Background Subtraction Technique Research Articles

Position Tracking and Fuzzy Logic

In this paper, the problem of tracking the position of an object from a remote place is solved by applying the rules of fuzzy logic. The domain of image analysis and artificial intelligence has been integrated so as to track the position of the object and get results that are more precise. The object has been identified and then tracked until it has exceeded the range of a camera. Various techniques have been used at various levels of the entire process. Fuzzy rules have been made so as to accomplish the goal. A video has been first converted into frames and then using technique of background subtraction, the object has been identified and then tracked which is an optimal solution as well as precise.

Crowd modeling framework using fast head detection and shape-aware matching

A framework for crowd modeling using a combination of multiple kernel learning (MKL)-based fast head detection and shape-aware matching is proposed. First, the MKL technique is used to train a classifier for head detection using a combination of the histogram of oriented gradient and local binary patterns feature sets. Further, the head detection process is accelerated by implementing the classification procedure only at those spatial locations in the image where the gradient points overlap with moving objects. Such moving objects are determined using an adaptive background subtraction technique. Finally, the crowd is modeled as a deformable shape through connected boundary points (head detection) and matched with the subsequent detection from the next frame in a shape-aware manner. Experimental results obtained from crowded videos show that the proposed framework, while being characterized by a low computation load, performs better than other state-of-art techniques and results in reliable crowd modeling.

In vivo molecular contrast OCT imaging of methylene blue.

An 830-nm spectral-domain optical coherence tomography (OCT) system with an integrated 663-nm diode pump laser has been developed to enable molecular contrast OCT imaging of methylene blue (MB), a common vital dye used clinically. The introduction of the 663-nm diode laser, which acts as the pump in this implementation of pump-probe OCT (PPOCT), represents a minor modification to an otherwise typical OCT system. A newly developed background subtraction technique completely removes all background from intensity noise at the pump modulation frequency, simplifying the interpretation of PPOCT images. These developments have enabled the first in vivo imaging of MB with PPOCT. Volumetric images of a zebrafish, stained by submersion in a 0.01% (w/v) solution of MB for 6h, show accumulation of MB in the mesonephros, the primordial filtration organ.

Evaluation of Recursive Background Subtraction Algorithms for Real-Time Passenger Counting at Bus Rapid Transit System

This work discusses a method to count the number of passengers waiting in Bus Rapid Transit station. The proposed system relies on computer vision technique to monitor the movement of passengers crossing doors on the station. In this work, three background subtraction techniques, namely, Running Gaussian Average, Gaussian Mixture Model, and Adaptive Gaussian Mixture Model, were used to count the passengers crossing an entrance on a BRT station from a pre-recorded motion picture. The results indicates that the tree algorithms are able to identify the passenger crossing with a reasonable high level of recall and but low level of precision. These results indicates that many false positives are identified by the three algorithms. In addition, the empirical data indicate that the three algorithms tend to have better performance with higher value of the learning rate.

Application of Object Detection and Tracking Techniques for Unmanned Aerial Vehicles

In this research, the information captured by Unmanned Aerial Vehicles (UAVs) are eminently utilized in detecting and tracking moving objects which pose a primary security threat against the United States southern border. Illegal trespassing and border encroachment by immigrants is a huge predicament against the United States border security force and the Department of Homeland Security. It becomes insurmountable to warranty suspicious behaviour, monitoring by human operators for long periods of time, due to the massive amount of data involved. The main objective of this research is to assist the human operators, by implementing intelligent visual surveillance systems which help in detecting and tracking suspicious or unusual events in the video sequence. The visual surveillance system requires fast and robust methods of detecting and tracking moving objects. In this research, we have investigated methods for detecting and tracking objects from UAVs. Moving objects were detected using adaptive background subtraction technique successfully and these detected objects were tracked by using Lucas-Kanade optical flow tracking, Continuously Adaptive Mean-Shift tracking based techniques. The simulation results show the efficacy of these techniques in detecting and tracking moving objects in the video sequences acquired by the UAV.

Compressive Background Modeling for Foreground Extraction

Robust and efficient foreground extraction is a crucial topic in many computer vision applications. In this paper, we propose an accurate and computationally efficient background subtraction method. The key idea is to reduce the data dimensionality of image frame based on compressive sensing and in the meanwhile apply sparse representation to build the current background by a set of preceding background images. According to greedy iterative optimization, the background image and background subtracted image can be recovered by using a few compressive measurements. The proposed method is validated through multiple challenging video sequences. Experimental results demonstrate the fact that the performance of our approach is comparable to those of existing classical background subtraction techniques.

Diffuse Galactic gamma-ray emission with H.E.S.S.

