Multi-camera Vehicle Tracking Using Local Image Features and Neural Networks
A method for tracking moving objects crossing fields of view of multiple cameras is presented. The algorithm utilizes Artificial Neural Networks (ANNs). Each ANN is trained to recognize images of one moving object acquired by a single camera. Local image features calculated in the vicinity of automatically detected interest points are used as object image parameters. Next, ANNs are employed to identify the same objects captured by other cameras. Object tracking is supplemented by spatial and temporal constraints defining possible transitions between cameras’ fields of view. Experiments carried out were focused on identification of the same vehicles in different cameras. The results achieved prove that the algorithm is sufficiently effective for multi-camera object tracking provided that the cameras’ orientations with respect to moving objects and to the ground are similar.
Keywordsmulti-camera object tracking moving object segmentation local image features SURF object identification
Unable to display preview. Download preview PDF.
- 3.Chilgunde, A., Kumar, P., Ranganath, S., WeiMin, H.: Multi-Camera Target Tracking in Blind Regions of Cameras with Non-overlapping Fields of View. In: British Machine Vision Conference BMVC, Kingston, September 7-9 (2004)Google Scholar
- 7.Czyzewski, A., Dalka, P.: Examining Kalman filters applied to tracking objects in motion. In: 9th Int. Workshop on Image Analysis for Mult. Interact. Services, pp. 175–178 (2008)Google Scholar