Subtraction Model Research Articles

Dynamic Deep Pixel Distribution Learning for Background Subtraction

Previous approaches to background subtraction usually approximate the distribution of pixels with artificial models. In this paper, we focus on automatically learning the distribution, using a novel background subtraction model named Dynamic Deep Pixel Distribution Learning (D-DPDL). In our D-DPDL model, a distribution descriptor named Random Permutation of Temporal Pixels (RPoTP) is dynamically generated as the input to a convolutional neural network for learning the statistical distribution, and a Bayesian refinement model is tailored to handle the random noise introduced by the random permutation. Because the temporal pixels are randomly permutated to guarantee that only statistical information is retained in RPoTP features, the network is forced to learn the pixel distribution. Moreover, since the noise is random, the Bayesian theorem is naturally selected to propose an empirical model as a compensation based on the similarity between pixels. Evaluations using standard benchmark demonstrates the superiority of the proposed approach compared with the state-of-the-art, including traditional methods as well as deep learning methods.

Conceptual models for integer addition and subtraction

In this article, we report the findings of a study conducted with 6 Grade 8 students in the United States. The students posed stories for open number sentences involving addition and subtraction of integers. We analysed the stories posed by the students to build models that describe the conceptual structures behind these posed stories – the conceptual models for integer addition and subtraction. These four conceptual models for thinking about and using integer addition and subtraction include Bookkeeping, Counterbalance, Relativity, and Translation, and are generated from the students’ posed stories. We also provide profiles of conceptual model use for two of the 6 students that describe how the students posed stories to accommodate conceptual model use, such as posing unconventional or unrealistic stories or changing the structure of the number sentences. The conceptual models and descriptions of how the students used them provide perspective into student thinking about integers and contexts, highlighting the mathematics of the students, and calling for a re-examination of contexts used in school mathematics.

Automatic space object detection on all-sky images from a synoptic survey synthetic telescope array

Detection, classification, and characterization of Space Objects (SOs) are important tasks in many areas of research. Detection of SOs is important in predicting collisions with debris that could become hazards for satellites or space missions in Near Earth orbits. This paper describes a flexible pipeline able to detect sunlit SOs automatically in images acquired using an all-sky camera with a large field of view (FoV). The proposed pipeline includes the following main steps: image distortion correction, filtering for noise reduction, generation of a background model for subtraction, star elimination using a star catalog, local-based contrast enhancement, and, finally, for automatic SO detection, two methodologies were developed to detect line segments. The first one uses a Canny edge detector and a Progressive Probabilistic Hough Transform, and the second is based on the Radon Transform for detecting line segments. The method was applied to a dataset of 22 × 3 images obtained from the Omnidirectional Space Situational Awareness (OmniSSA) Array at the Georgia Institute of Technology in downtown Atlanta. The OmniSSA array has 3 sensors that capture high-resolution images simultaneously (3352 × 2532 pixels) using a wide FoV for each camera. An intensity scaled by noise (ISN) signal was defined and measured to show improvement in SO detection objectively. Fusing images from the three OmniSSA sensors after the background subtraction step improved both the ISN and visualization during the detection stage. Ground-truth data were extracted from a Space-Track catalog and marked by human experts to validate the results of the pipeline, considering information from Astrometry.net. Results showed that almost all the SOs were correctly detected by the pipeline.

Localised quantitative structure–retention relationship modelling for rapid method development in reversed-phase high performance liquid chromatography