Diffuse $\gamma$-ray emission is the most prominent observable signature of celestial cosmic-ray interactions at high energies. While already being investigated at GeV energies over several decades, assessments of diffuse $\gamma$-ray emission at TeV energies remain sparse. After completion of the systematic survey of the inner Galaxy, the H.E.S.S. experiment is in a prime position to observe large-scale diffuse emission at TeV energies. Data of the H.E.S.S. Galactic Plane Survey are investigated in regions off known $\gamma$-ray sources. Corresponding $\gamma$-ray flux measurements were made over an extensive grid of celestial locations. Longitudinal and latitudinal profiles of the observed $\gamma$-ray fluxes show characteristic excess emission not attributable to known $\gamma$-ray sources. For the first time large-scale $\gamma$-ray emission along the Galactic Plane using imaging atmospheric Cherenkov telescopes has been observed. While the background subtraction technique limits the ability to recover modest variation on the scale of the H.E.S.S. field of view or larger, which is characteristic of the inverse Compton scatter-induced Galactic diffuse emission, contributions of neutral pion decay as well as emission from unresolved $\gamma$-ray sources can be recovered in the observed signal to a large fraction. Calculations show that the minimum $\gamma$-ray emission from $\pi^0$-decay represents a significant contribution to the total signal. This detection is interpreted as a mix of diffuse Galactic $\gamma$-ray emission and unresolved sources.

Modeling the Raman spectrum of graphitic material in rock samples with fluorescence backgrounds: accuracy of fitting and uncertainty estimation.

We propose a robust technique called Savitzky-Golay second-derivative (SGSD) fitting for modeling the in situ Raman spectrum of graphitic materials in rock samples such as carbonaceous chondrite meteorites. In contrast to non-derivative techniques, with assumed locally linear or nth-order polynomial fluorescence backgrounds, SGSD produces consistently good fits of spectra with variable background fluorescence of any slowly varying form, without fitting or subtracting the background. In combination with a Monte Carlo technique, SGSD calculates Raman parameters (such as peak width and intensity) with robust uncertainties. To explain why SGSD fitting is more accurate, we compare how different background subtraction techniques model the background fluorescence with the wide and overlapping peaks present in a real Raman spectrum of carbonaceous material. Then, the utility of SGSD is demonstrated with a set of real and simulated data compared to commonly used linear background techniques. Researchers may find the SGSD technique useful if their spectra contain intense background interference with unknown functional form or wide overlapping peaks, and when the uncertainty of the spectral data is not well understood.

Improved visual background extractor using an adaptive distance threshold

Camouflage is a challenging issue in moving object detection. Even the recent and advanced background subtraction technique, visual background extractor (ViBe), cannot effectively deal with it. To better handle camouflage according to the perception characteristics of the human visual system (HVS) in terms of minimum change of intensity under a certain background illumination, we propose an improved ViBe method using an adaptive distance threshold, named IViBe for short. Different from the original ViBe using a fixed distance threshold for background matching, our approach adaptively sets a distance threshold for each background sample based on its intensity. Through analyzing the performance of the HVS in discriminating intensity changes, we determine a reasonable ratio between the intensity of a background sample and its corresponding distance threshold. We also analyze the impacts of our adaptive threshold together with an update mechanism on detection results. Experimental results demonstrate that our method outperforms ViBe even when the foreground and background share similar intensities. Furthermore, in a scenario where foreground objects are motionless for several frames, our IViBe not only reduces the initial false negatives, but also suppresses the diffusion of misclassification caused by those false negatives serving as erroneous background seeds, and hence shows an improved performance compared to ViBe.

Advanced Moving Object Detection by Using K-means Algorithm

Moving Object Detection systems have long been in use to look at security sensitive areas. The manufacture of video investigating systems sensible desires fast, reliable and durable algorithms for moving object detection classification pursuit and activity analysis. Moving object detection is that the essential step for extra analysis of video. It handles segmentation of moving objects from stationary background objects. Object classification step categorizes detected objects into preened classes like human, vehicle, animal, clutter, etc. it's a necessity to inform apart objects from each other therefore on trace and analyse their actions reliably. In previous system we have got performed background subtraction by victimization clever Edge Detection. In clever Edge Detection technique we tend to tend to square measure taking 2 photos for comparison those square measure background image and foreground image. Previous strategies for object detection square measure methodology along side object detectors image segmentation and Background subtraction. Our methodology aims to section objects supported motion info and it contains a part of background modelling. In the previous strategies we tend to tend to square measure conducting background subtraction only for photos. For this we tend to tend to projected an image component wise background modelling and subtraction technique mistreatment multiple choices. Hence, throughout this color, gradient and Hear-Like choices square measure integrated to handle the variation element. Thus, motion segmentation and background subtraction square measure the foremost connected topics to the current paper.

