The Research Of Image Retrieval Algorithm By Combination Local Feature And Global Feature

In the wake of the arriving of the Internet plus period,the relationship between the network and people's learning ? life and work has a growing connection,at the same time,also causing an rapidly increasing in the number of images.So,how to search image from a large image database quickly and accurately become a very meaningful and challenging topic.Currently,image retrieval technology is divided into two categories: Text Based Image Retrieval(TBIR)and Content Based Image Retrieval(CBIR).Feature extraction is the most important part of CBIR research.The features of an image can generally be described by local features or global features,global features can take into account the influence factors of each part of the image,and local features are more able to reflect the details of the information.However,image content is so complex and varied that,it is very difficult to retrieve the image efficiently and precisely with only a single type of feature.To solve this issue,this paper proposed a new image retrieval algorithm based on combine the local features and global features.The main research works are included as follows:1.Study the local features and global features of images.Aimed at the local features,selects the most representative SIFT features,and focuses on the generation of an image feature descriptor.For global features,we discuss the color features after quantifying and improved texture features respectively.2.In view of the problems that a lot of the SIFT descriptors are described by using the gray information of the image,can't differentiate the objects with very similar shape but with different colors,the same shape but with different backgrounds commendably.The paper proposes an improved SIFT features descriptor,and combined color features with it to retrieval image.3.Aiming at the problem that SIFT algorithm is easily influenced by image noise and brightness change in image feature extraction and retrieval.In this paper,we propose an image retrieval algorithm by combination texture LBP algorithm and local SIFT algorithm.Through the simulation experimental results show that the fusion algorithms proposed in this paper has a good retrieval performance compared with the algorithm of single feature image retrieval,and the image retrieval algorithm based on local feature and global feature is proved accurate and effective.