Presentation | 2011-03-10 Abnormal Motion Detection Using Temporal and Spatial Information Hisashi MIKADO, Kazunori ONOGUCHI, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | This paper presents a method to detect abnormal motion by using spatio-temporal features. At first, silhouette images obtained from the backgrouund subtraction are separated into the upper part and the lower part. Each part is separated into some blocks so that the number of pixels in each block may be distributed equally. Next, the fourier transform is applied to time series data consisting of the pixel density in each block and the motion features are detected. In addition to the motion features, the shape features are detected by calculating the mean and the variance of the pixel density in each block. The classifier using one class SVM is created by leaning feature vectors generated from this process. Our method can detect both dynamic abnormal motions and static ones. In this experiment, walking, running or stopping was detected as normal motion,and the others were abnormal. The validity of the method was confirmed by experimental results. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Abnormal Detection / Discrete Fourier Transform / One Class SVM |
Paper # | PRMU2010-257 |
Date of Issue |
Conference Information | |
Committee | PRMU |
---|---|
Conference Date | 2011/3/3(1days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | |
Chair | |
Vice Chair | |
Secretary | |
Assistant |
Paper Information | |
Registration To | Pattern Recognition and Media Understanding (PRMU) |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | Abnormal Motion Detection Using Temporal and Spatial Information |
Sub Title (in English) | |
Keyword(1) | Abnormal Detection |
Keyword(2) | Discrete Fourier Transform |
Keyword(3) | One Class SVM |
1st Author's Name | Hisashi MIKADO |
1st Author's Affiliation | Faculty of Science and Technology, Hirosaki University() |
2nd Author's Name | Kazunori ONOGUCHI |
2nd Author's Affiliation | Faculty of Science and Technology, Hirosaki University |
Date | 2011-03-10 |
Paper # | PRMU2010-257 |
Volume (vol) | vol.110 |
Number (no) | 467 |
Page | pp.pp.- |
#Pages | 6 |
Date of Issue |