Algorithmic and Architectural Optimization of a 3D Reconstruction Medical Images Algorithm for a Hardware Implementation
This paper presents an optimization of an FPGA circuit implementation of 3D reconstruction algorithm of medicals images. It is based on an algorithmic specification in the shape of a Factorized and Conditioned Data Dependences Graph (GFCDD). An automatic and optimized implementation of the algorithm of marching cubes has been carried out. The repetitive property of the algorithm has been exploited, as much as possible, by means of the methodology "Adequacy Algorithm Structures". The 3D reconstruction consists in providing a volumetric representation of an object from a set of minimum information to ensure the description of the 3D volume.