The Ceramic Certification Method Based On Image Matching And Blockchain Storage

The use of ceramic products has greatly facilitated people’s lives,and the distinctive ceramic vessels not only have practical value,but also a certain collection value.While some ancient ceramics and modern art ceramics also bear the responsibility of spreading historical culture and ceramic culture.However,due to interest-driven,both ancient and modern ceramic products are subject to imitation or forgery.In order to prevent the imitation of ceramics from causing damage to ceramic culture and the profit of consumers,researchers have combined the knowledge of various disciplines to constantly optimize and update the identification methods of ceramics.Since the physical and chemical identification techniques used in traditional ceramic identification methods often cause irreversible damage to the ceramics themselves,more and more researchers are using modern artificial intelligence methods,such as image recognition technology,to try to identify and classify ceramics by their shape,pattern,and other characteristics.However,these external macroscopic features of ceramics are easy to imitate and sometimes difficult for professionals to distinguish,and the changes in camera conditions have a large impact on them,making it more difficult for computer vision methods to distinguish the specific identity of ceramics with these features.Based on preliminary research and empirical evidence,this paper attempts to find new perspectives for computer vision methods to recognize the specific identity of ceramics.The research shows that from the microstructure analysis of ceramic materials,the arrangement of tiny pores on their surfaces is randomly generated during the firing process of ceramics,and has stable and unique statistical features and topology.Therefore,this paper proposes a ceramic authentication method of image matching from the perspective of microscopic images,and demonstrates through experimental comparison that microscopic image features are more accurate compared with macroscopic image features of ceramics,and can match to specific ceramics.And the development of image matching is reviewed in depth,and a variety of classical image matching algorithms are applied to ceramic microscopic image matching with better results,showing the scientificity and superiority of authenticating ceramic identity from microscopic image matching to start with.On the other hand,image matching needs to extract the feature data of both the template image and the target image to be matched,while the task of ceramic authentication requires network storage of the template image and its features,which requires that the storage network used must have strong security so as to ensure that the saved image and feature information are not tampered with.Establishing a suitable storage network to achieve effective anti-counterfeit traceability is the main task of the second phase of the research work in this paper.After a comparative analysis of the traditional centralized storage and the emerging distributed storage,this paper introduces the principles of distributed storage networks such as blockchain,chooses to use a blockchain-based storage network as the storage network for ceramic information,and justifies the reasonableness of this choice.Finally,in view of the shortage of storage capacity of blockchain,this paper proposes a storage network using "IPFS（Inter Planety File System）+ blockchain" and verifies its feasibility and access rate through testing.