會議論文
| 學年 | 97 |
|---|---|
| 學期 | 1 |
| 發表日期 | 2008-11-21 |
| 作品名稱 | Moving Object Detection and Tracking |
| 作品名稱（其他語言） | |
| 著者 | Lin, Hwei-Jen; Liang, Feng-Ming; Wang, Chun-Wei; Yang, Fu-Wen |
| 作品所屬單位 | 淡江大學資訊工程學系 |
| 出版者 | 臺北縣淡水鎮 : 淡江大學 |
| 會議名稱 | 第十三屆人工智慧與應用研討會=The 13th conference on artificial intelligence and applications |
| 會議地點 | 宜蘭縣, 臺灣 |
| 摘要 | For object detection and tracking, we use amodified version of Gaussian Mixture Models(GMMs) to construct background, which is thensubtracted from the image to obtain the foregroundwhere the moving objects locate. We then performsome operations, including shadow removal, edgedetection, and connected component analysis, tolocalize each moving object in the foreground. As soon as an object is detected it is then trackedin the following frames by the use of Particle Filters(PF). PF is effective but the dimension of its statespace is high so as the tracked objects tend to beshifting. To reduce this problem we modify theparticle filtering by carrying out tracking over theforeground portion instead of the whole image. Withthe use of the modified versions of GMMs and PFs,our system was proved to have high accuracy rate ofdetection/tracking and satisfactory time efficiency. |
| 關鍵字 | Detection;Tracking;Gaussian mixture model (GMM);Particle filter (PF);Sequential K means algorithm;Expectation maximization (EM) |
| 語言 | en |
| 收錄於 | |
| 會議性質 | 國內 |
| 校內研討會地點 | 蘭陽校園 |
| 研討會時間 | 20081121~20081122 |
| 通訊作者 | |
| 國別 | TWN |
| 公開徵稿 | Y |
| 出版型式 | 紙本 |
| 出處 | 第十三屆人工智慧與應用研討會論文集=The 13th conference on artificial intelligence and applications, pp.809-813 |
| 相關連結 |
機構典藏連結 ( http://tkuir.lib.tku.edu.tw:8080/dspace/handle/987654321/95854 ) |