Motion object detection method using adaptive background model and computer-readable storage medium

A motion object detection method using an adaptive background model to detect a foreground object in a present image in accordance with at least one background image, wherein the motion object detection method comprises:

• establishing a background model in accordance with at least one background image, wherein the background model comprises a plurality of background brightness reference values;
• determining if a plurality of pixels of the present image are foreground pixels in accordance with the background model;
• detecting at least one foreground object in the present image in accordance with the pixels determined as foreground pixels;
• calculating an area of the at least one foreground object;
• determining if the area of the at least one foreground object is greater than a background interference threshold;
• performing a quick background updating step to update each of the background brightness reference values when the area of the at least one foreground object is greater than a background interference threshold, wherein the quick background updating step comprises:
• determining a brightness weight value in accordance with the brightness of the pixel of the present image;
• multiplying the background brightness reference value corresponding to the pixel by the brightness weight value to obtain a background brightness proportion value;
• multiplying a difference between 1 and the brightness weight value by the brightness of the pixel to obtain a foreground brightness proportion value;
• and
• updating the background brightness reference value corresponding to the pixel to the sum of the foreground brightness proportion value and the background brightness proportion value.

A motion object detection method using an adaptive background model to detect a foreground object in a present image in accordance with at least one background image, wherein the motion object detection method comprises:

• establishing a background model in accordance with at least one background image, wherein the background model comprises a plurality of background brightness reference values;
• determining if a plurality of pixels of the present image are foreground pixels in accordance with the background model;
• calculating a number of the pixels determined as foreground pixels;
• determining if the number of the pixels determined as foreground pixels is greater than a background varying threshold;
• performing a quick background updating step to update each of the background brightness reference values when the number of the pixels determined as foreground pixels is greater than a background varying threshold, wherein the quick background updating step comprises:
• determining a brightness weight value in accordance with the brightness of the pixel of the present image;
• multiplying the background brightness reference value corresponding to the pixel by the brightness weight value to obtain a background brightness proportion value;
• multiplying a difference between 1 and the brightness weight value by the brightness of the pixel to obtain a foreground brightness proportion value;
• and
• updating the background brightness reference value corresponding to the pixel to the sum of the foreground brightness proportion value and the background brightness proportion value.

A motion object detection, method using an adaptive background model to detect a foreground object in a present image in accordance with at least one background image, wherein the motion object detection method comprises:

• performing a background model establishing step to use the at least one background images to establish a background model, wherein the background model comprises a plurality of background brightness reference value;
• determining if a plurality;
• of pixels of the present image are foreground pixels;
• detecting at least one foreground pixies in the present image in accordance with the pixels determined as foreground pixels;
• and performing a shadow filtering step to filter off the shadow of the foreground object, wherein the shadow filtering step comprising:
• projecting the foreground object on a horizontal coordinate axis to obtain a pixels distribution curve wherein the pixels distribution curve represent a relation between horizontal positions of the pixels of the foreground object and pixel numbers corresponding to the horizontal positions and the horizontal coordinate axis represents the horizontal positions of the pixels of the foreground object;
• calculating a vertical distance between a maximum value point and the horizontal coordinate axis, wherein the maximum value point belongs to the pixels distribution curve and has a maximum pixel number value in the points of the pixels distribution curve;
• providing an object substance proportion threshold;
• determining a horizontal cut line in accordance with the object substance proportion threshold and the vertical distance wherein the horizontal cut line intercepts the pixel distribution curve at two interception points determining two vertical boundary lines in accordance with the two interception points;
• and determining a substance portion of the foreground object in accordance with the two vertical boundary lines.