FPGA IMPLEMENTATION OF NOVEL ARCHITECTURES FOR DIGITAL IMAGE PROCESSING CONVOLUTION FILTER: DEVELOPMENT OF NOVEL QUARTILE DIVISION ARCHITECTURES

Image convolution is a very important operation in signal and image processing. It is general purpose filter effect and an algorithm used to filter images. In this the mathematical concepts of convolution and the kernel matrix are used to apply filters to images, to perform functions such as blurring, sharpening, embossing, extracting edges and reducing unwanted noise. Convolution could operate in 1D (e.g. speech processing), 2D (e.g. image processing) or 3D (video processing). Image Convolution finds applications in the specialized areas of Digital Image Processing such as Computed tomography, Image recognition, Imaging Spectroscopy, Artificial Intelligence, feature extraction etc. In this research I have developed 8 different Novel hardware architectures of my own for Convolution Process, which are nowhere in literature before, and first of their kind. All these architectures are designed and implemented first as a prototypic 16 by 16 image and 3 by 3 integer kernels; 32 by 32 image and compared. The programming is done using Verilog HDL and implementation is done on Xilinx Artix-7 FPGAs using Vivado 2015.2 Tool.