The localization of interference associated with crossing modes in thetime-frequency(TF) plane is of significant interest when the objective is to separate these modes. Prior to localizing mode crossings, it is essential to determine whether the observed interference arises from closely spaced modes or from actual mode crossings. In this paper, we propose a novel technique to discriminate between these different types of interference in the TF plane and to subsequently localize mode crossings. The proposed approach exhibits strong adaptability across a wide range of scenarios and demonstrates robustness to noise, as confirmed by simulation results.

Frequency Modulated (FM) signals are usually characterized by two or more components having different waveforms and time-dependent frequencies. The estimation of the instantaneous frequency (IF) represents a very challenging topic when there are two or more interfering components. In this case, available approaches, like the reassignment-based ones, fail in correspondence to the interference region, where IFs of each individual mode are not far enough to be considered separable. In this paper, a novel approach for IF estimation based on spectrogram analysis is proposed. Firstly, it is proved that the projection of signal spectrogram on the temporal axis is itself an amplitude and frequency modulated signal. Its IF depends on the instantaneous frequencies of the individual signal components and the functional dependence can be explicitely given. Then, under proper assumptions, a regression method for the individual IFs estimation is provided using a few selected 'good' points lying in the interference region. The regression method benefits from a penalized-spline (P-spline) model where a second-order discrete penalty term is also used. Experimental results show that the proposed model is very effective to estimate the individual IFs of multicomponent FM signals in the interference region, making it a suitable method for correcting reassignment estimation where the separability condition is not satisfied.

Abstract This paper presents a method for Photo Response Non Uniformity (PRNU) pattern noise based camera identification. It takes advantage of the coherence between different PRNU estimations restricted to specific image regions. The main idea is based on the following observations: different methods can be used for estimating PRNU contribution in a given image; the estimation has not the same accuracy in the whole image as a more faithful estimation is expected from flat regions. Hence, two different estimations of the reference PRNU have been considered in the classification procedure, and the coherence of the similarity metric between them, when evaluated in three different image regions, is used as classification feature. More coherence is expected in case of matching, i.e. the image has been acquired by the analysed device, than in the opposite case, where similarity metric is almost noisy and then unpredictable. Presented results show that the proposed approach provides comparable and often better classification results of some state of the art methods, showing to be robust to lack of flat field (FF) images availability, devices of the same brand or model, uploading/downloading from social networks.

This paper deals with the {\em typical set} of an image quality assessment (IQA) measure. In particular, it focuses on the well known and widely used Structural SIMilarity index (SSIM). In agreement with Information Theory, the {\em visual distortion typical set} is composed of the least amount of information necessary to estimate the quality of the distorted image. General criteria for an effective and fruitful computation of the set will be given. As it will be shown, the typical set allows to increase IQA efficiency by considerably speeding up its computation.

The objective of the paper is to embed perception rules into the kernel-based target tracking algorithm and to evaluate to what extent these rules are able to improve the original tracking algorithm, without any additional computational cost. To this aim, the target is represented through features that are related to its visual appearance; then, it is tracked in subsequent frames using a metric that, again, correlates well with the human visual perception (HVP). The use of HVP rules are twofold advantageous: it allows us to both increase tracking efficacy and considerably reduce the computational cost of the tracking process-thanks to the reduced size of the perceptual feature space. Various tests on video sequences have shown the stability and the robustness of the proposed framework, also in the presence of both other moving objects and partial or complete target occlusion in a limited number of subsequent frames.

This paper proposes a novel approach for image and signal denoising that does not need any classical regularization. It subtracts a sorted realization of noise to the sorted noisy signal. The similarity between the noise that corrupted the signal and the selected noise realization allows us to denoise monotonic signals. The Minimum Description Length (MDL) is then adopted to get a piecewise monotonic representation of the original signal. Experimental results show that the proposed approach outperforms most of the classical denoising approaches, even though it is based on very simple operations.

This paper focuses on the link between a visible linear and local distortion in a pictorial scene and its cost in terms of information theory quantities. In particular, a formal relation between the Michelson visual contrast and the Jensen-Shannon divergence (JSD) will be provided. A universal just noticeable threshold is also derived by maximizing JSD constrained to the greatest Michelson contrast of an invisible distortion. Such a threshold is independent of both distortion parameters and the probability density function of the degraded image. It is then able to measure the bits budget to be spent for leaving the distortion invisible. Some tests on both synthetic and real distorted images validate the theoretical findings. They confirm that the proposed universal threshold is really able to get the lower bound of human perception and then it may be useful for many applications, such as bit allocation in coding procedures and automatic parameters setting in restoration algorithms.

This letter investigates the possibility of removing noise in correspondence to jump discontinuities using the sorted copy of the signal. It will be proved that sorting makes noise predictable so that it can be reproduced and subtracted from the sorted noisy signal. It will be also shown that the proposed method can substitute for the edge preserving term into an anisotropic diffusion scheme, gaining in terms of mean square error, edge preservation and computational effort.

This paper presents a novel approach for the extraction of the transients content of audio signals, usually represented as superposition of stationary, transient, and stochastic components. The proposed model exploits the predictable and peculiar time-scale behavior of transients by modeling them as superposition of suitable wavelet atoms. These latter allow to predict transients information even at scales where the tonal component is dominant. In this way it is possible to avoid, if required, the pre-analysis of the tonal component. Extensive experimental results show that the proposed model achieves good performances with a moderate computational effort and without any user's dependence.

This correspondence presents a novel approach for translational motion estimation based on the phase of the Fourier transform. It exploits the equality between the averaging of a group of successive frames and the convolution of the reference one with an impulse train function. The use of suitable space filling curves allows to reduce the error in motion estimation making the proposed approach robust under noise. Experimental results show that the proposed approach outperforms available techniques in terms of objective (PSNR) and subjective quality with a lower computational effort.

This paper presents an automatic technique to remove semi-transparent blotches (due to moisture) from archived photographs and documents. Blotches are processed in the HSV space. While chroma components are processed using a simple texture synthesis method, the intensity component is split into an over-complete wavelet representation. In the approximation band, the blotch is modelled as an alpha matte which reduces the intensity of the image in a non-uniform yet smooth manner. The alpha matte is estimated using a Bayesian approach and its effect reversed. Wavelet details are left unchanged in the case of perfect semi-transparency or attenuated using visibility laws whenever dirt and dust cause spurious edges. Experimental results achieved on many historical photographs show the effectiveness of the proposed approach.

WISDOW-Comp is a scheme for simultaneous image compression and denoising. Its key issue consists of firstly expanding the input image in a basis of wavelets. The corresponding coefficients can be then seen as waves that travel through both space and scales. Thus the overlapping effects principle can be applied to manage waves interference. Each atom is characterized by the amplitude and the location of its absolute modulus maximum. This powerful representation allows us to sent to the decoder just the low pass component of the image and the significance map of atoms absolute maxima for achieving a denoised image. Experimental results show that the proposed comp-denoiser outperforms the state of the art in terms of both rate and distortion.

This paper presents a generalization of Kokaram's model for scratch lines detection on digital film materials. It is based on the assumption that scratch is not purely additive on a given image but shows also a destroying effect. This result allows us to design a more efficacious scratch detector which performs on a hierarchical representation of a degraded image, i.e., on its cross section local extrema. Thanks to Weber's law, the proposed detector even works well on slight scratches resulting completely automatic, except for the scratch color (black or white). The experimental results show that the proposed detector works better in terms of good detection and false alarms rejection with a lower computing time.