Background Subtraction Research Articles

3D PET/CT Tumor Co-Segmentation Based on Background Subtraction Hybrid Active Contour Model

Accurate tumor segmentation in medical images plays an important role in clinical diagnosis and disease analysis. However, medical images usually have great complexity, such as low contrast of computed tomography (CT) or low spatial resolution of positron emission tomography (PET). In the actual radiotherapy plan, multimodal imaging technology, such as PET/CT, is often used. PET images provide basic metabolic information and CT images provide anatomical details. In this paper, we propose a 3D PET/CT tumor co-segmentation framework based on active contour model. First, a new edge stop function (ESF) based on PET image and CT image is defined, which combines the grayscale standard deviation information of the image and is more effective for blurry medical image edges. Second, we propose a background subtraction model to solve the problem of uneven grayscale level in medical images. Apart from that, the calculation format adopts the level set algorithm based on the additive operator splitting (AOS) format. The solution is unconditionally stable and eliminates the dependence on time step size. Experimental results on a dataset of 50 pairs of PET/CT images of non-small cell lung cancer patients show that the proposed method has a good performance for tumor segmentation.

Mother of dragons

Context. Core accretion models of massive star formation require the existence of massive, starless cores within molecular clouds. Yet, only a small number of candidates for such truly massive, monolithic cores are currently known. Aims. Here we analyse a massive core in the well-studied infrared-dark cloud (IRDC) called the ‘dragon cloud’ (also known as G028.37+00.07 or ‘Cloud C’). This core (C2c1) sits at the end of a chain of a roughly equally spaced actively star-forming cores near the center of the IRDC. Methods. We present new high-angular-resolution 1 mm ALMA dust continuum and molecular line observations of the massive core. Results. The high-angular-resolution observations show that this region fragments into two cores, C2c1a and C2c1b, which retain significant background-subtracted masses of 23 M⊙ and 2 M⊙ (31 M⊙ and 6 M⊙ without background subtraction), respectively. The cores do not appear to fragment further on the scales of our highest-angular-resolution images (0.2″, 0.005 pc ~ 1000 AU). We find that these cores are very dense (nH2 > 106 cm−3) and have only trans-sonic non-thermal motions (ℳs ~ 1). Together the mass, density, and internal motions imply a virial parameter of <1, which suggests the cores are gravitationally unstable, unless supported by strong magnetic fields with strengths of ~1–10 mG. From CO line observations, we find that there is tentative evidence for a weak molecular outflow towards the lower-mass core, and yet the more massive core remains devoid of any star formation indicators. Conclusions. We present evidence for the existence of a massive, pre-stellar core, which has implications for theories of massive star formation. This source warrants follow-up higher-angular-resolution observations to further assess its monolithic and pre-stellar nature.

Traffic Rules Violation Detection System

Abstract: The paper presents a system that uses surveillance videos to automatically detect bike riders without helmets by us-ing object segmentation and background subtraction techniques. A consolidation approach is introduced to improve accuracy, and three feature representations are compared. The system achieves a detection accuracy of 93.80 percentage and has an average processing time of 11.58 milliseconds, making it a cost-effective and real-time solution for managing traffic violations related to helmet usage

Determining the oxygen detection limit with magnetic sector dynamic secondary ion mass spectrometry (SIMS)

Dynamic SIMS is often used in evaluating the concentration of impurities in solids because of its high sensitivity and depth profiling capabilities with good depth resolution and high throughput. Continuous ion beam sputtering with a high-density primary beam provides high sensitivity and reduced background contribution from residual gases within the analytical chamber. Numerical simulations along with the data acquired in the real depth profiling experiment show a good agreement of the simple model, assuming that the surface uptake of the oxygen under Cs bombardment depends on the Cs surface retention concentration. In cases where high depth resolution is desired, requiring a low primary ion beam energy, a background subtraction option will help provide better detection limits. It is proposed to perform the background subtraction based on the difference between the continuous and interleaving sputtering modes of dynamic SIMS. The experiments performed on different instruments using different sputtering rate conditions (and different impact energies) show the power function dependence of oxygen detection limit on the reversed sputter rate variation or sputtering beam density. The proposed detection limit dependence curve can be used for analytical specifications benchmarking in most magnetic sector SIMS instruments.

