Abstract | Keywords | Fulltext | Export | TOC |

The explosive growth of image data leads to the need of research anddevelopment of Image retrieval. Image retrieval researches are moving fromkeyword, to low level features and to semantic features. Drive towards semanticfeatures is due to the problem of the keywords which can be very subjective andtime consuming while low level features cannot always describe high levelconcepts in the users’ mind. This paper is proposed a novel technique for objectsspatial relationships semantics extraction and representation among objectsexists in images. All objects are identified based on low level features extractionintegrated with proposed line detection techniques. Objects are representedusing a Minimum Bound Region (MBR) with a reference coordinate. Thereference coordinate is used to compute the spatial relation among objects.There are 8 spatial relationship concepts are determined: “Front”, “Back”, “Right”,“Left”, “Right-Front”, “Left-Front”, “Right-Back”, “Left-Back” concept. The userquery in text form is automatically translated to semantic meaning andrepresentation. Besides, the image similarity of objects spatial relationshipssemantic has been proposed.

Semantic Gap --- Objects Spatial Relationships Semantic --- Automatic Image Semantic Extraction --- Image Retrieval