Study And Application On Digital Image And Video Retrieval

With the development of multimedia, multimedia information and media form in the field of computer are increasing quickly. Image and video are the most intuitionistic information in multimedia. Image and video retrieval and query is an important aspect in the multimedia processing. The information in database is text information previously, it is multimedia information now. Multimedia information's data quantity especially as image and video's data quantity is immense compare with traditional text information, and multimedia information's configuration is also more complicated. The research of image and video is very important whereas their characteristic.At present, the prevalent method of image and video retrieval is based on the key word, the multimedia in database is described with text and key word, matched with text, this retrieval method required much manpower. CBIR(content-based image retrieval) is rising fast with the development of computer vision, image processing, mode identify and image apperception. As this retrieval method makes use of many characters of image and video itself, the mostly work of CBIR is accomplished by computer, economizing much manpower. This dissertation presents a streets video database system based on Vb and Access, this system provides streets browse, streets query and searches. A CBIR method utilizes color and texture synthetically is introduced in the end, the core codes of Matlab are presented.