Continuous tracking within and across camera streams

Advances in Pattern Recognition - Proceedings of the Sixth International Conference, 2007

Continuously Adaptive Mean shift (CAMSHIFT) is a popular algorithm for visual tracking, providing speed and robustness with minimal training and computational cost. While it performs well with a fixed camera and static background scene, it can fail rapidly when the camera moves or the background changes since it relies on static models of both the background and the tracked object. Furthermore it is unable to track objects passing in front of backgrounds with which they share significant colours. We describe a new algorithm, the Adaptive Background CAMSHIFT (ABCshift), which addresses both of these problems by using a background model which can be continuously relearned for every frame with minimal additional computational expense. Further, we show how adaptive background relearning can occasionally lead to a particular mode of instability which we resolve by comparing background and tracked object distributions using a metric based on the Bhattacharyya coefficient.

2009 IEEE Applied Imagery Pattern Recognition Workshop (AIPR 2009), 2009

Transferring responsibility for object tracking in a video scene to computer vision rather than human operators has the appeal that the computer will remain vigilant under all circumstances while operator attention can wane. However, when operating at their peak performance, human operators often outperform computer vision because of their ability to adapt to changes in the scene. While many tracking algorithms are available, background subtraction, where a background image is subtracted from the current frame to isolate the foreground objects in a scene, remains a well proven and popular technique. Under some circumstances, a background image can be obtained manually when no foreground objects are present. In the case of persistent surveillance outdoors, the background has a time evolution due to diurnal changes, weather, and seasonal changes. Such changes render a fixed background scene inadequate. We present a method for estimating the background of a scene utilizing a Kalman filter approach. Our method applies a one-dimensional Kalman filter to each pixel of the camera array to track the pixel intensity. We designed the algorithm to track the background intensity of a scene assuming that the camera view is relatively stationary and that the time evolution of the background occurs much slower than the time evolution of relevant foreground events. This allows the background subtraction algorithm to adapt automatically to changes in the scene. The algorithm is a two step process of mean intensity update and standard deviation update. These updates are derived from standard Kalman filter equations. Our algorithm also allows objects to transition between the background and foreground as appropriate by modeling the input standard deviation. For example, a car entering a parking lot surveillance camera field of view would initially be included in the foreground. However, once parked, it will eventually transition to the background. We present results validating our algorithm's ability to estimate backgrounds in a variety of scenes. We demonstrate the application of our method to track objects using simple frame detection with no temporal coherency.

Lecture Notes in Computer Science, 2007

Robust tracking of objects in video is a key challenge in computer vision with applications in automated surveillance, video indexing, human-computer-interaction, gesture recognition, traffic monitoring, etc. Many algorithms have been developed for tracking an object in controlled environments. However, they are susceptible to failure when the challenge is to track multiple objects that undergo appearance change to due to factors such as variation in illumination and object pose. In this paper we present a tracker based on Bayesian estimation, which is relatively robust to object appearance change, and can track multiple targets simultaneously in real time. The object model for computing the likelihood function is incrementally updated and uses background-foreground segmentation information to ameliorate the problem of drift associated with object model update schemes. We demonstrate the efficacy of the proposed method by tracking objects in image sequences from the CAVIAR dataset.

2006

frames. The systems was tested on a large ground-truthed data set containing hundreds of color and FLIR image sequences. A performance evaluation for the system was performed and the average evaluation results are reported in this paper.

Multiple object tracking in image sequences is an emerging topic of research in the field of image processing and computer vision. This paper presents a robust tracking algorithm for multiple moving object detection and tracking them in dynamic scenes. Our algorithm consists of two important steps. Firstly adaptive GMM background modeling with noise reduction is used for foreground object segmentation in the noisy scenes. Also a background updating method is applied to correct the detection of foregorund objects that bends to the background instantly. Secondly, our object tracking framework uses Extended Kalman Filter to model the nonlinear dynamics and measurement models. Also a new approach for solving data association problem is introduced to determine appropriate association between objects and Kalman Filters which is necessary in multipleobject scenario. The adaptive GMM with noise reduction greatly reduced the false detections in the scenes. The experiment of our proposed algorithm shows good results for multiple object motion tracking in complex dynamic background.

Journal of Visual Communication and Image Representation, 2015

This paper presents an effective method for the detection and tracking of multiple moving objects from a video sequence captured by a moving camera without additional sensors. Moving object detection is relatively difficult for video captured by a moving camera, since camera motion and object motion are mixed. In the proposed method, the feature points in the frames are found and then classified as belonging to foreground or background features. Next, moving object regions are obtained using an integration scheme based on foreground feature points and foreground regions, which are obtained using an image difference scheme. Then, a compensation scheme based on the motion history of the continuous motion contours obtained from three consecutive frames is applied to increase the regions of moving objects. Moving objects are detected using a refinement scheme and a minimum bounding box. Finally, moving object tracking is achieved using a Kalman filter based on the center of gravity of a moving object region in the minimum bounding box. Experimental results show that the proposed method has good performance.

2004

We present a novel approach for continuous detection and tracking of moving objects observed by multiple stationary cameras. We address the tracking problem by simultaneously modeling motion and appearance of the moving objects. The object's appearance is represented using color distribution model invariant to 2D rigid and scale transformation. It provides an efficient blob similarity measure for tracking. The motion models are obtained using a Kalman Filter process, which predicts the position of the moving object in both 2D and 3D. The tracking is performed by the maximization of a joint probability model reflecting objects' motion and appearance. The novelty of our approach consists in integrating multiple cues and multiple views in a Joint Probability Data Association Filter for tracking a large number of moving people with partial and total occlusions. We demonstrate the performance of the proposed method on a soccer game captured by two stationary cameras.