Traffic Accident Detection Using Background Subtraction and CNN Encoder–Transformer Decoder in Video Frames

Artificial intelligence plays a significant role in traffic-accident detection. Traffic accidents involve a cascade of inadvertent events, making traditional detection approaches challenging. For instance, Convolutional Neural Network (CNN)-based approaches cannot analyze temporal relationships among objects, and Recurrent Neural Network (RNN)-based approaches suffer from low processing speeds and cannot detect traffic accidents simultaneously across multiple frames. Furthermore, these networks dismiss background interference in input video frames. This paper proposes a framework that begins by subtracting the background based on You Only Look Once (YOLOv5), which adaptively reduces background interference when detecting objects. Subsequently, the CNN encoder and Transformer decoder are combined into an end-to-end model to extract the spatial and temporal features between different time points, allowing for a parallel analysis between input video frames. The proposed framework was evaluated on the Car Crash Dataset through a series of comparison and ablation experiments. Our framework was benchmarked against three accident-detection models to evaluate its effectiveness, and the proposed framework demonstrated a superior accuracy of approximately 96%. The results of the ablation experiments indicate that when background subtraction was not incorporated into the proposed framework, the values of all evaluation indicators decreased by approximately 3%.

Real-World Deep Local Motion Deblurring

Most existing deblurring methods focus on removing global blur caused by camera shake, while they cannot well handle local blur caused by object movements. To fill the vacancy of local deblurring in real scenes, we establish the first real local motion blur dataset (ReLoBlur), which is captured by a synchronized beam-splitting photographing system and corrected by a post-progressing pipeline. Based on ReLoBlur, we propose a Local Blur-Aware Gated network (LBAG) and several local blur-aware techniques to bridge the gap between global and local deblurring: 1) a blur detection approach based on background subtraction to localize blurred regions; 2) a gate mechanism to guide our network to focus on blurred regions; and 3) a blur-aware patch cropping strategy to address data imbalance problem. Extensive experiments prove the reliability of ReLoBlur dataset, and demonstrate that LBAG achieves better performance than state-of-the-art global deblurring methods and our proposed local blur-aware techniques are effective.

AMEBaS: Automatic Midline Extraction and Background Subtraction of Ratiometric Fluorescence Time-Lapses of Polarized Single Cells

Cell polarity is a macroscopic phenomenon established by a collection of spatially concentrated molecules and structures that culminate in the emergence of specialized domains at the subcellular level. It is associated with developing asymmetric morphological structures that underlie key biological functions such as cell division, growth, and migration. In addition, the disruption of cell polarity has been linked to tissue-related disorders such as cancer and gastric dysplasia. Current methods to evaluate the spatiotemporal dynamics of fluorescent reporters in individual polarized cells often involve manual steps to trace a midline along the cells' major axis, which is time consuming and prone to strong biases. Furthermore, although ratiometric analysis can correct the uneven distribution of reporter molecules using two fluorescence channels, background subtraction techniques are frequently arbitrary and lack statistical support. This manuscript introduces a novel computational pipeline to automate and quantify the spatiotemporal behavior of single cells using a model of cell polarity: pollen tube/root hair growth and cytosolic ion dynamics. A three-step algorithm was developed to process ratiometric images and extract a quantitative representation of intracellular dynamics and growth. The first step segments the cell from the background, producing a binary mask through a thresholding technique in the pixel intensity space. The second step traces a path through the midline of the cell through a skeletonization operation. Finally, the third step provides the processed data as a ratiometric timelapse and yields a ratiometric kymograph (i.e., a 1D spatial profile through time). Data from ratiometric images acquired with genetically encoded fluorescent reporters from growing pollen tubes were used to benchmark the method. This pipeline allows for faster, less biased, and more accurate representation of the spatiotemporal dynamics along the midline of polarized cells, thus advancing the quantitative toolkit available to investigate cell polarity. The AMEBaS Python source code is available at: https://github.com/badain/amebas.git

Efficient parallel implementation of Gaussian Mixture Model background subtraction algorithm on an embedded multi-core Digital Signal Processor

Scene segmentation is crucial in video processing applications, such as vehicle detection. The Gaussian Mixture Model (GMM) based background subtraction (BS) is a widely employed technique for scene segmentation. However, due to its high computational demands, an efficient parallel implementation is necessary. This paper proposes a parallelized implementation of the GMM algorithm on the C6678 digital signal processor (DSP), utilizing its eight cores. Three different parallelization approaches are presented: Open Multiprocessing (OpenMP), manually optimized method, and manually optimized method involving Direct Memory Access (DMA) engine. To evaluate their effectiveness, a metrics comparison is conducted using background truth images provided by the literature dataset references. The experimental results demonstrate that the proposed manually optimized method surpasses the current state-of-the-art. It achieves a remarkable parallel efficiency of approximately 99% for low and medium frame sizes and 83% and 79% for high-definition (HD) and full-HD resolution frames, respectively, when all eight cores are enabled. In contrast, the parallel efficiency obtained using OpenMP is only 45%.