Infrared Target Detection Based on Temporal-Spatial Domain Fusion

In order to improve the accuracy and stability of infrared target detection, a novel moving target detection approach based on temporal-spatial domain fusion is presented. A multi-level spatial-temporal median filter is utilized to extract the background frame, with which the background clutters are suppressed by using the background subtraction technique. Then a local weighted operator is applied to enhance the targets. Lastly, the otsu thresholding algorithm is utilized to detect the targets. Experimental results demonstrate that the proposed approach is capable of detecting infrared moving targets effectively for F1 measurement up to 92.8%.

Block-Sparse RPCA for Salient Motion Detection.

Recent evaluation [2], [13] of representative background subtraction techniques demonstrated that there are still considerable challenges facing these methods. Challenges in realistic environment include illumination change causing complex intensity variation, background motions (trees, waves, etc.) whose magnitude can be greater than those of the foreground, poor image quality under low light, camouflage, etc. Existing methods often handle only part of these challenges; we address all these challenges in a unified framework which makes little specific assumption of the background. We regard the observed image sequence as being made up of the sum of a low-rank background matrix and a sparse outlier matrix and solve the decomposition using the Robust Principal Component Analysis method. Our contribution lies in dynamically estimating the support of the foreground regions via a motion saliency estimation step, so as to impose spatial coherence on these regions. Unlike smoothness constraint such as MRF, our method is able to obtain crisply defined foreground regions, and in general, handles large dynamic background motion much better. Furthermore, we also introduce an image alignment step to handle camera jitter. Extensive experiments on benchmark and additional challenging data sets demonstrate that our method works effectively on a wide range of complex scenarios, resulting in best performance that significantly outperforms many state-of-the-art approaches.

WHITEBOARD IMAGE RECONSTRUCTION USING MATLAB

Online video lectures are becoming commonplace in higher education. One problem is that the instructor might block what he has written on the board. In our project, we reconstruct the online lecture video so that the information on the whiteboard is always available to the viewers. The background subtraction, object tracking, and object replacement technique have been used to reconstruct the video. The complete access of the whiteboard information is obtained by replacing the lecturer/object with whiteboard information from the future frames or the past. The object is detected by using background subtraction method and is tracked by object tracking technique. The algorithm can be implemented for different board conditions and video resolutions. Keywordsbackground subtraction, object tracking, whiteboard reconstruction, object replacement, online lectures, and video transformation -------------------------------------------------------------------***-------------------------------------------------------------------

Statistically tuned Gaussian background subtraction technique for UAV videos

Background subtraction is one of the efficient techniques to segment the targets from non-informative background of a video. The traditional background subtraction technique suits for videos with static background whereas the video obtained from unmanned aerial vehicle has dynamic background. Here, we propose an algorithm with tuning factor and Gaussian update for surveillance videos that suits effectively for aerial videos. The tuning factor is optimized by extracting the statistical features of the input frames. With the optimized tuning factor and Gaussian update an adaptive Gaussian-based background subtraction technique is proposed. The algorithm involves modelling, update and subtraction phases. This running Gaussian average based background subtraction technique uses updation at both model generation phase and subtraction phase. The resultant video extracts the moving objects from the dynamic background. Sample videos of various properties such as cluttered background, small objects, moving background and multiple objects are considered for evaluation. The technique is statistically compared with frame differencing technique, temporal median method and mixture of Gaussian model and performance evaluation is done to check the effectiveness of the proposed technique after optimization for both static and dynamic videos.

Moving Object Detection using Lab2000HL Color Space with Spatial and Temporal Smoothing

In order to detect moving objects such as vehicles in motorways, background subtraction techniques are commonly used. This is completely solved problem for static backgrounds. However, real-world problems contain many non-static components such as waving sea, camera oscillations, and sudden changes in daylight. Gaussian Mixture Model (GMM) is statistical based background subtraction method, in which values of each pixels features are represented with a few normal distributions, partially overcame such problems at least. To improve performance of GMM model, using spatial and temporal features in Lab2000HL color space which have linear hue band, is proposed in this study. The spatial and temporal featu res performed by using spatial low-pass filter and temporal kalman filter respectively. As a performance metric, the area under the Precision Recall (PR) curve is used. In addition to videos existing in the I2R dataset, a new dataset which images gained from traffic surveilla nce cameras placed over the entrance of the Istanbul FSM Bridge at different times of the day used for compare proposed method against other well-known GMM version. According to our tests proposed method has been more successful to the other methods in most cases.

