Device and method for image processing
AssigneeSamsung Electronics Co., Ltd.
InventorsXun Sun, Rong Liu, Maolin Chen, Fan Zhang, Ji Yeun Kim, Lujin Gong, Zhilan Hu
Life of this patent
Estimating the term…
Abstract
A device and a method for image processing include an image processing device that may extract a foreground moving object from a depth map of a three-dimensional (3D) image that may include an image depth map acquirer to obtain the depth map of a successive 3D image over a period of time, a moving object segmenter to segment a moving object from the obtained depth map, and a moving object tracker to identify and track the segmented moving object.
Claims22
An image processing device to extract a foreground moving object from a depth map of a three-dimensional (3D) image, the device comprising:
- an image depth map acquirer to obtain the depth map of a successive 3D image,
- wherein the depth map expresses feature information of a target in the 3D image and comprises a two-dimensional (2D) pixel matrix sign of the 3D image;
- and
- a moving object segmenter to
- segment a moving object from the obtained depth map;
- construct a background model using the obtained depth map and determine an initial seed point of a foreground object;
- obtain a foreground object region by performing region growing from the initial seed point;
- obtain an initial region of the moving object by merging foreground object regions belonging to the same moving object among obtained foreground object regions;
- and
- obtain a complete region of the moving object by performing precise segmentation on adjacent initial regions of the moving object exhibiting a complex motion.
An image processing method of extracting a foreground moving object from a depth map of a three-dimensional (3D) image, the method comprising:
- obtaining the depth map of a successive 3D image,
- wherein the depth map expresses feature information of a target in the 3D image and comprises a two-dimensional (2D) pixel matrix sign of the 3D image;
- and
- segmenting a moving object from the obtained depth map;
- constructing a background model using the obtained depth map and determining an initial seed point of a foreground object;
- obtaining a foreground object region by performing region growing from the initial seed point;
- obtaining an initial region of the moving object by merging foreground object regions belonging to the same moving object among obtained foreground object regions;
- and
- obtaining a complete region of the moving object by performing precise segmentation on adjacent initial regions of the moving object exhibiting a complex motion.
Dependencies are inferred from the claim text.