FPGA implementation of spatial filtering techniques for 2D images

Various filtering techniques have been the core of image processing since the very inception of image enhancement techniques. Spatial domain filtering in image processing deals with various practical applications like image sharpening, blurring, noise removal etc. Former refers to the filter dealing with high frequencies while the latter is resulted due to low frequency operations. The versatility of spatial filtering is more compared to transform domain as it can be used for both linear and nonlinear filtering. Image smoothing is realized directly by manipulating the intensity values of the original image by the application of spatial masks or windows. It results in reduction of small details and noise from an image. This paper proposes hardware realization of various spatial domain masking operations. The results are based on the hardware consumption, speed of the design of each architecture. Measure of the image quality is given by comparing the outputs of both MATLAB & Xilinx FPGA and then by calculating the MSE. Practical implementation all the techniques are represented and explained through appropriate outputs.