Research On Anti-counterfeiting Algorithm Of Micro Texture Infor-mation In Handwritten Signature Modulated Images

With the rise of awareness of property rights protection and the harm of counterfeit and inferior products to economic order,there is an urgent need for practical new technologies in the field of anti-counterfeiting.Therefore,this thesis designs an algorithm to automatically generate miniature images with handwritten signature textures to achieve anti-counterfeiting functions.This algorithm applies non-photorealistic rendering technology and vectorization algorithm to simulate artist handwritten signature drawing images,solving problems in image features,representation of grayscale,artistic effect expression,micro texture,and other aspects.Firstly,to address the differences in expression between contours and backgrounds,a filtering model based on Sobel operator optimization is proposed to obtain the line texture and contours of the original image,and use this as a mask to achieve image segmentation.Secondly,based on the weighted Voronoi algorithm,a high-quality point drawing image of the background image is generated to obtain the relative coordinate data of the point set.Then,a handwritten signature is used as a variable pixel carrier parameter to modulate the background dot image,generating a background image under the definition of the handwritten signature pixel.Among them,the distribution of pixels needs to be constrained by the absolute coordinate values of the converted points,while the size of each pixel changes based on the grayscale integration of Voronoi units,and the direction is adjusted based on the texture features of the image.Then,using local spatial invariance,the fusion of the contour image and the modulated background image is achieved.Finally,using Potrace to vectorize the image and obtain the attributes of the micro texture,the entire algorithm process is completed.After experimental testing,the image generated by this algorithm achieves a similar artistic expression effect compared to the portrait drawn by manual handwritten signature.In addition,as a digital product,the algorithm in this thesis generates highly efficient spatial semantics for the generated miniature images.While ensuring uniqueness and security,it also has unique application value in printing information anti-counterfeiting.Not only can it quantify the value of digital works or physical objects,provide solutions for the ownership of items,but it can also be used as an art collection to achieve anti-counterfeiting or authentication value-added services.