A VHDL Implementation of Effective Impulse Noise Filtration Technique
Sarala Singh · 2013
Abstract— Digital imaging is now replacing the traditional imaging techniques because of its smaller size and low cost. The digital imaging provided many advantages over traditional techniques the however the digital imaging system may suffer from number of problems caused by sensing elements to the communication channel one of them is impulsive noise. The filtration of such type of noise is not very difficult but for real time applications it is required to be performed in very short time duration. This paper presents a FPGA based median filter architecture which filters the image without losing the edge information. The FPGA and CPLD's are gaining applications in all fields of engineering because of their high speed parallel operations (unlike microprocessor which execute operations sequentially). Because of its operating speed it is especially preferred for the heavy and time critical processing tasks such as real time image processing. The digital image contains millions of bytes of data and in real time applications it is needed to be processed in very limited time (frame interval). This paper presents a fast and efficient processing architecture based on FPGA for the filtration of the impulsive noise. Traditionally, the impulse noise is removed by a median filter which is the most popular nonlinear filter. Its hardware implementation is straightforward and does not require many resources. However, the standard median filter gives a poor performance for images corrupted by impulse noise with higher intensity. A simple median utilizing 3×3 or 5×5-pixel window is sufficient only when the noise intensity is less than approx. 10-20%. When the intensity of noise is increasing, a simple median filter remains many shots unfiltered (1). To overcome the limitation of the median filter, numbers of advancements have been proposed which not only enhance the performance but also optimized to efficiently utilize the FPGA features. The rest of the paper is arranged as that the second section presents a brief introduction to impulsive noise followed by the basic median filter architecture in third section. The fourth section presents the proposed architecture and fifth chapter presents the RTL test analysis waveforms followed by conclusion in sixth section.