In image retrieval systems, user information needs is exressed using multiple types of query unfortunately, due to user subjectivity perception to visual features and semantic depths of images, the conventional query submitted to the system encounter difficulties to identify user information need. The blooming of interest in semantic image retrieval requires current research direction to be more concerned into semantics. This paper describes our on-going research work in formulating a new query approach for image databases. The enabling technologies of the semantic web formed the building blocks of our query specification. Using the MRML as the communication protocol, XML in the form of SVG as the visual content description (sketch based) and RDF for binding the semantic meanings of images, this research provides an initial framework towards semantic based query formulation framework.