Design And Implementation of Retrieval of Images Based on Color, Text And Shape

Abstract

Retrieving the image based on the content is one of the most significant research fields in the digital image processing area. The image retrieval is becoming an innovative and quickest way of retrieving the image. Many digital images and videos are uploaded and transitioned everyday, because of the social networking medium development. Google, Yahoo and many search engines are accessing the large collection of database based on the textual annotation. These images are annotated manually with keywords, descriptions and then the images are retrieved using text-based searching methods. Therefore, performances of these systems are not satisfactory. So, we need new techniques. These techniques will easy to handle the data and can easily access the data. The CBIR method is working on the basis of particular image feature such as color, texture and shape which are derived automatically. It uses digital processing and analysis to automatically generate descriptions directly from the media data. This paper represents how the implementation of image retrieval is done based on the content.