An overview of median and stack filtering

Within the last two decades a small group of researchers has built a useful, nontrivial theory of nonlinear signal processing around the median-related filters known as rank-order filters, order-statistic filters, weighted median filters, and stack filters. This required significant effort to overcome the bias, both in education and research, toward linear theory, which has been dominant since the days of Fourier, Laplace, and Convolute.We trace the development of this theory of nonlinear filtering from its beginnings in the study of noise-removal properties and structural behavior of the median filter to the recently developed theory of optimal stack filtering.The theory of stack filtering provides a point of view which unifies many different filter classes, including morphological filters, so it is discussed in detail. Of particular importance is the way this theory has brought together, in a single analytical framework, both the estimation-based and the structural-based approaches to the design of these filters.Some recent applications of median and stack filters are provided to demonstrate the effectiveness of this approach to nonlinear filtering. They include: the design of an optimal stack filter for image restoration; the use of vector median filters to attenuate impulsive noise in color images and to eliminate cross luminance and cross color in TV images; and the use of median-based filters for image sequence coding, reconstruction, and scan rate conversion in normal TV and HDTV systems.     

An effective CU size decision method for quality scalability in SHVC

The Scalable extension of the High Efficiency Video Coding (known as SHVC) combines the high compression efficiency with the possibility of encoding different resolutions of the same encoded video in a single bitstream. However, this is accompanied with a high computational complexity. In this paper, we propose an effective coding unit (CU) size decision method by restricting the CU depth range to reduce the encoding time for quality scalability in SHVC. Since the optimal depth level in the enhancement layer (EL) is highly correlated to that in the base layer (BL), we can determine the CU depth range in the EL according to the depth of the co-located CU in the BL. Based on the high correlation between the current CU and its spatio-temporal neighboring CUs, the proposed method skips some specific depth levels which are rarely used in the previous frame and neighboring CUs to further reduce the computational complexity. Experimental results demonstrate that the proposed method can efficiently reduce computational complexity while maintaining similar rate distortion (RD) performance as the original SHVC encoder.     

Screening and optimization of the factors of a detergent admixture preparation

The present research studies the optimization of the preparation of a hydrotrope (mixture of sodium toluenesulfonate and sodium xylenesulfonates) for liquid detergents using experimental designs. The preparation is carried out by selective sulfonation of toluene and xylene present in the BTX fraction of a natural gas. The optimization of five responses has been carried out in two steps. First, we carry out an asymmetrical screening design 2¹3⁶//18 to select from seven factors, those factors which have significant influence on the amount of residual sodium sulfate in the product and on the total conversion of toluene and xylenes. Secondly, we carry out a Box–Behnken design to optimize three retained factors (temperature, molar ratio of acid to toluene plus xylenes, and the amount of desulfation agent), using response surface methodology. In addition, and in order to reach near-optimal sulfonation conditions, we use the desirability functions to optimize the five responses simultaneously. This allows us to determine for the whole process, sulfonation conditions leading to a high yield of toluene and xylene conversion and acceptable amounts of residual sodium sulfates and sodium benzenesulfonate.     

Forecasting Financial Time Series Using Robust Deep Adaptive Input Normalization

Deep Learning provided powerful tools for forecasting financial time series data. However, despite the success of these approaches on many challenging financial forecasting tasks, it is not always straightforward to employ DL-based approaches for highly volatile and non-stationary time financial series. To this end, in this paper, an adaptive input normalization layer that can learn to identify the distribution from which the input data were generated and then apply the most appropriate normalization scheme is proposed. This allows for promptly adapting the input to the subsequent DL model, which can be especially important, given recent findings that hint at the existence of critical learning periods in neural networks. Furthermore, the proposed method operates on a sliding window over the time series allowing for overcoming non-stationary issues that often arise. It is worth noting that the main difference with existing approaches is that the proposed method does not just learn to perform static normalization, e.g., using a fixed set of parameters, but instead it adaptively calculates the most appropriate normalization parameters, significantly improving the robustness of the proposed approach when distribution shifts occur. The effectiveness of the proposed formulation is verified using extensive experiments on three challenging financial time-series datasets.

     

Parallel watershed transformation algorithms for image segmentation

The watershed transformation is a mid-level operation used in morphological image segmentation. Techniques applied on large images, which must often complete fast, are usually computationally expensive and complex entailing efficient parallel algorithms. Two distributed approaches of the watershed transformation are introduced in this paper. The algorithms survey in a Single Program Multiple Data (SPMD) model both local and global connectivity properties of the morphological gradient of a gray-scale image to label connected components. The sequentiality of the serial algorithm is broken in the parallel versions by exploiting the ordering relation between two neighboring pixels successively incorporated in the same region. Thus, a path is traced, for every unlabeled pixel, down to its region of inclusion (whose label is then propagated backwards); in the second algorithm, regions grow independently around their seeds. In both cases only pixels which satisfy the ordering relation are incorporated in any region. This way, not only different regions are explored in a parallel fashion, but also different parts of the same region, when the latter extends to neighboring subdomains, are treated likewise. Running time and relative speedup evaluated on a Cray T3D parallel computer are used to appreciate the performance of both algorithms.     

