Justin Varghese

Salt & Pepper Impulse Noise Removal using Adaptive Switching Median Filter

An effective median filter for salt & pepper impulse noise removal is presented. This computationally efficient filtering technique is implemented by a two pass algorithm; In the first pass, identification of corrupted pixels that are to be filtered are perfectly detected into a flag image using a variable sized detection window approach; In the second pass, using the detected flag image, the pixels to be modified are identified and corrected by a more suitable median. Experimental results show that the proposed algorithm performs far more superior than many of the median filtering techniques reported in terms of retaining the fidelity of the image highly corrupted by impulse noises even to the tune of ninety percent impulse noise. The proposed algorithm is free from patchy effects, does not extend black or white blocks in the image as has been found in many other adaptive median based techniques and is very effective in cases when images are corrupted with large percentage of impulse noises. This algorithm works very well for images with lower percentage of impulse noises also.

Selective switching median filter for the removal of salt & pepper impulse noise

In this paper, a new median based filtering algorithm is presented for the removal of impulse noise from digital images. A good analysis of the limitations of the top ranking median filters, the progressive switching median filter, PSMF and the rank-order based adaptive median filter, RAMF is made and are overcome very effectively by the proposed filter which cleans the impulse corruptions of a digital image in two distinct phases of impulse detection and impulse correction. The detection phase identifies the corrupted pixels into a flag image by a spatial rank ordered approach and the correction phase modifies the corrupted pixels identified in the flag image by a more suitable rank ordered value by considering the neighboring features. The filter shows better performance than most other impulse filters, and in particular, that of the top ranking RAMF and PSMF in terms of computational aspect and objectively. The filtering technique gives good results with less patchy effects from images corrupted at higher noise levels and the results are more appealing with good fidelity at lower noise levels

Iterative Adaptive Switching Median Filter

The paper presents an improved median filter for salt & pepper impulse noise removal. This computationally efficient filtering technique is implemented as a two pass algorithm: in the first pass, identification of corrupted pixels that are to be filtered are perfectly detected into a flag image using an iterative fixed sized smaller window approach; in the second pass, using the detected flag image, the pixels to be modified are identified and corrected by a valid median. Experimental results have shown that the proposed algorithm performs far more superior than many of the median filtering techniques reported in terms of retaining the fidelity of the image highly corrupted by impulse noises even to the tune of ninety percent impulse noise. The proposed algorithm is free from patchy effects, does not extend black or white blocks in the image as has been found in many other adaptive median based techniques and is very effective in cases when images are corrupted with large percentage of impulse noises. This algorithm works very well for images with lower percentage of impulse noises

An efficient Switching Median Filter for Salt & Pepper Impulse Noise Reduction

An improved and efficient adaptive median filter is proposed for salt & pepper impulse noise removal from digital images. The more perfect and computationally efficient filtering scheme takes care to restore only the impulse corrupted pixels by a perfect mid-ranking image statistic from an appropriate local neighborhood while keeping the signal content of the uncorrupted pixels as such. The corrupted signals are identified by an adaptive technique and the signal loss due to misinterpretation of uncorrupted signals as corrupted is almost avoided to sustain the image details intact. Experimental results show that the proposed algorithm which detects impulses and restores the corrupted signals in a single efficient stretch is far more superior than many of the other median filters in terms of retaining the fidelity of the images corrupted by salt & pepper impulse noise even to the tune of ninety five percent. The objectiveness of the proposed filter in terms of peak signal to noise ratio and mean absolute error at all noise levels shows the significance of this filter.

An efficient Multiscale Morphological Watershed Segmentation using Gradient and Marker Extraction

A novel segmentation algorithm is proposed in this paper for the segmentation of digital images with either bright or dark or both type of features corrupted at very high impulse noise ratios. The specialized pre-processing set-up completely restores the impulse corrupted image and generates all the information required for perfect segmentation. The algorithm works on the marker extractions of the gradient images obtained by multiscale morphological reconstruction to avoid oversegmentation vivid in watershed algorithm. Experimental results add to the computational efficiency of the algorithm, its shape maintaining, edge preserving and scale-calibrating features. The performance is also superior to most other standard segmentation algorithms

Fuzzy optimal thresholded multiscale morphological segmentation of digital images

A new fuzzy based multiscale morphological segmentation is proposed in this paper. The technique works satisfactorily on gray scale images containing bright and dark features of various scales even in an impulse corrupted environment. The segmentation algorithm involves three passes. In the first pass, the image is preprocessed by using an iterative adaptive switching median filter which reduces the impact of impulse that causes over segmentation. In the second pass the multiple scales of bright and dark features of different objects are extracted by the respective opening and closing of the preprocessed image. The resultant image is binarized using an optimum threshold, obtained by the fuzzy Gaussian measure. The process is repeated for multiple scales of the structuring element until all the features are extracted. In the last pass, valid segments of the bright top-hat and dark bottom-hat images are detected and the contours of these images are combined to give the final segmented image. The scheme is implemented on a set of test images and the performance of the algorithm is proved better both objectively and subjectively than the standard methods. The problems of over segmentation and under segmentation are also addressed by the proposed segmentation technique

Adaptive Rank-ordered Switching Median Filter for Salt & Pepper Impulse Noise Reduction

An adaptive rank-order based switching median filter is proposed in this paper for the restoration of digital images corrupted with salt & pepper impulse noise. A more ideal and efficient impulse noise filter, the ARSMF detects all the corrupted pixels even when the image is highly corrupted and restores them with a more perfect signal detected from an appropriate region whose dimensions vary adaptively with noise levels. This noise replacing signal does the dual functions of noise suppression and detail preservation than many other variants of impulse noise filters. Experimental results, the subjective and objective analysis add to the efficiency and fidelity claim of the proposed ARSMF

An efficient extension to chain codes for external image representation

A novel contour representation algorithm for binary images which is an extension of the conventional chain code is proposed in this paper. The algorithm exploits the features of the conventional chain code when advantageous, is precise and simple with lesser computational complexity than most of the other standard shape representation algorithms. It uses very low bit-rate compared to the conventional chain code and is more robust to noise. The corresponding reconstruction algorithm is completely reversible to give the lossless reconstruction of the contour. Experimental results on various binary images add to the improved efficiency claim of the proposed contour representation algorithm in terms of visual fidelity and bit-rate

An efficient approach to Salt & Pepper Impulse Noise Reduction using Iterative Impulse Detection and Correction Phases

The paper presents an improved median filter for salt & pepper impulse noise removal. The computationally efficient filtering technique is implemented as a two pass algorithm involving identification of corrupted pixels that are to be filtered are perfectly detected into a flag image using an iterative fixed sized smaller window approach and modification of them by a valid median. Experimental results have shown that the proposed algorithm performs far more superior than many of the median filtering techniques reported in terms of retaining the fidelity of the image highly corrupted by impulse noises even to the tune of ninety percent impulse noise.

An efficient approach to data hiding based on lattice vector quantization

This paper describes a robust, oblivious data embedding scheme which uses lattice vector quantization. The embedding is done in the DCT domain. The signature image is vector quantized and the indices obtained are coded using error-correcting codes (BCH codes). The codewords obtained are multiplied by a scaling factor and perturbed into the DCT coefficients of the host image which is lattice vector quantized. Experimental results of the peak signal to noise ratio (PSNR) of the watermarked image for different host images and the PSNR of the recovered signature image for different values of scaling factors stand a testimony for the validity of this method