VISTA: achieving cumulative VIsion through energy efficient Silhouette recognition of mobile Targets through collAboration of visual sensor nodes

Visual sensor networks (VSNs) are innovative networks founded on a broad range of areas such as networking, imaging, and database systems. These networks demand well-defined architectures in terms of sensor nodes and camera deployment, image capturing and processing, and well-organized distributed systems. This makes existing VSN architectures deficient because these are limited in approach and in design. In this paper, we propose VISTA, a distributed vision multi-layer architecture aimed at constructing the cumulative vision of mobile objects (MOs). VISTA realizes silhouette recognition of mobile targets through (a) pre-meditated deployment of sensor nodes (SNs) that are equipped with sonar sensors and fixed view (FV) on-board cameras present at the periphery of region of interest (RoI) and SNs with only on-board cameras within RoI, (b) pre-distribution of silhouettes of known objects across SNs, (c) sonar-based presence detection of MO at the outskirts of RoI, (d) MO silhouette capturing and matching at interior node to determine the % age match, (e) subsequent activation of next interior cameras in order to improve % age match, and (f) terminating further activation upon threshold recognition of MO. Experimental evaluation of our image processing algorithms against baseline algorithms with respect to execution time and memory shows significant reduction in image data and memory occupancy. Also, experiments show that true match is achieved fully under broad daylight conditions and large backgrounds when our proposed background subtraction and pixel reduction techniques are used. The mobility-driven behavior of associated network layer algorithms of VISTA is simulated in a network simulator (NS2) by representing the surety of MO identification as a function of number of cameras, database size and distribution, MO’s trajectory, stored perspectives, and network depth. The simulation results show that doubling and, in some situations, manifold increase is observed in the surety of the target with an increase in the number of silhouettes deployed against the baselined database size and mobility model. The results substantiate that VISTA is a suitable architecture for low-cost, autonomous and efficient human and asset monitoring surveillance, friend-or-foe (FoF) identification, and target tracking systems.

REAL-TIME OBJECT TRACKING AND LEARNING USING TEMPLATE MATCHING

The system that has been proposed here uses a live video stream to enable tracking, learning and detection of real time objects. The object of interest is selected in a cropping action and then it is subsequently tracked with the help of an indicator in the form of a rectangular bounding box. The process mentioned above is a long term process in the sense that the system is perpetually alert regarding the reappearance of the object after a departure from the frame or regarding the deformities in the physical appearance of the object. The proposed system uses the Template Matching algorithm to match the selected object with the ‘Region of Interest’ in the frame to mark the object’s location. If a match is found then the Principle Component Analysis algorithm is used. PN discrimination algorithm has been proposed in this system which uses background subtraction technique to increase the speed of frame processing for object detection. This reduction in frame processing time and the reduction in average localization errors improve the template matching percentage irrespective of scaling of the input image. Thus the proposed system is expected to overcome the drawbacks of existing system which range from loss of information caused by complex shapes, rapid motion, illumination changes, scaling and projection of 3D world on 2D image, etc.

Moving Object Detection with Fixed Camera and Moving Camera for Automated Video Analysis

Detection of moving objects in a video sequence is a difficult task and robust moving object detection in video frames for video surveillance applications is a challenging problem. Object detection is a fundamental step for automated video analysis in many vision applications. Object detection in a video is usually performed by object detectors or background subtraction techniques. Frequently, an object detector requires manual labeling, while background subtraction needs a training sequence. To automate the analysis, object detection without a separate training phase becomes a critical task. This paper presents a survey of various techniques related to moving object detection and discussed the optimization process that can lead to improved object detection and the speed of formulating the low rank model for detected object.

Implementation and Performance Evaluation of Background Subtraction Algorithms

The study evaluates three background subtraction techniques. The techniques ranges from very basic algorithm to state of the art published techniques categorized based on speed, memory requirements and accuracy. Such a review can effectively guide the designer to select the most suitable method for a given application in a principled way. The algorithms used in the study ranges from varying levels of accuracy and computational complexity. Few of them can also deal with real time challenges like rain, snow, hails, swaying branches, objects overlapping, varying light intensity or slow moving objects.

Efficient Motion Detection Algorithm in Video Sequences

Human motion analysis concerns the detection, tracking and recognition of people behaviors, from image sequences involving humans. A reference frame is initially used and considered as background information. While a new object enters into the frame, the foreground information and background information are identified using the reference frame as background model.In this paper, an efficient algorithm is proposed for objects detection in real time video sequences. The method aims at tracking an object like (human) in motion using background subtraction technique. The tracked objects are subject to different pre and post image processing in order to extract the most important features (frame preprocessing is used for shadow removal using morphological operations). They can be used later for recognition in different security system such as (human detection for surveillance). Experimental results show that the proposed method is very efficient in terms of reliability and accuracy of detection.