Design And Implementation Of Content Based Image Retrieval On The Computing Platform

With the instant development of multimedia and internet techoology, the amount ofdigital images is increasing with an astonishing speed. In order to make full use of themassive image data, the content-based image retrieval (CBIR) has been studiedextensively. However, there are two draowbacks in the current conten-base imageretrieval systems: one is the impossibility of the traditional input methods to meet theneed of users’ complex and combined retrieval; the other is the little opportunity for theconventional stand-alone operating environment to fill the speed requirements of massimages retrieval.In this paper, an image retrieval system with high speed and effective isconstructed based on the research of CBIR and the cloud computing platform. Itprovides users a new Image Query Language (IQL), which with the image as itsretrieval elements is similar to SQL, to meet the need of their complex retrieval, as aresult the high retrieval precise and satisfactory of users are reached. Furthermore, itimproves the efficiency of image retrieval greatly, by integrating the conventional imageretrieval technology into Hadoop (the open source cloud computing platform). Finally,the experimental results show that the effectiveness and efficiency of the system meetthe design requirements.