Image Retrieval System Based On Mixed Programming Of MATLAB And C#

In recent years, the capacity of the world’s digital images with the rapid development of computer network and multimedia technology and growing at an alarming rate. These digital images contains a great deal of useful information, they disorderly distributed in various parts of the world, resulting in a large number of these digital image information can not be accessed and utilized effectively. Research based on the traditional workload text retrieval method, the expression of the image information is not accurate enough, the lower the degree of retrieval, in this background, it is proposed content-based image retrieval. Content-based image retrieval features are a small amount of work, time to reduce image retrieval, image feature extraction data is more accurate, high degree of retrieval. In the field of computer vision, image retrieval is a content-based image retrieval technology, an important research direction, has a very broad application prospects in the future development of the society.Based on previous studies, in-depth study of the method of design content-based image retrieval system, under the premise of low-level feature extraction, integration of low-level features as a new feature, and MATLAB mixed programming with C# improve retrieval efficiency. Specific studies are as follows:(1)Describes the background and significance of the development of content-based image retrieval system, through the development of domestic and international research, analysis of the advantages and disadvantages of different methods based on the content of image retrieval, as well as a variety of methods.(2)Discuss the structure and the general flow of content-based image retrieval system. Analysis of low-level image feature extraction and analysis of image retrieval similarity matching methods.(3)Establish an image database and image database indexing in MATLAB. Choose a color histogram feature extraction and edge histogram feature extraction methods and to establish the corresponding feature database. Proposed color histogram features and edge histogram feature weighting methods and select the appropriate similarity matching algorithm.(4)The system provides an interface for the basic query a user-friendly interface, using.NET FrameWork 4.0 environment Windows Forms application and call some of the data in the MATLAB environment to achieve, to achieve a content-based image retrieval system.Through a large number of experimental results and data analysis showed that low-level feature extraction retrieval results were better under certain circumstances into account, the overall effect of the weighted low-level feature is better, better able to retrieve images associated with the query image.