Vladimir Zlokolica

Wavelet-Domain Video Denoising Based on Reliability Measures

This paper proposes a novel video denoising method based on nondecimated wavelet band filtering. In the proposed method, motion estimation and adaptive recursive temporal filtering are performed in a closed loop, followed by an intra-frame spatially adaptive filter. All processing occurs in the wavelet domain. The paper introduces new wavelet-based motion reliability measures. We make a difference between motion reliability per orientation and reliability per wavelet band. These two reliability measures are employed in different stages of the proposed denoising scheme. The reliability per orientation (horizontal and vertical) measure is used in the proposed motion estimation scheme while the reliability of the estimated motion vectors (MVs) per wavelet band is utilized for subsequent adaptive temporal and spatial filtering. We propose a novel cost function for motion estimation which takes into account the spatial orientation of image structures and their motion matching values. Our motion estimation approach is a novel wavelet-domain three-step scheme, where the refinement of MVs in each step is determined based on the proposed motion reliabilities per orientation. The temporal filtering is performed separately in each wavelet band along the estimated motion trajectory and the parameters of the temporal filter depend on the motion reliabilities per wavelet band. The final spatial filtering step employs an adaptive smoothing of wavelet coefficients that yields a stronger filtering at the positions where the temporal filter was less effective. The results on various grayscale sequences demonstrate that the proposed filter outperforms several state-of-the-art filters visually (as judged by a small test panel) as well as in terms of peak signal-to-noise ratio

Salient Motion Features for Video Quality Assessment

Design of algorithms that are able to estimate video quality as perceived by human observers is of interest for a number of applications. Depending on the video content, the artifacts introduced by the coding process can be more or less pronounced and diversely affect the quality of videos, as estimated by humans. While it is well understood that motion affects both human attention and coding quality, this relationship has only recently started gaining attention among the research community, when video quality assessment (VQA) is concerned. In this paper, the effect of calculating several objective measure features, related to video coding artifacts, separately for salient motion and other regions of the frames of the sequence is examined. In addition, we propose a new scheme for quality assessment of coded video streams, which takes into account salient motion. Standardized procedure has been used to calculate the Mean Opinion Score (MOS), based on experiments conducted with a group of non-expert observers viewing standard definition (SD) sequences. MOS measurements were taken for nine different SD sequences, coded using MPEG-2 at five different bit-rates. Eighteen different published approaches related to measuring the amount of coding artifacts objectively on a single-frame basis were implemented. Additional features describing the intensity of salient motion in the frames, as well as the intensity of coding artifacts in the salient motion regions were proposed. Automatic feature selection was performed to determine the subset of features most correlated to video quality. The results show that salient-motion-related features enhance prediction and indicate that the presence of blocking effect artifacts and blurring in the salient regions and variance and intensity of temporal changes in non-salient regions influence the perceived video quality.

Automatic functional TV set failure detection system

Enormous technological innovation in recent years has put a great challenge upon manufacturers in consumer electronics industry: how to efficiently and accurately verify the functionality of a product handed to a customer. To address this problem, we propose a fully automated system for functional failure detection in TV sets. In the proposed framework, the automatic assessment of TV set functionality is performed by processing captured images of the TV set under inspection and the captured images from the reference TV set (which is considered to be with no functional failures). A new algorithm for image similarity is proposed and incorporated in the system for evaluating the difference between the tested image and the reference one, in respect to specific degradations that can occur in TV sets. The similarity detection algorithm is a full-reference intra-frame video quality assessment scheme, which is based on spatially local mean square error and variance between the reference and the tested image. Based on the similarity detection, an acceptance decision for the TV set functionality is made. The proposed embedded system for automatic TV functional failure detection includes (i) central control unit through which the testing methodology is carried out: The TV set under inspection is fed by the test video signals defined by the (ii) specified test-case-scenario, while the picture is simultaneously captured from the tested TV set and compared by the (iii) proposed similarity detection algorithm to the reference one, for the same type of TV set. Finally, the results are stored in the (iv) data base for TV set debugging purposes. The usability and effectiveness of the proposed methodology has been experimentally evaluated and detailed analysis of the results has been reported. The proposed similarity detection algorithm has shown to be superior to the other state-of-the-art image quality assessment algorithms in terms of detecting TV picture degradations (such as picture misalignment, illumination change, aliasing, blurring, etc.) in captured images from TV set, under the methodology which incorporates defined test-case-scenario.

Video Denoising Using Multiple Class Averaging with Multiresolution

This paper presents a non-linear technique for noise reduction in video that is suitable for real-time processing. The proposed algorithm automatically adapts to detected levels of detail and motion, but also to the noise level, provided it is short-tail noise, such as Gaussian noise. It uses a one-level wavelet decomposition, and performs independent processing in four different bands in the wavelet domain. The non-decimated transform is used because it leads to better results for image/video denoising than the decimated transform. The results show that from both a PSNR and a visual quality, the proposed filter outperforms the other state of the art filters for different image sequences.