Quantitative structure–retention relationships (QSRR) predicting the values of solute “hydrophobicity” coefficient η′ in the approximate hydrophobic subtraction model (HSM) can be used to predict retention times of compounds on numerous reversed-phase (RP) columns, provided that column parameters on the corresponding stationary phases are available. In the present study, we propose a new dual clustering-based localised QSRR approach, combining P-ratio clustering (where P is the octanol–water partition coefficient) with second dominant interaction (SDI)-based clustering, to produce predictive models with an acceptable level of prediction accuracy for in silico column scoping in RP method development. QSRR models for η′ values were derived for 49 compounds out of 63 in a dataset extracted from the literature, where retention data were measured under one isocratic mobile phase condition (i.e., acetonitrile-water, 50:50 [v/v]). These models gave a predictive squared correlation coefficient Qext(F2)2 of 0.83 and a root mean square error of prediction (RMSEP) of 0.14. For the modelling, a genetic algorithm-partial least square regression (GA-PLS) approach was performed using the η′ values and their relevant molecular descriptors. The corresponding retention times were predicted by applying the predicted η′ values of the models and the stationary phase “hydrophobicity” parameter H values for the corresponding columns to the approximate HSM, resulting in excellent accuracy and predictability (Qext(F2)2 of 0.90 and RMSEP of 0.72 min). The established QSRR approach was experimentally verified for six Thermo Scientific columns (Acclaim™ 120 C18, Acclaim Polar Advantage, Acclaim Polar Advantage II, Accucore™ aQ, Accucore Phenyl-X, and Hypersil Gold C18 columns) using two types of datasets. The first dataset consisted of eight model compounds extracted from the original dataset and retention time predictions for those compounds were then evaluated on the above columns. The result showed good agreement between predicted and observed retention times with an acceptable error in retention time predictions (slope of 0.97, Qext(F2)2 of 0.95, a mean absolute error (MAE) of 0.43 min and RMSEP of 0.61 min). The second dataset included eight test compounds not included in the original dataset, which were all classified into the η′ cluster by applying a Tanimoto similarity (TS) threshold of 0.7. Similarly, predicted retention times of the test compounds were compared with their corresponding observed retention times, resulting in acceptable retention time predictions with the slope of 0.99, Qext(F2)2 of 0.93 and RMSEP of 0.52 min. Comparisons of resolution values between columns were utilised to select the most suitable columns for separations of the compounds in the respective test sets. Actual chromatograms obtained on the chosen columns showed the feasibility for effective column scoping without experimentation on numerous RP stationary phases available in the USP website, based on the predicted resolution values.

P: 80 Optimizing the Performance of the Psychometric Hepatic Encephalopathy Score (PHES) for the Diagnosis of Hepatic Encephalopathy

BACKGROUND: PHES is the most established test for the diagnosis of hepatic encephalopathy (HE). The availability of normative data to adjust for potential confounding variable is a necessity. However, there are other variables which may confound the performance of the PHES test and might result in the misclassification of patients, including: (i) the confounding variables used for test correction; (ii) possible interactions between the corrections variables; (ii) the scoring of the line tracing test (LTT) and (iii) the final summation of the component scores. The aim of this study was to produce a series of models to determine the influence of these variables on the performance of the PHES test. METHODS: PHES testing was undertaken in 524 permanent UK residents aged > 18 years, who spoke good English, and were generally healthy. A UK PHES model was constructed as follows: (i) a reciprocal model for the LTT time/error relationship was devised; (ii) univariate/multivariate analyses identified significant correction factors for the individual test scores, without applying interaction terms; and (iii) each corrected test score was transformed to a Z-Score and then summed. The final PHES score was expressed as a Z-Score; values below −2.0 were considered abnormal. The effects of changing the basic PHES model by adding in interaction terms, changing the LTT scoring system and the final summation method were examined by applying the models to PHES data collected in 226 patients with cirrhosis. RESULTS: Seventy-six (33.6%) of the 226 patients had abnormal UK PHES scores. Variation in the scoring of the LTT test (German, Spanish and Italian versions) resulted in a minor reduction in the number of abnormal results (73 [32%]). Add-in and subtraction models for the independent correction factors - age, sex, ethnicity, alcohol consumption, years/place of education -were associated with changes of −11% to +9.8% and +2.4 to −8.8% respectively, in the proportion of abnormal result; ethnicity and age had the greatest effects. Adding interaction terms for the correction factors marginally increased the number of abnormal results to 84 (37.2%). Changing the final score summation from Z-scores to integerized scores significantly increased the number of abnormal tests to 100 (44.2%) P = 0.026. CONCLUSIONS: The major factor affecting the proportion of abnormal PHES results is the method used for final summation of the scores. Integerization of the scores does not allow for the exact application of thresholds and its use should perhaps be reconsidered.

Inside Back cover: Retention prediction using quantitative structure‐retention relationships combined with the hydrophobic subtraction model in reversed‐phase liquid chromatography

Inside Back cover: Retention prediction using quantitative structure‐retention relationships combined with the hydrophobic subtraction model in reversed‐phase liquid chromatography

Colours in your pocket: smartphone-based spectrometers to investigate the quantum world

We designed a low cost spectrometer to be used by groups of students with an inquiry approach. This simple equipment allows to investigate the additive and subtractive models of colour formation, to study the selective absorption of a material and explain it from a microscopic point of view. The home made apparatus was used in several experiments showing the quantum nature of the absorption and production of light with the aim of bridging optics and modern physics.

