Hardware Approach for Morphological Image Processing Using Xilinx System Generator

The steady evolution of VLSI has led to drastic growth in variety of applications especially for computationally intensive applications in both time and resources like image and video processing. The FPGAs being reconfigurable, have capabilities like enormous resources and fast processing capability. This paper employs Xilinx Block Sets of Simulink to perform morphological operations like erosion and masking on gray scale images. The image erosion removes unnecessary parts and retains only the necessary information of the image while masking selects a region and ignores irrelevant regions. Also, the related parameters assessment is performed using XSG for Kaggle dataset and simplest possible blocks are integrated. The results are evaluated for 28nm Xilinx 7000 series FPGA with part number XC7z020-1CLG484 using Xilinx System Generator and are found to have 30% less resource utilization and 14% improved speed. Hence, the proposed XSG Designs prove to be better options for less computational overheads.