A framework for grid-based mobile flower recognition

Identifying an organic object like a flower against the background picture is known to be challenging. Several methods have been proposed so that unique characteristics of flowers can be extracted with increased accuracy. Establishing this process as a mobile flower identification service for the public requires major adaptation of the system due to numerous picture requests being received simultaneously. This paper presents a framework for grid-based mobile flower recognition that employs 1) grid computing technology for further optimization of multiple concurrent image processing and 2) mobile wireless communication as the primary data transfer for accessing this mobile service. Globus Toolkit is applied as the middleware in the distributed computing environment whereby Short Message Services (SMS) and Multimedia Messaging Services (MMS) technology are used for picture and result delivery.