Automatic foreground detection based on KDE and binary classification

In the recent decades, several methods have been developed to extract moving objects in the presence of dynamic background. However, most of them use a global threshold, and ignore the correlation between neighboring pixels. To address these issues, this paper presents a new approach to generate a probability image based on Kernel Density Estimation (KDE) method, and then apply the Maximum A Posteriori in the Markov Random Field (MAP-MRF) based on probability image, so as to generate an energy function, this function will be minimized by the binary graph cut algorithm to detect the moving pixels instead of applying a thresholding step. The proposed method was tested on various video sequences, and the obtained results showed its effectiveness in presence of a dynamic scene, compared to other background subtraction models.

Adaptive etalon suppression technique for long‐term stability improvement in high index contrast waveguide‐based laser absorption spectrometers

The authors present an adaptive algorithm based on a non-linear regression model for mitigating time-varying etalon drifts in line-scanned optical absorption spectrometers. By dynamically varying the etalon spectral background using physically realistic degrees of freedom, the authors’ dynamic etalon fitting-routine (DEF-R) significantly increases the spectral baseline recalibration interval as compared to conventional fringe subtraction models. They provide an empirical demonstration of the efficacy of DEF-R using an on-chip 10 cm silicon waveguide for near-infrared methane absorption spectroscopy at 6057 cm−1, which suffers significant etalon spectral noise due to reflections and multi-path interference from stochastic line-edge roughness imperfections. They demonstrate the corresponding improvement in both spectral clean-up and long-term stability via Allan-variance analysis. For the sensor presented here, application of DEF-R enables Gaussian-noise limited performance for more than 102 s and provides almost an order-of-magnitude improvement in stability time with respect to conventional baseline subtraction. Although DEF-R is applied here to an on-chip sensor embodiment, they envision their technique to be applicable to any absorption sensor limited by time-varying etalon drifts.

A novel background subtraction algorithm based on parallel vision and Bayesian GANs

To address the challenges of change detection in the wild, we present a novel background subtraction algorithm based on parallel vision and Bayesian generative adversarial networks (GANs). First, we use the median filtering algorithm for background image extraction. Then, we build the background subtraction model by using Bayesian GANs to classify all pixels into foreground and background, and use parallel vision theory to improve the background subtraction results in complex scenes. The proposed algorithm has been evaluated on the well-known, publicly available changedetection.net dataset. Experiment results show that the proposed algorithm results in better performance than many state-of-the-art ones. In addition, our model trained on CDnet dataset can generalize very well to unseen datasets, outperforming multiple state-of-art methods. The major contribution of this work is to apply parallel vision and Bayesian GANs to solve the difficulties in background subtraction, achieving high detection accuracy.

Tailoring the Chemical Composition of LiMPO4 (M = Mg, Co, Ni) Orthophosphates To Design New Inorganic Pigments from Magenta to Yellow Hue.

New inorganic pigments with intense and saturated coloration have been prepared by a solid-state route and exhibit a large color scale from magenta to yellow. Indeed, yellow and magenta are two of the three subtractive model's colors with wide application in printing or displays as e-book readers. To develop yellow and magenta hue, we focused on cobalt- and nickel-based orthophosphates thanks to the chemical stability, low density, low price, and easy preparation of such a pigment class. All of these orthophosphates crystallize with the well-known olivine-type structure (orthorhombic Pnma space group) where transition metals are stabilized in a distorted octahedral site. This paper deals with the optical absorption properties of various orthophosphates, the correlations with structural features, and their colorimetric parameters (in L*a*b* color space). The LiCo1- xMg xPO4 series show near-magenta color with tunable luminosity, while the LiNiPO4 compound exhibits a frank yellow coloration. Co2+ (4T1) and Ni2+ (4A2) chromophore ions occupy a more or less distorted octahedral site, leading to tuning of the intensity of the d-d electronic transitions in the visible and near-IR ranges and providing a subtractive color scale; i.e., a LiCo1- xNi xPO4 solid solution possesses a very rich panel of colors between the two yellow and magenta extremes. It is worth noting that the crystal-field splitting and B Racah parameter have been estimated in a first approximation on the basis of the Tanabe-Sugano diagram and lead to the conclusion of a slightly higher crystal-field splitting of around 0.9 eV for Ni2+ ions and similar β covalent parameters, despite the same crystallographic sites of both of these transition metals.

Retention prediction using quantitative structure-retention relationships combined with the hydrophobic subtraction model in reversed-phase liquid chromatography.