– Computer Vision is the part of " Artificial Intelligence " concerned with the theory behind artificial systems that extract information from images. Within which, Video Surveillance is a term given to monitor the behavior of any kind through videos. It requires person to monitor the CCTV and huge volume of memory to record it. One of the major challenges involved is the huge volume of video storage and retrieval of the same on demand. In order to avoid the depletion of human resources and to detect the suspicious behaviors that threaten safety and security, Intelligent Video Surveillance system (IVS) is required. The proposed work is focused on bringing effective and efficient video surveillance system with added intelligence to avoid human intervention in identifying security threats. In IVS, Extended Kalman filters the Gaussian mixture models are used to detect the moving objects. A tracking algorithm is proposed for tracking the moving objects. It implements position of each group, the recognition of the same group, and the newly appearing and disappearing groups. So, the proposed work IVS, promises the robustness against the environmental influences and speed, which are suitable for the real-time surveillance in detecting and tracking moving objects.

Procedia Computer Science, 2015

In video analytics based systems, an efficient method for segmenting foreground objects from video frames is the need of the hour. Currently, foreground segmentation is performed by modelling background of the scene with statistical estimates and comparing them with the current scene. Such methods are not applicable for modelling the background of a moving scene, since the background changes between the scenes. The scope of this paper includes solving the problem of background modelling for applications involving moving camera. The proposed method is a non-panoramic background modelling technique that models each pixel with a single Spatio-Temporal Gaussian. Experimentation on various videos promises that the proposed method can detect foreground objects from the frames of moving camera with negligible false alarms.

2010 6th International Conference on Emerging Technologies (ICET), 2010

The aim of this paper is to present an algorithm for multiple object tracking and video summarization in a scene filmed by one or several cameras. We propose a computationally efficient real time human tracking algorithm, which can 1) track objects inside the field of view (FOV) of a camera even in case of occlusions; 2) recognize objects that quit and then return on a camera's FOV; 3) recognize objects passing through different cameras FOV. We propose a simple 1-D appearance model, called vertical feature (VF), view and size invariant, which is stored in a database in order to help object recognition. We combine it with other motion features like position and velocity for real-time tracking. We find the k closest matches of current object and select the one whose predicted position is closest to the current object position. Our algorithm shows good capabilities for objects tracking even with the change of object view angle and also with the partial change of shape. We compare our algorithm with appearance based and motion based algorithms and show the advantage of a combined approach.

Indonesian Journal of Electrical Engineering and Computer Science

Detection of Moving Objects and Tracking is one of the most concerned issue and is being vastly used at home, business and modern applications. It is used to identify and track of an entity in a significant way. This paper illustrates the way to detect multiple objects using background subtraction methods and extract each object features by using Speed-Up Robust Feature algorithm and track the features through k-Nearest Neighbor processing from different surveillance videos sequentially. In the detection of object of each frame, pixel difference is calculated with respect to the reference background frame for the detection of an object which is only suitable for any ideal static condition with the consideration of lights from the environment. Thus, this method will detect the complete object and the extracted feature will be carried out for the tracking of the object in the multiple videos by one by one video. It is expected that this proposed method can commendably abolish the impact of the changing of lights.

BT Technology Journal, 2004

This paper aims to address two of the key research issues in computer vision -the detection and tracking of multiple objects in the cluttered dynamic scene -that underpin the intelligence aspects of advanced visual surveillance systems aiming at automated visual events detection and behaviour analysis. We discuss two major contributions in resolving these problems within a systematic framework. Firstly, for accurate object detection, an efficient and effective scheme is proposed to remove cast shadows/highlights with error corrections based on a conditional morphological reconstruction. Secondly, for effective tracking, a temporal-template-based tracking scheme is introduced, using multiple descriptive cues (velocity, shape, colour, etc) of the 2-D object appearance together with their respective variances over time. A scaled Euclidean distance is used as the matching metric, and the template is updated using Kalman filters when a matching is found or by linear mean prediction in the case of occlusion. Extensive experiments are carried out on video sequences from various realworld scenarios. The results show very promising tracking performance.

In this paper, we propose a real-time video-surveillance system for image sequences acquired by a moving camera. The system is able to compensate the background motion and to detect mobile objects in the scene. Background compensation is obtained by assuming a simple translation of the whole background from the previous to the actual frame. Dominant translation is computed on the basis of the tracker proposed by Shi-Tomasi and Tomasi-Kanade. Features to be tracked are selected according to a new intrinsic optimality criterion. Badly tracked features are rejected on the basis of a statistical test. The current frame and the related background, after compensation, are processed by a change detection method in order to obtain a binary image of moving points.Results are presented in the contest of a visual-based system for outdoor environments.

1999

A common method for real-time segmentation of moving regions in image sequences involves "background subtraction," or thresholding the error between an estimate of the image without moving objects and the current image. The numerous approaches to this problem differ in the type of background model used and the procedure used to update the model. This paper discusses modeling each pixel as a mixture of Gaussians and using an on-line approximation to update the model. The Gaussian distributions of the adaptive mixture model are then evaluated to determine which are most likely to result from a background process. Each pixel is classified based on whether the Gaussian distribution which represents it most effectively is considered part of the background model.