The recognition of laser anti-counterfeiting code with complex background

Anti-counterfeiting code of cigarettes can prevent the fake, shoddy and the sales in wrong area. And it provides a lot of important information such as delivery date, sales area, and nature of customer. A recognition method for anti-counterfeiting code with complex background using Log-Gabor filters and Support Vector Machine (SVM) classifiers is proposed in this paper. Log-Gabor transform is used to extract the character image's local texture features without the illumination impact, and SVM classifiers are superior to other methods to solving small example size recognition problems. The experimental results show that the proposed method performs effectively for recognition of laser anti-counterfeiting code images with noises, non-uniform illumination, backgrounds or stroke distortions in low quality grayscale at the rate of 97%.