Dance Video Motion Recognition Based on Computer Vision and Image Processing

ABSTRACT In recent years, motion recognition has become a hot research topic in the field of computer vision and has great research worth. Scholars and research institutions at home and abroad have made a lot of research and achieved good results. However, the research results of applying motion recognition technology to dance video are relatively few, mainly because the complexity of human motion in dance video is relatively high. In article, an optimized design scheme for motion recognition research of dance video based on computer vision and image processing is proposed. The collected dance video is preprocessed, including grayscale, background subtraction and filter denoising, so as to extract the character features in the video image. Then, the method of self-organizing mapping neural network (SOM) in computer vision is used to realize the recognition of dance movements. Finally, the simulation test and analysis are carried out. The simulation results show that the proposed arithmetic has a certain accuracy, which is 9.34% higher than the traditional arithmetic. The research of motion recognition method based on dance video will also play a certain reference role for the research of human motion recognition in a large number of real and complex environments, and enrich the application field of motion recognition technology.

Quantifying high-dimensional spatial entanglement with a single-photon-sensitive time-stamping camera.

High-dimensional entanglement is a promising resource for quantum technologies. Being able to certify it for any quantum state is essential. However, to date, experimental entanglement certification methods are imperfect and leave some loopholes open. Using a single-photon-sensitive time-stamping camera, we quantify high-dimensional spatial entanglement by collecting all output modes and without background subtraction, two critical steps on the route toward assumptions-free entanglement certification. We show position-momentum Einstein-Podolsky-Rosen (EPR) correlations and quantify the entanglement of formation of our source to be larger than 2.8 along both transverse spatial axes, indicating a dimension higher than 14. Our work overcomes important challenges in photonic entanglement quantification and paves the way toward the development of practical quantum information processing protocols based on high-dimensional entanglement.

NeuroHSMD: Neuromorphic Hybrid Spiking Motion Detector

Vertebrate retinas are highly-efficient in processing trivial visual tasks such as detecting moving objects, which still represent complex challenges for modern computers. In vertebrates, the detection of object motion is performed by specialised retinal cells named Object Motion Sensitive Ganglion Cells (OMS-GC). OMS-GC process continuous visual signals and generate spike patterns that are post-processed by the Visual Cortex. Our previous Hybrid Sensitive Motion Detector (HSMD) algorithm was the first hybrid algorithm to enhance Background subtraction (BS) algorithms with a customised 3-layer Spiking Neural Network (SNN) that generates OMS-GC spiking-like responses. In this work, we present a Neuromorphic Hybrid Sensitive Motion Detector (NeuroHSMD) algorithm that accelerates our HSMD algorithm using Field-Programmable Gate Arrays (FPGAs). The NeuroHSMD was compared against the HSMD algorithm, using the same 2012 Change Detection (CDnet2012) and 2014 Change Detection (CDnet2014) benchmark datasets. When tested against the CDnet2012 and CDnet2014 datasets, NeuroHSMD performs object motion detection at 720 × 480 at 28.06 Frames Per Second (fps) and 720 × 480 at 28.71 fps, respectively, with no degradation of quality. Moreover, the NeuroHSMD proposed in this article was completely implemented in Open Computer Language (OpenCL) and therefore is easily replicated in other devices such as Graphical Processing Units (GPUs) and clusters of Central Processing Units (CPUs).

Static High Target-Induced False Alarm Suppression in Circular Synthetic Aperture Radar Moving Target Detection Based on Trajectory Features

The new mode of Circular Synthetic Aperture Radar (CSAR) has several advantages including multi-aspect and long-time observation, which can generate high-frame-rate image sequences to detect moving targets with a single-channel system. Nonetheless, due to CSAR being sensitive to 3D structures, static high targets are observed in scene display rotational motion within CSAR subaperture image sequences. Such motion can cause false alarms rising when utilizing image sequence-based moving target detection methods like logarithm background subtraction (LBS). To address this issue, this paper first thoroughly analyzes the moving target and static high target’s difference for the trajectory in an image sequence. Two new trajectory features of the rotation angle and moving distance are proposed to differentiate them. Based on the features, a new false alarm suppression method is proposed. The method first utilizes LBS to obtain coarse binary detection results comprising both moving and static high targets, then employs morphological filtering to eliminate noise. Next, DBSCAN and target tracking steps are employed to extract the trajectory features of the target and false alarm. Finally, false alarms are suppressed with trajectory-based feature discriminators to output detection results. The W-band CSAR open dataset is used to validate the proposed method’s effectiveness.

