Method and apparatus for color-based object tracking in video sequences
0 Assignments
0 Petitions
Accused Products
Abstract
A method and apparatus for tracking a color-based object in video sequences are provided. According to the method, an initial object area in one frame of video sequences desired to be tracked is assigned, and an initial object effective window containing the initial object area is assigned. A frame following the frame containing the assigned initial object area is received as a newly input image, and an object search window containing the initial object area for tracking and the initial object effective window in the newly input image is assigned. Then, the model histogram of the initial object area corresponding to a predetermined bin resolution value and the input histogram of the image in the object search window are calculated. From the calculated object probability image, using a predetermined method, a new object area to which the initial object area moved is obtained in the next frame in which the initial object area of the frame desired to be tracked is given as a previous (tracked) object area. By doing so, the object in video sequences is tracked. Accordingly, using the continuously extracted video object region information, an object-based interactive-type additional information service function in movies, TV programs, and CFs can be implemented effectively.
-
Citations
21 Claims
-
1-10. -10. (canceled)
-
11. A method for optimizing the bin resolution of a color histogram for robustly tracking an object in video sequences, the method comprising:
-
(a) assigning an initial object area in one frame of the video sequences desired to be tracked and assigning an initial object effective window containing the initial object area;
(b) assigning an object search window containing the initial object effective window assigned in the step (a), in the frame containing the object area assigned in the step (a);
(c) calculating a model histogram of the initial object area corresponding to a predetermined bin resolution value and an input histogram of an image in the object search window;
(d) calculating an object probability image of the image in the object search window, by using the model histogram and the input histogram;
(e) detecting an object area, existing in the frame of the step (a) by using a predetermined method, from the object probability image calculated in the step (d);
(f) calculating an object detection performance index which represents how much the initial object area assigned in the step (a) and the new object area obtained in the step (e) coincide, and how well the initial object area is distinguished from an adjacent background area in the object probability image calculated in the step (d); and
(g) calculating the object detection performance index for each of all bin resolution values, by repeatedly performing the steps (c) through (e using all bin resolution values that are available for assigning, and determining a bin resolution value that provides the maximum object detection performance index, as an optimized bin resolution value. - View Dependent Claims (12)
-
-
13-17. -17. (canceled)
-
18. An apparatus for optimizing the bin resolution of a color histogram for robustly tracking an object in video sequences comprising:
-
an image input means through which the video sequences are input;
a histogram calculating means which calculates a histogram of an image in an area assigned in the video sequence frame, by using a predetermined bin resolution;
a means which assigns an initial object area desired to be tracked, in one frame of the video sequences input through the image input means, and sets an initial object effective window containing the initial object area;
an object search window assigning means which assigns an object search window containing the set initial object effective window in the frame containing the initial object area;
an object probability image calculating means which calculates an object probability image of an image in the object search window, by using a model histogram which is a histogram of the initial object area calculated by using the histogram calculating means, and an input histogram which is a histogram of an image in the object search window;
a means which again detects the object area in the first frame by binarizing the calculated object probability image;
a means which determines an object detection performance index which indicates how much the initial object area and the object area obtained in the means, which again detects the object area, coincide, and how well the initial object area is distinguished from an adjacent background area in the object probability image obtained in the object probability image calculating means; and
a means which calculates the object detection performance index for each of all bin resolution values that the object detection performance index determining means is capable of assigning, and, determines a bin resolution value which provides the maximum object detection performance index among the calculated indexes, as an optimized bin resolution.
-
-
19-20. -20. (canceled)
-
21. A computer readable medium having embodied thereon a computer program for optimizing the bin resolution of a color histogram for robustly tracking an object in video sequences, wherein the optimizing of the bin resolution of a color histogram comprises:
-
(a) assigning an initial object area in one frame of the video sequences desired to be tracked and assigning an initial object effective window containing the initial object area;
(b) assigning an object search window containing the initial object effective window assigned in the step (a), in the frame containing the object area assigned in the step (a);
(c) calculating a model histogram of the initial object area corresponding to a predetermined bin resolution value and an input histogram of an image in the object search window;
(d) calculating an object probability image of the image in the object search window, by using the model histogram and the input histogram;
(e) again detecting an object area in the frame of the step (a) by using a predetermined method, from the object probability image calculated in the step (d);
(f) calculating an object detection performance index which represents how much the initial object area assigned in the step (a) and the new object area obtained in the step (e) coincide, and how well the initial object area is distinguished from an adjacent background area in the object probability image calculated in the step (d); and
(g) calculating the object detection performance index for each of all bin resolution values, by repeatedly performing the steps (c) through (f) using all bin resolution values that are available for assigning, and determining a bin resolution value that provides the maximum object detection performance index, as an optimized bin resolution value.
-
Specification