Recognition And System Implementation Of Low Quality Laser Anti-counterfeiting Code

The widespread existence of fake and inferior commodities affects the order of market and the brand effect of manufacturers,and threatens the rights and interests of consumers.It is of great significance to identify the authenticity of goods correctly and quickly,for the fight against fake.The anti-counterfeiting code,as a kind of effective anti-counterfeiting technology,is the unique code for the commodity identification,which is stored in the database.Some manufacturers print laser anti-counterfeit code on the packaging as a means of security.Dealers and consumers can get the official information and identify the authenticity of goods by querying the anti-counterfeiting code.Because of the instability of laser marking equipment,camera performance,light condition and the decorative pattern on wrapping paper,the quality of security code images is not good,affected by the complex background,shading and noise interference,low contrast,uneven illumination and adhesion or broken characters.At present,it is not easy to solve these problems with the traditional character recognition technology,which is aimed at the high quality characters.The artificial recognition method is poor in accuracy and low in efficiency.Therefore,it is of great significance to study the recognition technology and application system of low quality anti-counterfeiting image.In this paper,based on the relevant research and development projects,the image enhancement and denoising,character region location,character segmentation and recognition algorithm and the corresponding system for the low quality anti-counterfeiting code image is designed to identify the authenticity of goods.The main research work is as follows.(1)In order to improve the quality of the image,improve the contrast and reduce the noise interference,an image enhancement algorithm based on bilateral filtering and contrast limited adaptive histogram equalization(CLAHE)is proposed.The details and contrast of the image is enhanced by using the bilateral filter with the edge preserving feature,combined with unsharp mask.The noise is suppressed by introducing the local mean square deviation.Besides,the CLAHE algorithm is used to adjust contrast of the low frequency components of the image.In addition,the basic principle of curvature filtering is studied and applied to image denoising.In thispaper,the method is compared with the enhancement method based on Gaussian blur and unsharp mask,the CLAHE algorithm,the retinex algorithm(theory of retina and cerebral cortex),and the range limited bi-histogram equalization(RLBHE)algorithm.The experimental results show that the proposed algorithm can enhance the contrast of low quality images and the edge information of characters while the noise is well suppressed.(2)For the low quality anti-counterfeiting code image,the Susan operator with good anti-noise performance is used to extract the image edge.The sliding window is used in the binarization of the edge image.By analyzing the distribution characteristics of the character and non-character region,the segmentation threshold is calculated to realize the location of the anti-counterfeiting code.In order to adapt to the randomness of the tilt angle of the image,the improved probabilistic Hough transform algorithm is used to realize the tilt correction of anti-counterfeiting code.The experiments indicate that the proposed method can locate the character region from the image which with uneven illumination,noise and background interference.(3)The basic principle of the template matching algorithm is studied and applied to locate the single character of the anti-counterfeiting code,combined with the red component of the image.The template matching method based on gray feature is adopted to recognize the characters of anti-counterfeiting code.For the easily confused characters,the histogram of oriented gradient(HOG)descriptors is extracted to recognize and distinguish them.The test samples with different tilt angle and position deviation are selected.The results show that the proposed algorithm can achieve good classification results.In addition,the recognition and application system of laser anti-counterfeit code is designed to recognize the characters,identify the authenticity of commodities,monitor the running status of the recognition equipment,statistic and push and query the distribution information of fakes.