Near-Infrared Blood Vessel Image Segmentation Using Background Subtraction and Improved Mathematical Morphology.

The precise display of blood vessel information for doctors is crucial. This is not only true for facilitating intravenous injections, but also for the diagnosis and analysis of diseases. Currently, infrared cameras can be used to capture images of superficial blood vessels. However, their imaging quality always has the problems of noises, breaks, and uneven vascular information. In order to overcome these problems, this paper proposes an image segmentation algorithm based on the background subtraction and improved mathematical morphology. The algorithm regards the image as a superposition of blood vessels into the background, removes the noise by calculating the size of connected domains, achieves uniform blood vessel width, and smooths edges that reflect the actual blood vessel state. The algorithm is evaluated subjectively and objectively in this paper to provide a basis for vascular image quality assessment. Extensive experimental results demonstrate that the proposed method can effectively extract accurate and clear vascular information.

Awesome SOSS: transmission spectroscopy of WASP-96b with NIRISS/SOSS

ABSTRACT The future is now – after its long-awaited launch in 2021 December, JWST began science operations in 2022 July and is already revolutionizing exoplanet astronomy. The Early Release Observations (ERO) programme was designed to provide the first images and spectra from JWST, covering a multitude of science cases and using multiple modes of each on-board instrument. Here, we present transmission spectroscopy observations of the hot-Saturn WASP-96 b with the Single Object Slitless Spectroscopy (SOSS) mode of the Near Infrared Imager and Slitless Spectrograph, observed as part of the ERO programme. As the SOSS mode presents some unique data reduction challenges, we provide an in-depth walk-through of the major steps necessary for the reduction of SOSS data: including background subtraction, correction of 1/f noise, and treatment of the trace order overlap. We furthermore offer potential routes to correct for field star contamination, which can occur due to the SOSS mode’s slitless nature. By comparing our extracted transmission spectrum with grids of atmosphere models, we find an atmosphere metallicity between 1× and 5× solar, and a solar carbon-to-oxygen ratio. Moreover, our models indicate that no grey cloud deck is required to fit WASP-96 b’s transmission spectrum, but find evidence for a slope shortward of 0.9 μm, which could either be caused by enhanced Rayleigh scattering or the red wing of a pressure-broadened Na feature. Our work demonstrates the unique capabilities of the SOSS mode for exoplanet transmission spectroscopy and presents a step-by-step reduction guide for this new and exciting instrument.

Twin background subtraction technique: A novel infrared technique for simultaneous thickness and temperature measurements of micro films

The study proposes a novel non-intrusive infrared technique to measure simultaneously the thickness and temperature of a thin layer that is semi-transparent to infrared radiation. The measurement method has been called Twin Background Subtraction technique. The technique simultaneously exploits the mid-wave IR emission from the film to be measured and the radiation attenuation by the same film of the emission coming from two backgrounds at different constant temperatures. This study discusses the method theoretically and experimentally, focusing on the uncertainties and the dependence on different parameters of the set-up and fluid physical properties.The fluid adopted for the film is water, and various tests have been carried out by changing its temperature between 15 and 50 °C and its thickness from 0 to 500 µm. The TBS technique was proven to be of interest for studying micro to mini films both from a qualitative and quantitative point of view for water films ranging between 10 µm and 350 µm.

Kernel-Induced Possibilistic Fuzzy Associate Background Subtraction for Video Scene

The background subtraction (BGS) technique is popularly used for many surveillance systems, segmenting the foreground by subtracting the modeled background from the image sequences. The effectiveness of any BGS technique depends on the robustness of the constructed background model. It is to be noted that many BGS schemes are affected by the inclusion of either noisy pixels in background construction or parameters of generative models. In this regard, we propound an idea of a kernel-induced possibilistic fuzzy associated BGS scheme for local change detection from a fixed camera captured sequence. The proposed scheme follows two stages: background training and foreground segmentation. In the background construction stage, each pixel is modeled using a possibilistic fuzzy cost function in kernel-induced space. The use of the induced kernel function will project the low-dimensional data into a higher dimensional space and the use of the possibilistic function will construct a robust background model based on the density of the data in the temporal domain avoiding the noisy and outlier points. The performance of the proposed scheme is tested on three benchmark databases. The effectiveness of the proposed scheme is evaluated on different performance evaluation measures: precision, recall, F-measure, and average similarity. We corroborate our findings by comparing them against 19 state-of-the-art existing BGS techniques.

