White Paper
Standards-Based System for High-End Imaging Applications
High-end image processing has long been the domain of specialized and proprietary hardware. While this approach has delivered the required performance, it comes with some important drawbacks. First is that development of proprietary hardware is expensive and time-consuming. Second is that the software (or firmware) is generally not portable and with every change of hardware platform the software needs to be significantly recoded and re-tested. To a certain extent, the use of proprietary systems has been unavoidable, given the complexity of the required computations and practical limits on processing time. However, as standards-based systems continue to increase in performance, more and more image processing applications become possible on these platforms. By augmenting general-purpose systems with application-specific accelerators it is possible to design very versatile and economical image processing solutions. The advantages of standards-based systems over proprietary ones are significant. Software investments can be preserved from one generation to the next. System designers can tap into a large pool of available hardware design and software programming talent. Finally, component parts are generally available at much lower cost. The end result is that products reach the market faster and at lower cost.
