| 摘要 |
A method for detecting new objects in a scene illuminated by lights not necessarily constrained. The method detects and produces, in particular with a video camera, in the form of electric signals, video images, at successive times, digitizes the resulting electric signals, and produces image digital data, in particular pixels, at successive times. Then, the image digitalis stored and the type of object, the trajectory, and attitude changes of the object are estimated. |
