A high-level tool for the design of custom image processing systems

Abstract

Real-time image processing is a computational intensive task with applications in various engineering fields. In several image processing applications, a significant amount of computing power is committed to image enhancement operations, basic segmentation and identification of regions of interest for further analysis. Such type of front-end processing can be done efficiently by custom data-flow processors closely coupled to an image sensor. This paper proposes a visual design environment to support the high-level design of custom data-flow processors for real-time image analysis applications. The tool is embedded in Matlab/Simulink, and the system modeling is done using a library of blocks that implement common low-level image processing operations. Functional validation is performed efficiently by the simulation engine of Simulink in a frame by frame basis, using the functions provided by the image processing toolbox in Matlab. The automatic generation of a synthesizable RTL model guarantees a logic implementation of the system that complies to the high-level model validated, under constraints imposed by the user and the target reconfigurable device.