Research And Application Of Content-based Image Retrieval Technology

With the continuous development of big data industry,the digital information has developed in an explosive manner.Under this context,the image search engine technology emerges.Since the text-based image search engine technology depends too much on the image annotation information,it was replaced by the content-based image search engine technology.Combing with the content-based image search technology,this paper proposes a new algorithm in order to improve the seep and robustness of image search.The main contents of this paper are as follows:(1)The content-based image search technology is discussed and the twin-histogram algorithm based on Hash is utilized to obtain the global information of image combining with the fast coding of Hash in image search technology.By reducing the complexity of the image features and limiting the search range of target image,the search speed is improved.(2)In order to compensate for the hash code is not sensitive to the local characteristics of the shortcomings,on the basis of the first step,the Surf algorithm is selected to make the second selection and comparison of image information.So that in the image search comparison,from the global information to the local feature points to compare the two steps to complete the image matching work,in order to ensure the efficiency of the image search at the same time,improve the robustness.(3)The study is conducted by artificially changing the brightness and color of the target picture,transforming the visual angle and the rotation scale and fuzzificating the watermarking as well as by analyzing the speed of the system matching the image and its noise immunity.Compared with the Sift algorithm and the Surf algorithm,it is proved that the algorithm in this paper can greatly shorten the matching time and achieve good results while ensuring the completeness of global information and local feature of the picture.And its application are used by face data experience project in Guizhou Science and Technology Museum's big data exhibition area.