Edge Preserving Techniques for Efficient Removal of Impulse Noise Impulse noise can corrupt the image signals in the process of signal acquisition and transmission. VLSI design needs low computational complexity and two line memory buffers. The cost of hardware is quite low. This new design has a better image quality with low cost.
The impulse noise corrupts the image in the applications such as medical imaging, printing skills, image segmentation, face recognition, and scanning techniques acquisition and transmission. Therefore, denoising efficient technique plays an important role in the image processing applications. Different image denoising methods employ the standard median filter to implement denoising process however with these approaches the image is blurred since both noisy and noise-free pixels are modified.
The switching median filter generally has two steps. They are impulse detection and noise filtering. In general, the denoising methods for impulse noise suppression has classified into two techniques. First one is lower-complexity techniques and second one is higher-complexity techniques.
Due to simplicity and easy implementation of denoising with the VLSI circuit, denoising has the lower-complexity techniques. The proposed new impulse detector (NID) for switching median filter needs the minimum absolute value of four convolutions obtained by 1-D Laplacian operators to detect noisy pixels. Other impulse detectors are DRID, ERIN, ATMBM, DBA, and SEPD.
The extreme data detector detects the minimum and maximum luminance values (MINinW and MAXinW ) from the first to the current one in the image. If fixed-value impulse noise corrupts a pixel then luminance value will become minimum or maximum value in gray scale.
Conclusion:
This experiment demonstrates our design has an excellent performance in quantitative evaluation and visual quality thought the noise ratio is 90%. As per real-time applications, 7-stage pipeline architecture for SEPD and 5-stage pipeline architecture for RSEPD are implemented and developed. RSEPD like other chips has lowest hardware cost and works with monochromatic images and extended with RGB color images.
Click Here To Download Edge Preserving Techniques for Efficient Removal of Impulse Noise Electronics and Communication Engineering ECE Final Year Project Report.