21-08-2014, 04:12 PM
This report presents architecture for different type of image proessing techniques using Xilinx System Generator (XSG) Matlab and VHDL. Simulink offers an alternative that combines MATLAB, Simulink and Xilinx System Generator and explores important aspects concerned to hardware implementation. This report focuses on Enhancement of image in spatial domain which is the method of processing an image directly on the pixels. Different types of image enhancement techniques such as image brightening, image negative, image averaging, image sharpening, image zooming etc have been covered here. The fundamental of all the image processing techniques is reading the image pixel by pixel and then process each pixel depending on its value or the value of its neighbors. For image averaging we pass the image from a low pass filter and for image sharpening we pass the image from high pass filter. Similarly, by passing the image through different filters we get different results.In the end, edge detection using Sobel and Prewitt methods have been discussed which are very popular edge detection methods. Sobel and Prewitt method finds edges using the sobel and Prewitt approximation, respectively to the derivative. It returns edges at those points where the gradient of I(graph of image intensity) is maximum. After finding the edges of the image we can also find lines in the image. For this problem hough transform has been explained as it is widely used for finding the lines in the image.After the designing is over, we can implement the design on various devices such as Sea of gates, FPGA, PLA. Using FPGA is benificial for implementing the hardware, as it many advantages over the other devices. We can also implement the design in microblaze which is inbuit in FPGA kit. The area that is dedicated to Microblaze works very similar to a processor and is very easy to design.