The hydrophobic subtraction model (HSM) combined with quantitative structure-retention relationships (QSRR) methodology was utilized to predict retention times in reversed-phase liquid chromatography (RPLC). A selection of new analytes and new RPLC columns that had never been used in the QSRR modeling process were used to verify the proposed approach. This work is designed to facilitate early prediction of co-elution of analytes in pharmaceutical drug discovery applications where it is advantageous to predict whether impurities might be co-eluted with the active drug component. The QSRR models were constructed through partial least squares regression combined with a genetic algorithm (GA-PLS) which was employed as a feature selection method to choose the most informative molecular descriptors calculated using VolSurf+ software. The analyte hydrophobicity coefficient of the HSM was predicted for subsequent calculation of retention. Clustering approaches based on the local compound type and the local second dominant interaction were investigated to select the most appropriate training set of analytes from a larger database. Predicted retention times of five new compounds on five new RPLC C18 columns were compared with their measured retention times with percentage root-mean-square errors of 15.4 and 24.7 for the local compound type and local second dominant interaction clustering methods, respectively.

A pH-stable, crosslinked stationary phase based on the thiol-yne reaction

Stationary phases that can withstand extremes of pH and temperature are needed to allow a single column to accommodate a wider set of solutes and separation criteria. We used a simple multi-step process using the thiol-yne reaction following the modification of the silica surface with a thiol-containing silane. The monomers 1,4-diethynylbenzene (DEB) and 1,6-hexanedithiol were used to create a crosslinked thiol-yne (CTY) stationary phase along the surface of the thiol functionalized silica. In the Tanaka test characterization, the CTY phase showed a low phase ratio, methylene selectivity typical of a reversed phase, and extremely high shape selectivity compared to commercial reversed phases. The hydrophobic subtraction model characterization showed a high positive steric resistance, a low hydrogen bond acidity, and a high cation-exchange capacity compared to most reversed phases. At pH 0.5 with an 85% aqueous mobile phase the phase showed no significant change over 114 h. With a 50% aqueous mobile phase the phase took four more days than a sterically protected C18 phase for the k’ to decline 25%. At pH 12.6, 50% aqueous mobile phase, a sterically protected C18 phase showed a 20% decrease in k’ and more than a 60% decrease in theoretical plates per meter in three hours. The CTY phase actually showed modest increases in k’ and theoretical plates per meter after three hours.

A Coarse-to-Fine Model for Rail Surface Defect Detection

Computer vision systems have attracted much attention in recent years for use in detecting surface defects on rails; however, accurate and efficient recognition of possible defects remains challenging due to the variations shown by defects and also noise. This paper proposes a coarse-to-fine model (CTFM) to identify defects at different scales. The model works on three scales from coarse to fine: subimage level, region level, and pixel level. At the subimage level, the background subtraction model exploits row consistency in the longitudinal direction, and strongly filters the defect-free range, leaving roughly identified subimages within which defects may exist. At the next level, the region extraction model, inspired by visual saliency models, locates definite defect regions using phase-only Fourier transforms. At the finest level, the pixel subtraction model uses pixel consistency to refine the shape of each defect. The proposed method is evaluated using Type-I and Type-II rail surface defect detection data sets and an actual rail line. The experimental results show that CTFM outperforms state-of-the-art methods according to both the pixel-level index and the defect-level index.

Extended Codebook with Multispectral Sequences for Background Subtraction.

The Codebook model is one of the popular real-time models for background subtraction. In this paper, we first extend it from traditional Red-Green-Blue (RGB) color model to multispectral sequences. A self-adaptive mechanism is then designed based on the statistical information extracted from the data themselves, with which the performance has been improved, in addition to saving time and effort to search for the appropriate parameters. Furthermore, the Spectral Information Divergence is introduced to evaluate the spectral distance between the current and reference vectors, together with the Brightness and Spectral Distortion. Experiments on five multispectral sequences with different challenges have shown that the multispectral self-adaptive Codebook model is more capable of detecting moving objects than the corresponding RGB sequences. The proposed research framework opens a door for future works for applying multispectral sequences in moving object detection.

Adaptive Cluster based Model for Fast Video Background Subtraction

Background subtraction (BGS) is one of the impor�tant steps in many automatic video analysis applications. Several researchers have attempted to address the challenges due to illumination variation, shadow, camouflage, dynamic changes in the background and bootstrapping requirement. In this paper, a method to perform BGS using dynamic clustering is proposed. A background model is generated using the K0 -means algorithm. The normalized γ corrected distance values and an automatic threshold value is used to perform the background subtraction. The background models are updated online to handle slow illu�mination changes. The experiment was conducted on CDNet2014 dataset. The experimental results show that the proposed method is fast and performs well for baseline, camera-jitter and dynamic background categories of video.

