Development and Implementation of VLSI Reconfigurable Architecture for Gabor Filter in Medical Imaging Application

The Gabor filter is a very effective tool in visual search approaches and multimedia applications. This filter provides high resolution in time-frequency domains and thus finds use in object recognition, character recognition and pattern recognition applications. Medical Image analysis using image processing algorithms is one of the best ways of diagnosing diseases inside human body. The Gabor wavelets resemble the visual cortex cell operation of mammalian brains and hence are best suited for biological image analysis. A Tonsillitis detection system is proposed here using Gabor filtering approach. This system detects the presence of Tonsillitis from the tonsils images. A suitable VLSI architecture for the implementation of the Gabor filter was modeled in Verilog using Xilinx tool and simulated using the tonsils test images. The proposed system was successful in detecting the presence of Tonsillitis from the diseased tonsils image. The complete system was then synthesized and implemented on FPGA Artix 7. The design was capable of operating at a maximum frequency of 394.563 MHz.