈值,用此阈值二值化分割彩色条码图像的颜色分量达到色彩恢复的目的。该方法克服了单一使用Otsu算法确定阈值不够准确的缺陷,相比同类算法色彩恢复效果更好,算法运行时间也较短。仿真结果表明,对于光照不足或光照不均达到一定程度下的彩色条码图像,该方法能及时有效地恢复出其真实颜色。

Aiming at the chrominance deviation problem that color barcode is affected from the abnormal light, we present an adaptive restoration method which is based on Otsu method and iterative algorithm of information entropy to determine threshold, according to the features that it has double colors, four colors or eight colors. Firstly, an initial threshold is determined by Otsu method. Then, on the basis of this threshold, a new threshold is further determined accurately by using iterative algorithm of information entropy. Eventually, the color components of color barcode are divided by the binarization to achieve the purpose of color restoration. The method overcomes the defects of using alone Otsu method to obtain threshold,and its effect is better than that of similar method, and running time of the algorithm is shorter. The simulation results show that this method can timely and effectively recover to the true color of color barcode image under a certain extent insufficient light or uneven illumination.