Detection and Recognition of Moving Objects by Using Motion Invariants

Motion invariant features, which are called either absolute or relative invariant linear features, are useful for detection and recognition of moving objects from image streams. We conducted simple experiments using these invariant features. The first experiment tested our method for moving object recognition with a fixed camera using absolute invariant linear features. In the second experiment, another method was used to detect moving objects with a moving camera using both relative and absolute invariant linear features. The results of these experiments confirmed the effectiveness of the proposed methods