Background model update method for image process
The present invention relates to a background model update method for image process. The accuracy of distinguishing a foreground image from the background model can be improved by adjusting the background model update level according to image variations such as brightness variation, motion variation, color variation, etc.
1. A background model update method for an image sensing chip including an image sensor, comprising:
determining whether an object enters a detection region of the image sensor, wherein the detection region is a three-dimensional space; and
when it is determined that the object enters the detection region of the image sensor and at least one of the following judgments is determined yes:
(1) that the object includes a plurality of pixels having brightness exceeding a predetermined difference value over the background model brightness;
(2) that the object has a predetermined appearance feature; or
(3) that the object shows a movement, reducing a level of updating a background model, whereby an accuracy to distinguish a foreground image in a new image sensed by the image sensor is increased;
wherein the step of reducing the level of updating the background model includes: prolonging an update time of the background model.
2. The background model update method according to claim 1 , wherein the step of reducing the level of updating the background model further includes: reducing a weighting of the new image as a component for updating background model.