| 摘要 |
A moving object tracking method is provided to simplify an image acquirement process in an image extraction algorithm by removing implementation complexity. A still image of one frame of a video frame received from a camera is divided into a predetermined size of a matrix form(S530). With respect to each partial area, pattern data of the area is extracted by an image extraction algorithm(S540). A still image of a continuously received video frame is divided into a predetermined size of a matrix form(S630). With respect to each partial area, pattern data of the area is extracted by the image extraction algorithm. The extracted pattern data is compared with the pre-extracted pattern data. If a weighted value is differently extracted in pattern data of a specific matrix area, it is determined that a moving object moves to the corresponding matrix area(S760). By comprehensively comparing average pattern data with the above pattern data, a movement direction of the moving object is detected(S800).
|