A Novel Framework to Generate Synthetic Video for Foreground Detection in Highway Surveillance Scenarios

Foreground detection (FD) plays an important role in the domain of video surveillance for highway. The design of advanced FD algorithms requires large-scale and diverse video dataset. However, collecting and labeling real dataset is still time-consuming, labor-intensive, and highly subjective. To address this issue, we first use computer graphics (CG) to clone real highway scenarios (HS) and generate synthetic multi-challenge video datasets, called “Synthetic-HS (CG)”, automatically labeled with accurate pixel-level ground truth. The Synthetic-HS (CG) dataset contains eight imaging condition sequences for computer vision research. Then, we design an image translation (IT) model that translates source domain (Synthetic-HS (CG)) to target domain (real). This model uses skip connections and attention module to generate realistic synthetic images “Synthetic-HS (IT)”. We use publicly available Synthetic-HS in combination with the corresponding real video sequence to conduct experiments. The experiment results suggest that: 1) The Synthetic-HS (CG) dataset enables us to provide precise quantitative evaluation of the drawbacks of foreground detection methods 2) The realistic Synthetic-HS (IT) images can be used to promote the visual perception in highway video surveillance.

An automated approach for counting groups of flying animals applied to one of the world's largest bat colonies

AbstractEstimating animal populations is essential for conservation. Censusing large congregations is especially important since these are priorities for protection, but efficiently counting hundreds of thousands of moving animals remains a challenge. We developed a deep learning‐based system using consumer cameras that not only counts but also records behavioral information for large numbers of flying animals in a range of lighting conditions including near darkness. We built a robust training set without human labeling by leveraging data augmentation and background subtraction. We demonstrate this approach by estimating the size of a straw‐colored fruit bat (Eidolon helvum) colony in Kasanka National Park, Zambia with cameras encircling the colony to record evening emergence. Detection of bats was robust to deteriorating lighting conditions and changing backgrounds. Combined over five days, our population estimates ranged between 750,000 and 976,000 bats with a mean of 857,233. In addition to counts, we extracted wingbeat frequency, flight altitude, and local group polarity for 639,414 individuals. This open access method is an inexpensive but powerful approach that, in addition to radial emergences from central locations, can also be applied to unidirectional movements of flying groups, such as migratory streams of birds.

Traffic Monitoring System for Vehicle Detection in Day and Night Conditions

Abstract In this work, a day and night time vehicle detection system for traffic surveillance is proposed. Our system is composed of two main processes, day time and night time processes. In the night time, the vehicles are detected based on their taillights and headlights. First of all, the 2D-DWT (Two Dimensional Discrete Wavelet Transform) and the background subtraction are applied to the input image. Then, the connected component technique is used to extract the region of interest. If it is the daytime, the connected component candidates are taken as potential vehicles after applying a pre-processing algorithm to improve the result. If it is the night-time, a filtering operation is used to keep only the bright white and red connected component candidates (which represent potential headlights and taillights, respectively). Finally, potential lamp sets are formed by grouping the extracted components on the basis of their positions, sizes, and colours. The potential extracted vehicles are classified as a vehicle or non-vehicle by using a pre-trained CNN (Convolutional Neural Network) classifier. The proposed system was tested and evaluated using different works from the literature. The experiments show that our proposed system has reached a high accuracy in terms of vehicle detection process whether in day or night time. The experiments were performed using four different videos and were implemented using the C++ language, which facilitates mathematical computation, and its OpenCV library, which is used to run the image processing algorithms used, as well as the TensorFlow library, which facilitates transfer learning of pre-trained CNN models.

Detection of Multiple Phase Space Overdensities of GSE Stars by Orbit Integration

In N-body simulations, nearly radial mergers can form shell-like overdensities in the sky position and phase space (r − v r ) due to the combination of dynamical friction and tidal stripping. The merger event of Gaia-Sausage-Enceladus (GSE) has provided a unique opportunity to study the shells in the phase space. To search for them, we integrate the orbits of 5949 GSE-related halo K giants from the LAMOST survey and record their positions at all time intervals in the r − v r diagram. After the subtraction of a smoothed background, we find six significant and complete thin chevron-like overdensities. The apocenters r apo of stars in the six chevrons are around 6.75, 12.75, 18.75, 25.25, 27.25, and 30.25 kpc. These chevrons reveal the multiple pileups of GSE stars at different apocenters. The application of a different Milky Way mass M vir will change the opening angles of these chevrons, while leaving their apocenters almost unchanged. By comparing with a recent study of the phase space overdensities of local halo stars from the Gaia Radial Velocity Spectrometer survey, our results are more inclined to a medium M vir of 1012 M ⊙. The application of a nonaxisymmetric Galactic potential with a steadily rotating bar has a blurring effect on the appearance of these chevron-like overdensities, especially for the chevrons with r apo > 20 kpc.