Implementation of a large data processing method for embedded system and CMOS SNR application

Abstract

The embedded system is the future trends of instrument and the larger memory capacity of embedded system is favor to different variety of applications. The commercial embedded systems are always restricted by their embedded memory capacity and required to be upgraded, especially in image application. This article reports a new design and development of an embedded system built in a CMOS image SNR measurement instrument. The new developed approach using the mix technique of large data processing (MLDP) method for CMOS SNR calculation is described. The MLDP method uses an external memory device as auxiliary memory in the regular embedded system to break the memory capacity limitation. The experimental results show the new method is applied successfully in CMOS SNR measurement and the calculated speed is increased almost 200 times compared to that of the traditional method even thought the processing data size is over the embedded system memory. © 2011 Springer-Verlag.