System and method for object identification and behavior characterization using video analysis

作者: Linda Crnic , Yiqing Liang , Vikrant Kobla , Wayne Wolf

DOI:

关键词: Motion detectionVideo cameraObject (computer science)Artificial intelligenceVideo trackingComputer visionIdentification (information)SegmentationProbabilistic logicComputer scienceBackground subtraction

摘要: In general, the present invention is directed to systems and methods for finding position shape of an object using video. The includes a system with video camera coupled computer in which configured automatically provide segmentation identification, motion tracking (for moving objects), classification, behavior identification. preferred embodiment, may use background subtraction identification tracking, probabilistic approach expectation-maximization detection decision tree classification Thus, capable monitoring image identify, track classify actions various objects object's movements within image. be provided real time or from storage. particularly useful classifying animal testing drugs genetic mutations, but used any number other surveillance applications.

参考文章(92)
Alan P. Jones, Linda S. Crnic, Maternal Mediation of the Effects of Malnutrition Springer, Boston, MA. pp. 409- 425 ,(1986) , 10.1007/978-1-4613-2189-7_17
LINDA S. CRNIC, Behavioral Consequences of Virus Infection Psychoneuroimmunology (Second Edition). pp. 749- 769 ,(1991) , 10.1016/B978-0-12-043780-1.50033-6
Gheorghe Tecuci, Yinqing Liang, Apprenticeship Learning of Domain Models. software engineering and knowledge engineering. pp. 54- 62 ,(1995)
Linda S. Crnic, Nutrition and mental development. American journal of mental deficiency. ,vol. 88, pp. 526- 533 ,(1984)
Katsumi Higuchi, Keizo Maeda, Hiroshi Hakata, Apparatus for automatically measuring the movement of an animal ,(1965)
Michael Philips, Wayne Wolf, A multi-attribute shot segmentation algorithm for video programs Telecommunication Systems. ,vol. 9, pp. 393- 402 ,(1998) , 10.1023/A:1019164327291