Pepsinogen and pepsin from the stomach of smooth hound (Mustelus mustelus): Purification,characterization and amino acid terminal sequences

Pepsinogen from the stomach of smooth hound (Mustelus mustelus) was purified to homogeneity by 20–70% ammonium sulphate precipitation, Sephadex G-100 gel filtration and DEAE-cellulose anion exchange chromatography with a 9.4-fold increase in specific activity and 38.36% recovery. Upon activation at pH 2.0, M. mustelus pepsinogen was converted to active form in one-step pathway. Molecular weights of the purified pepsinogen and the active pepsin were estimated to be 40,000 and 35,000 Da using SDS-PAGE and gel filtration, respectively. The optimum pH and temperature for the pepsin activity were pH 2.0 and 40 °C, respectively, using haemoglobin as a substrate. Activity was completely inhibited by Pepstatin A but not by phenylmethylsulphonyl fluoride, a serine-protease inhibitor and ethylenediaminetetraacetic acid, a metalloenzyme inhibitor. The N-terminal amino acid sequences of the first 15 amino acids of the activation segment of the pepsinogen and the first 20 amino acids of the active pepsin were LLRVPLRKGKSTLDV and ATEPLSNYLDSSYFGDISIG, respectively. M. mustelus pepsinogen, which showed high homology to rat C pepsinogen, had Thr-Leu-Asp sequence at amino acid positions 12–14 not found in all pepsinogen sequences. A remarkable substitution was found in the activation segment of M. mustelus pepsinogen: the Arg-13 conserved in all gastric proteinases, whose sequences are known, is replaced by Leu-13.     

Growth and characterization of Cr2N/CrN multilayer coatings

A series of monolithic and multilayer coatings of chromium nitride with various compositions and architectures were deposited at low temperatures (<200°C) on silicon substrates using ion-assisted reactive magnetron sputtering. All coatings had a total thickness in the 1.5±0.3 μm range. The multilayer coatings were designed such that their period and CrN fraction varied in the range 30–150 nm and 0.50–0.93, respectively. Real-time in situ ellipsometry was used to monitor and control the deposition process. The deposited coatings were characterized post-deposition using X-ray diffraction (XRD), Rutherford backscattering (RBS), X-ray photoelectron spectroscopy (XPS), and spectroscopic ellipsometry (SE). The primary chromium nitride phases (Cr₂N and CrN) in the films were identified using XRD. The chemical composition of selected samples was determined from RBS and XPS measurements. The phase composition of the deposited layers was deduced from the analysis of the SE data. The mechanical properties of the coatings were evaluated using a nanoindenter. The measured hardness values were in excess of 20 GPa. The results of the different characterization and testing techniques were correlated and follow-up work on this project suggested.     

The coupled theory of micropolar thermoelastic diffusion

The general equations of motion and constitutive equations are derived for a general homogeneous anisotropic medium with a microstructure, taking into account the effects of heat and diffusion. We establish a reciprocal relation, which involves two thermoelastic diffusion processes at different instants. We show that this relation can be used to obtain reciprocity, uniqueness and continuous dependence theorems. The reciprocity theorem avoids both the use of the Laplace transform and the incorporation of initial conditions into the equations of motion. The uniqueness theorem is derived without the positive definiteness assumption on the elastic, conductivity and diffusion tensors. We prove also that the reciprocal relation leads to a continuous dependence theorem studied on external body loads. Finally we prove the existence of a generalized solution by means of the semigroup of linear operators theory.     

Phototransformation of the insecticide fipronil: identification of novel photoproducts and evidence for an alternative pathway of photodegradation

Fipronil is a recently discovered insecticide of the phenylpyrazole series. It has a highly selective biochemical mode of action, which has led to its use in a large number of important agronomical, household, and veterinary applications. Previous studies have shown that, during exposure to light, fipronil is converted into a desulfurated derivative (desulfinyl-fipronil), which has slightly reduced insecticidal activity. In this study, the photodegradation of fipronil was studied in solution at low light intensities (sunlight or UV lamp). In addition to desulfinyl-fipronil, a large number of minor photoproducts were observed, including diversely substituted phenylpyrazole derivatives and aniline derivatives that had lost the pyrazole ring. Desulfinylfipronil itself was shown to be relatively stable under both UV light and sunlight, with only limited changes occurring in the substitution of the aromatic ring. Since this compound accumulated to levels corresponding to only 30-55% of the amount of fipronil degraded, it was concluded that one or more alternative pathways of photodegradation must be operating. On the basis of the structurally identified photoproducts, it is proposed that fipronil photodegradation occurs via at least two distinct pathways, one of which involves desulfuration at the 4-position of the pyrazole ring giving the desulfinyl derivative and the other of which involves a different modification of the 4-substituent, leading to cleavage of the pyrazole ring and the formation of aniline derivatives. The latter compounds do not accumulate to high levels and may, therefore, be degraded further. The ecological significance of these results is discussed, particularly with regard to the insecticidal activity of the photoproducts.     

Exploratory investigation on mechanical anchors for connecting SMA bars to steel or FRP bars

Using superelastic shape memory alloys (SMAs) as reinforcing bars in concrete structures proved to have a great potential in seismic areas because of its recentering capability. However, using them in an entire structure is generally not economically feasible due to their high cost. Therefore, it is more practical to limit their use to the plastic hinge zones, while regular steel can be used in the other regions of the structure. Connections between SMA and steel are critical, and need to be strong enough to transfer the full force from SMA bars to steel bars. Various mechanical couplers are available in the market to splice bars in reinforced concrete (RC) structures, each of which has several advantages and disadvantages. The efficiency of these couplers for connecting steel bars is tested and reported in this paper. Since these couplers are intended for connecting steel bars only, another experimental investigation has been performed to determine the suitability of these couplers for connecting SMA with steel bars. Commercially available screw-lock couplers are found to be unsuitable for connecting SMA to steel bars. An existing coupler has been modified for SMA–steel splicing to allow SMA bars to achieve their full superelastic strain. Additional tests have also been performed for connecting FRP bars to SMA bars. A new generation mechanical-adhesive type coupler has been developed for splicing FRP to SMA bars.