System-level memory management for weakly parallel image processing

Abstract

Application studies in the domain of image and video processing indicate that between 50 and 80% of the area cost in (applicationspecific) architectures for multi-dimensional (M-D) signal processing is due to memory units. This is true for both single-processor and weakly parallel processor realizations. This paper has two main contributions. First, to reduce this dominant cost, we propose to address the systemlevel storage organization for the M-D signals as a first step in the overall methodology to map these applications. Secondly, we will demonstrate the usefulness of this novel approach based on a realistic image processing test-vehicle, namely a cavity detection algorithm. The novel design results for this relevant application are useful as such.