Anti-counterfeiting Identification Technology And System Based On Surface Texture Image And Appearance Feature Image

Nowadays,many people are puzzled by such problems as fake contracts,fake diplomas,fake invoices,fake calligraphy and paintings,and fake wills,which have caused great harm to the country,society and individuals.In recent years,the emergence of digital image technology has brought convenience to distinguish the texture patterns of the real and fake objects.The existing texture anti-counterfeiting identification technology mainly depends on the randomness and non-replicability of the texture of the article base material,and it is only suitable for the paper articles with light transmission,but it can't do anything for the articles with non-random distribution texture and non-light transmission,and the degree of automation of texture identification relying on human eyes is not deep.In order to solve the shortcomings of the existing technology,this paper studies and proposes an object authenticity identification technology based on the surface texture image and appearance feature image.The experiment proves the effectiveness and feasibility of this technology.The technology proposed in this paper has no damage to the goods,high degree of automation in authenticity identification,convenient and fast use,wider application range and large-scale application potential.In this paper,an anti-counterfeiting identification method based on the uniqueness of the materials used by the objects is proposed by using the surface texture image and the appearance feature image of the objects.This method is not only applicable to randomly distributed texture materials,but also effective for non-randomly distributed texture materials.It is not only applicable to transparent objects,but also to non-transparent objects.In this paper,the light source small angle forward illumination is used to identify the object,and the camera captures the appearance characteristic image to highlight the macro appearance characteristic of the object;the light source large angle grazing illumination is used to identify the object,and the camera captures the surface texture image of the opaque object to highlight the fine texture structure of the object surface;the light source uniform back illumination is used to identify the object,and the camera captures Obtain the surface texture image of transparent objects to highlight the fine texture structure inside the surface of the objects,and discuss the optical imaging principle of these photographed appearance feature images and surface texture images.In order to realize the automatic authenticity identification,the paper develops the image processing and analysis algorithm design,including the registration algorithm of the appearance feature image of the authentic and the items to be identified,the identity identification algorithm of the texture image of the authentic and the items to be identified,etc.Based on the similarity of the appearance features of the real products and the imitated products,it is proposed to use the appearance features of the products as the registration benchmark of the real images and the images to be identified,which improves the accuracy of image registration.The correlation coefficient of the texture image after registration is used as the basis to distinguish the texture identity of the real and fake objects,which makes the identification of the texture identity independent of the random distribution of the texture,and expands the scope of application of the anti-counterfeiting technology based on the texture.The experimental system of anti-counterfeiting identification is established,which mainly includes the optical system of object surface texture image and appearance feature image shooting,image shooting control system and graphical user interface.In order to verify the effectiveness of the proposed anti-counterfeiting identification technology,the established experimental system was used to carry out a variety of different material identification experiments.The objects selected in the experiment include paper bills with good transparency,paper documents,Chinese calligraphy works,etc.,as well as the back cover of degree thesis and oil painting canvas with no transparency.The objects selected include randomly distributed texture materials and non-randomly distributed texture materials.In the experiment,the translation and rotation,zooming in and out,and different light intensity of the object image to be identified that may occur in the identification process are also considered.The experimental results show that the anti-counterfeiting identification system can automatically identify the authenticity of the goods.