Video analysis of human dynamics is an important area of research devoted to detecting people and understanding their dynamic physical behavior in a complex environment that can be used for biometric applications. This paper provides a detailed survey of the various studies in areas related to the tracking of people and body parts such as face, hands, fingers, legs, etc., and modeling behavior using motion analysis.

Video analysis of human dynamics—a survey

This paper describes a new real-time approach for detecting changes in grey level image sequences, which were taken from stationary cameras. This new method combines a temporal difference method with an adaptive background model subtraction scheme. When changes in illumination occur the background model is automatically adapted to suit the new conditions. For the adaptation of the background model a new method is proposed, which avoids reinforcement of adaptation errors by performing the adaptation solely on those regions that were detected by the temporal difference method rather than using the regions resulting from the overall algorithm. Thus the adaptation process is separated from the results of its own background subtraction algorithm. The change detector was successfully tested both in a vision-based workspace monitoring system for different kinds of non-autonomous service robots and in a surveillance scenario, in which it was the task to detect people in a subway-platform scenario. The proposed real-time algorithm showed recognition rates of up to 90% in the foreground and 84% in the background and performed in all cases at least 12% better than the alternative method of adaptive background estimation which uses a modified Kalman filtering technique.

Adaptive change detection for real-time surveillance applications

Presents a novel approach to the tracking of multiple articulate objects in the presence of occlusion in moderately complex scenes. Most conventional tracking algorithms work well when only one object is tracked at a time. However, when multiple objects must be tracked simultaneously, significant computation is often introduced in order to handle occlusion and to calculate the appropriate region correspondence between successive frames. We introduce a near-real-time solution to this problem by using a probabilistic mixing of low-level features and components. The algorithm mixes coarse motion estimates, change detection information and unobservable predictions to create accurate trajectories of moving objects. We implement this multifeature mixing strategy within the context of a video surveillance system using a modified Kalman filtering mechanism. Experimental results demonstrate the efficacy of the proposed tracking and surveillance system.

Tracking multiple objects in the presence of articulated and occluded motion

This paper presents a novel approach to the tracking of multiple articulate objects in the presence of occlusion in moderately complex scenes. Most conventional tracking algorithms work well when only one object is tracked at a time. However, when multiple objects must be tracked simultaneously, significant computation is often introduced in order to handle occlusion and to calculate the appropriate region correspondence between successive frames. We introduce a near real-time solution to this problem by using a probabilistic mixing of low-level features and components. The algorithm mixes coarse motion estimates, change detection information, and unobservable predictions to create accurate trajectories of moving objects. We implement this multi-feature mixing strategy using a modified Kalman filtering mechanism. The proposed technique is presented within the context of a video surveillance system, where we further introduce a simple color-based technique for automatically extracting facial regions of moving persons. The single, most descriptive, facial snapshot for each person is added to a database that, when used in conjunction with the tracking data, can be used as a compact, semantic representation for the video sequence. Experimental results demonstrate the efficacy of the proposed tracking and surveillance system.

On the tracking of articulated and occluded video object motion

Computer vision has been a widely developed research area in the last years, and it has been used for a broad range of applications, including surveillance systems. In the pursuit of an autonomous and smart motion detection system, a reliable segmentation algorithm is required. The main problems of present segmentation solutions are their high execution time and the lack of robustness against changes in the environment due to variations in lighting, shadows, occlusions or the movement of secondary objects. This paper proposes a new algorithm named Mixture of Merged Gaussian Algorithm (MMGA) that aims to achieve a substantial improvement in execution speed to enable real-time implementation, without compromising the reliability and accuracy of the segmentation. The MMGA is based on the combination of a probabilistic model for the background, similar to the Mixture of Gaussian Model (MGM), with the learning processes of Real-Time Dynamic Ellipsoidal Neural Networks (RTDENN) for the update of the model. The proposed algorithm has been tested for different videos and compared to the MGM and SDGM algorithms. Results show a reduction of 30 to 50 % in execution times. Furthermore, the segmentation is more robust against the effect of noise and adapts faster to lighting changes.

/pdf/mixture-of-merged-gaussian-algorithm-using-rtdenn-4hnrr18mrs.pdf

Mixture of Merged Gaussian Algorithm using RTDENN

The proposed system STABIL combines four levels of abstraction. In the first level the foreground is extracted by using a Kalman filter technique. The second level uses the foreground regions in order to seek subsequently for parts of the skin of a human. The three channel color signal is transformed into a 2D color space best representing the color of the skin. The Kalman-filtering speeds up the classification in the case of a stationary camera; in the case of a moving camera the classification is directly applied on the sequence. The regions representing the skin serve as input for the third level estimating the position of the person in the 3D space relative to the camera. The fourth level handles a model of a human using statistical data of their size. The model is adapted to a person in an iterative process taking into account the limits of the movements of the person, the restrictions on the model in order to match the skin regions in the image to the correct person. The result of processing the n-the image of an image sequence is a scaled model projected and superimposed on the n-1-th image showing the correct estimation of the position of the person.© (1996) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

Olaf Muenkelt

Papers

STABIL: a system for monitoring persons in image sequences