اهرم مدل رفتار انسان برای پیش بینی مسیر در محیط های داخلی
|کد مقاله||سال انتشار||تعداد صفحات مقاله انگلیسی||ترجمه فارسی|
|28117||2011||12 صفحه PDF||سفارش دهید|
نسخه انگلیسی مقاله همین الان قابل دانلود است.
هزینه ترجمه مقاله بر اساس تعداد کلمات مقاله انگلیسی محاسبه می شود.
این مقاله تقریباً شامل 6640 کلمه می باشد.
هزینه ترجمه مقاله توسط مترجمان با تجربه، طبق جدول زیر محاسبه می شود:
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Pervasive and Mobile Computing, Volume 7, Issue 3, June 2011, Pages 319–330
One of the most powerful constraints governing many activity recognition problems is that imposed by the human actor. It is well known that humans have a large set of physical and cognitive limitations that constrain their execution of various tasks. In this article, we show how prior knowledge of these perception and locomotion limitations can be exploited to enhance path prediction and tracking in indoor environments for pervasive computing applications. We demonstrate an approach for path prediction based on a model of visually guided steering that has been validated on human obstacle avoidance data. Our approach outperforms standard motion models in a particle filter tracker during occlusion periods of greater than one second and results in a significant reduction in SSD tracking error.
The ability to predict the path of a moving human is a crucial element in a wide range of applications, including video surveillance, assisted living environments (smart homes), and simulation environments. Two tasks, tracking (finding the user’s current location) and goal prediction (identifying the final destination) are particularly relevant to many problems. Human trajectory prediction can serve as a useful supplement to activity recognition in facilitating location-aware user notifications. For instance, recognizing activities of daily living might reveal that the user is engaged in housecleaning; trajectory prediction would enable a home monitoring system to warn the user about a possible tripping hazard. In some cases, predicting the user’s trajectory is more important for recognizing user intent than activity recognition; imagine a parking lot surveillance system attempting to match cars to pedestrians. In this case, the recognition that the users are walking would provide comparatively little information relevant to predicting the user’s trajectory and/or destination. Although standard path planning approaches can be used to predict human behavior at a macroscopic level, they do not accurately model human path preferences. In this article, we demonstrate an approach for path prediction based on a model of visually guided steering that has been validated on human obstacle avoidance data. By basing our path prediction on egocentric features that are known to affect human steering preferences, we can improve on strictly geometric models such as Voronoi diagrams . Our approach outperforms standard motion models in a particle filter tracker and can also be used to discriminate between multiple user destinations. To track humans with sensor networks , detect behavior anomalies , and offer effective navigational assistance , we need to be able to predict the trajectory that a human will follow in an environment. Although human paths can be approximated by a minimal distance metric, humans often exhibit counter-intuitive behaviors; for instance, human paths can be non-symmetric and depend on the direction of path traversal (e.g., humans walking one route and returning via a different one) . Obviously tracking and goal prediction algorithms that assume distance-minimizing behavior will generate errors in environments where the humans’ behavior diverges from this model. To address this problem, we sought a psychologically grounded model of human steering and obstacle behavior to incorporate into our tracking and goal prediction system. Our selected model, originally proposed by Fajen et al.  incorporates environmental features that are accessible to the human vision system into a second-order dynamical model; all calculations are based on local perceptual information and do not require global knowledge of the environment. In this article, we demonstrate that a particle filter tracking system based on this human steering model outperforms other commonly used motion models. We trained and evaluated our model in two different scenarios (1) individual subjects navigating an obstacle course in a simulated environment and (2) multiple humans moving around an office environment.
نتیجه گیری انگلیسی
Using physical constraints to limit the search space across models of human activity is a potentially powerful tool since there are many activity recognition problems that utilize human motion data from video, motion capture, and body-mounted inertial measurement units. Pre-existing models of human physical and information processing from human psychology could be leveraged toward improving recognition in a wide variety of problems. The research challenge is in determining how to use models to aid the recognition and prediction process. In this article, we demonstrate an approach for path prediction and user tracking in both the Second Life virtual world and an indoor office environment; our method is applicable to any environment (real or virtual) with known geometry. The ability to predict the path of a moving human is a crucial element in a wide range of pervasive computing applications, including video surveillance, assisted living environments (smart homes), and simulation systems. Although standard path planning approaches can be used to predict human behavior at a macroscopic level, they do not accurately model human angular acceleration preferences. We propose an alternate approach for path prediction based on a second-order dynamical human steering model originally introduced by Fajen et al.  and confirm the applicability of the HSM for our task by showing statistically significant path fit improvements over standard planning models. By basing the tracker’s motion model on human steering preferences, we improve on strictly geometric models such as Voronoi diagrams. Additionally, our model is robust to occlusions and outperforms the commonly used constant velocity motion model during sensor blackouts.