Motion and detail adaptive denoising of video

Non-linear techniques for denoising images and video are known to be superior to linear ones. In addition video denoising using spatio-temporal information is considered to be more efficient compared with the use of just temporal information in the presence of fast motion and low noise. Earlier, we introduced a 3-D extension of the K-nearest neighbor filter and have investigated its properties. In this paper we propose a new, motion- and detail-adaptive filter, which solves some of the potential drawbacks of the non-adaptive version: motion caused artifacts and the loss of fine details and texture. We also introduce a novel noise level estimation technique for automatic tuning of the noise-level dependent parameters. The results show that the adaptive K-nearest neighbor filter outperforms the none-adaptive one, as well as some other state-of-the-art spatio-temporal filters such as the 3D alpha-trimmed mean and the state-of-the-art rational filter by Ramponi from both a PSNR and visual quality point of view.

Multiresolution Denoising for Optical Coherence Tomography: A Review and Evaluation

Recently emerging non-invasive imaging modality - optical coherence tomography (OCT) - is becoming an increasingly important diagnostic tool in various medical applications. One of its main limitations is the presence of speckle noise which obscures small and low-intensity features. The use of multiresolution techniques has been recently reported by several authors with promising results. These approaches take into account the signal and noise properties in different ways. Approaches that take into account the global orientation properties of OCT images apply accordingly different level of smoothing in different orientation subbands. Other approaches take into account local signal and noise covariances. So far it was unclear how these different approaches compare to each other and to the best available single-resolution despeckling techniques. The clinical relevance of the denoising results also remains to be determined. In this paper we review systematically recent multiresolution OCT speckle filters and we report the results of a comparative experimental study. We use 15 different OCT images extracted from five different three-dimensional volumes, and we also generate a software phantom with real OCT noise. These test images are processed with different filters and the results are evaluated both visually and in terms of different performance measures. The results indicate significant differences in the performance of the analyzed methods. Wavelet techniques perform much better than the single resolution ones and some of the wavelet methods improve remarkably the quality of OCT images.

Noise estimation for video processing based on spatio-temporal gradients

We propose an efficient and accurate wavelet-based noise estimation method for white Gaussian noise in video sequences. The proposed method analyzes the distribution of spatial and temporal gradients in the video sequence in order to estimate the noise variance. The estimate is derived from the most frequent gradient in the two distributions and is compensated for the errors due to the spatio-temporal image sequence content, by a novel correction function. The spatial and temporal gradients are determined from the finest scale of the spatial and temporal wavelet transform, respectively. The main application of the noise estimation algorithm is in wavelet-based video processing. The results show that the proposed method is more accurate than other state-of-the-art noise estimation techniques and less sensitive to varying spatio-temporal content and noise level.

Robust non-linear filtering for video processing

Noise removal techniques, such as the K-nearest neighbour filter and the /spl alpha/-trimmed mean filter, are known to be very robust in still image noise removal, but they have not been exploited in video processing. We investigate their 3D-extension for use in video sequence noise removal. We also determine the optimal balance between temporal and spatial window size, the optimal values of the other parameters and finally we investigate the artefacts introduced by the filters. The results show that the new video K-nearest neighbour filter outperforms the video version of the /spl alpha/-trimmed mean and the state-of-the-art rational filter by G. Ramponi from both a PSNR and a visual quality point of view.

Recursive temporal denoising and motion estimation of video

In this paper, we present a new technique for video denoising which is based on a novel motion estimation algorithm. First, a recursive temporal denoising is performed through the estimated motion trajectory. After that, appropriate spatial filtering is done. The proposed algorithm automatically adapts to the detected noise level, provided it is short-tail noise, such as Gaussian noise. It uses a one-level wavelet decomposition where both motion estimation and denoising is performed. The non-decimated transform is used because it is nearly shift invariant and thus yields better motion estimation and denoising results than the decimated transform. The results on different image sequences demonstrate that the proposed filter outperforms the other state-of-the-art filters both in terms of PSNR and visual aspect.

Video denoising by fuzzy motion and detail adaptive averaging

A new fuzzy-rule-based algorithm for the denoising of video sequences corrupted with additive Gaussian noise is presented. The proposed method constitutes a fuzzy-logic-based improvement of a recent detail and motion adaptive multiple class averaging filter (MCA). The method is first explained in the pixel domain for grayscale sequences, and is later extended to the wavelet domain and to color sequences. Experimental results show that the noise in digital image sequences is efficiently removed by the proposed fuzzy motion and detail adaptive video filter (FMDAF), and that the method outperforms other state of the art filters of comparable complexity on different video sequences.