Background subtraction based on deep convolutional neural networks features

Background modeling and subtraction, the task to detect moving objects in a scene, is a fundamental and critical step for many high level computer vision tasks. However, background subtraction modeling is still an open and challenge problem particularly in practical scenarios with drastic illumination changes and dynamic backgrounds. In this paper, we propose a novel foreground detection method based on CNNs(Convolutional Neural Networks) to deal with challenges confronted with background subtraction. Firstly, given a cleaned background image without moving objects, constructing adjustable neighborhood of each pixel in the background image to form windows; CNN features are extracted with a pre-trained CNN model for each window to form a features based background model. Secondly, for the current frame of a video scene, extracting features with the same operation as the background model. Euclidean distance is adopted to build distance map for current frame and background image with CNN features. Thirdly, the distance map is fed into graph cut algorithm to obtain foreground mask. In order to deal with background changes, the background model is updated with a certain rate. Experimental results verify that the proposed approach is effective to detect foreground objects from complex background environments, and outperforms some state-of-the-art methods.

Column selection for comprehensive two-dimensional liquid chromatography using the hydrophobic subtraction model

Two-dimensional (2D) liquid chromatography (2DLC) methods have grown in popularity due to their enhanced peak capacity that allows for resolving complex samples. Given the large number of commercially available column types, one of the major challenges in implementing 2DLC methods is the selection of suitable column pairs. Column selection is typically informed by chemical intuition with subsequent experimental optimization. In this work a computational screening method for 2DLC is proposed whereby virtual 2D chromatograms are calculated utilizing the Snyder–Dolan hydrophobic subtraction model (HSM) for reversed-phase column selectivity. Towards this end, 319 225 column pairs resulting from the combination of 565 columns and 100 sets of 1000 diverse analytes are examined. Compared to other screening approaches, the present method is highly predictive for column pairs that are able to resolve the largest number of analytes. This approach shows a strong sensitivity to the choice of the second dimension column (having a shorter operating time) and a preference for those with embedded polar moieties, whereas a relatively weak preference for C18 and phenyl columns is found for the first dimension.

Σpider diagram: A universal and versatile approach for system comparison and classification. Part 2: Stationary phase properties

Many different sorts of bonded phase chemistries may be used in high-performance liquid chromatography (HPLC) in the reversed-phase mode: C8, C18 (type A or B), polar-embedded C18, phenyl, pentafluorophenyl, or cyanopropyl. To assess their retention and selectivity properties, chromatographic tests exist. The data obtained from these tests may be presented in three ways. First, simple classification diagrams may be plotted, when only two or three parameters are studied. Secondly, chemometric treatments such as principal component analyses (PCA) or hierarchical cluster analyses (HCA) may be computed, when at least 4 parameters are studied. These are sometimes uneasy to interpret. Thirdly, the “distance” between one column and a reference column may be estimated, through calculated ranking (F or CDF) or selectivity (s) factors.In this paper, another treatment type is applied to the data of Euerby (Tanaka test) and Snyder (Hydrophobic Subtraction Model), each of these tests having 6 parameters. This treatment produces a visual classification, called spider diagram. In the first part of this series, this type of classification was applied to the classification of solvents. A logical and easily comprehensible classification is obtained for the varied types of bonded phases, with a clear location, which can be related to the chromatographic properties. The comparison of these diagrams shows that the classification based on Snyder’s hydrophobic subtraction model discriminates the stationary phases more effectively than the one based on the Tanaka test.Finally, on the basis of the parameter relevance and in order to favour comparison between these two tests and a third one called the carotenoid test, simplified classification maps are proposed. For Tanaka test, the selected parameters are the pentylbenzene retention factor (hydrophobicity), the benzylamine/phenol separation factor at pH 7.6 (polar surface activity) and the triphenylene/ortho-terphenyl (shape selectivity) separation factor. For Snyder test, the parameters selected are the ethylbenzene retention factor (hydrophobicity) the C term at pH 7.0 (polar surface activity) and the S* term (steric electivity). The location of some stationary phases onto the maps and their rankings are compared and shown to agree well between the three tests.

Side Channel Steganalysis: When Behavior is Considered in Steganographer Detection

This paper intents to solve the challenging problem of steganographer detection in the real world from a new perspective: side channel attack. We propose utilizing the behavior of actors in the social network to identify the steganographer. While there are many behavior information may expose the steganographer, we just consider the correlation between images sequence as an example in this paper. Base on the assumption that the steganographer choosing images for communication randomly, we design the feature of subtractive images adjacent model (SIAM) to represent the correlations between images sequence of each actor. And then a binary classifier is used to identify the steganographer. To simulate the real world, the images in the experiment are all crawled from twitter. The experimental result shows good performance of our method.