Target tracking using transputers as parallel processors
Yazarlar (3)
| Bildiri Türü | Tebliğ/Bildiri | Bildiri Dili | İngilizce |
| Bildiri Alt Türü | Tam Metin Olarak Yayınlanan Tebliğ (Uluslararası Kongre/Sempozyum) | ||
| Bildiri Niteliği | Alanında Hakemli Uluslararası Kongre/Sempozyum | ||
| Kongre Adı | 9th IFAC Symposium on Identification and System Parameter Estimation | ||
| Kongre Tarihi | 08-07-1991 / | ||
| Basıldığı Ülke | Macaristan | Basıldığı Şehir | |
| UAK Araştırma Alanları |
Bilgisayar Yazılımı ve Yazılım Mühendisliği
Gömülü Sistemler
Makine Öğrenmesi
|
||
| Özet |
| The paper discusses several aspects of the problem of tracking multiple targets from noisy measurement data using the track splitting algorithm. Particular attention is focussed on the performance of the sub-optimal α — β filter relative to that of the Kalman filter. The former is often used to speed up the computation time but it is shown that due to the difficulty of data association in multiple target tracking the benefits of using the α — β filter in place of the optimal Kalman filter are not as great as expected. Because the track splitting algorithm requires multiple filters it is particularly appropriate for implementation on parallel processors. Several ways of implementing the algorithm on transputer arrays are given and the simulation results show their relative advantages. |
| Anahtar Kelimeler |
