کد مقاله کد نشریه سال انتشار مقاله انگلیسی نسخه تمام متن
526980 869268 2014 13 صفحه PDF دانلود رایگان
عنوان انگلیسی مقاله ISI
Recognizing activities in multiple views with fusion of frame judgments
ترجمه فارسی عنوان
تشخیص فعالیت ها در دیدگاه های متعدد با هماهنگی قضاوت های فریم؟
کلمات کلیدی
تجزیه و تحلیل ویدئو، شناسایی فعالیت های انسانی، نمایش چندگانه، دوربین چندگانه
موضوعات مرتبط
مهندسی و علوم پایه مهندسی کامپیوتر چشم انداز کامپیوتر و تشخیص الگو
چکیده انگلیسی


• Data fusion based method for activity recognition using multiple views
• Straightforward architecture to incorporate new cameras or new features
• Performance generally increases when there are more cameras and features.
• Comparable performance with that produced by reconstruction
• Detailed experiments to answer different system considerations

This paper focuses on activity recognition when multiple views are available. In the literature, this is often performed using two different approaches. In the first one, the systems build a 3D reconstruction and match that. However, there are practical disadvantages to this methodology since a sufficient number of overlapping views is needed to reconstruct, and one must calibrate the cameras. A simpler alternative is to match the frames individually. This offers significant advantages in the system architecture (e.g., it is easy to incorporate new features and camera dropouts can be tolerated). In this paper, the second approach is employed and a novel fusion method is proposed. Our fusion method collects the activity labels over frames and cameras, and then fuses activity judgments as the sequence label. It is shown that there is no performance penalty when a straightforward weighted voting scheme is used. In particular, when there are enough overlapping views to generate a volumetric reconstruction, our recognition performance is comparable with that produced by volumetric reconstructions. However, if the overlapping views are not adequate, the performance degrades fairly gracefully, even in cases where test and training views do not overlap.

Figure optionsDownload high-quality image (142 K)Download as PowerPoint slide

ناشر
Database: Elsevier - ScienceDirect (ساینس دایرکت)
Journal: Image and Vision Computing - Volume 32, Issue 4, April 2014, Pages 237–249
نویسندگان
, ,