摘要
Pictures have become the most important element in a lot of applications such as web sites, discussion forums and digital photo albums. Nowadays, people want to share their own photos with others, and to search interesting pictures on the net. The hand-held devices with camera capabilities also make people to acquire and share their pictures easily. But the huge amount of the pictures grows exponentially, and the difficulty to manage such a huge amount picture archive becomes a time-consuming task. Thus, how to retrieve the correct picture to achieve the efficient searching speed is a challenging research issue.
Although the commonly used image searching engines can search the digital photos with their filename or the EXIF information, but the application is limited because of the lack of the content retrieval capability. Content-based image retrieval (CBIR) attracts lots of attention and has become one of the most active research areas in the past decade. A CBIR system allows users retrieve relevant images not only by the keywords but also by the image content. Numerous applications such as art archives, digital museum and medical image management system, etc…, have been developed.
In this project, we have developed a content-based image retrieval mobile agent which utilizes the hand-held device as the searching client, and implement the content-based image retrieval techniques over the 3G network. The proposed system adopts the web browser as the client interface. The users can take the pictures with their 3G cell-phone, and then transmit the pictures via 3G network to a home image database server. The proposed system also develops an E-mail file uploading function. Users can write their e-mail in any platform such as PC, cell-phone, or PDA to transfer their pictures to the photo album. With this function, an image blog can be built to allow user post their pictures on a web server or share with other friends. They can also add some text tags or voice notes on this picture as a reference. In addition, users can search the image database by using the captured picture from the built-in camera as the query image. That means user can retrieve the similar images by the text tags, the voice notes or the image content from the image database anytime, anywhere.
公告: