Compressibility-aware media retargeting with structure preserving

• US 20110211770A1
• Filed: 03/01/2010
• Published: 09/01/2011
• Est. Priority Date: 03/01/2010
• Status: Active Grant

• Current Assignee: National Tsing Hua University
• Original Assignee: National Tsing Hua University
• Inventors: Lai, Shang-Hong, Wang, Shu-Fan
• Granted Patent: US 8,295,647 B2
• Time in Patent Office: Days
• US Class Current: 382/299
• CPC Class Codes: G06T 3/18 Image warping, e.g. rearranging pixels individually

Abstract

The present invention provides a method for retargeting an image, comprising determining a total block structure energy of an input image content. A compressibility rate of the input image content is determined based on the total block structure energy. An optimal scaling factor of the input image content is obtained. The input image content is warped by using a new coordinate matrices and uniformly scaling the input image content to a target image resolution.

Claims

1. A method for retargeting an image, comprising:

determining a total block structure energy of an input image content by a processing unit;

determining a compressibility rate of said input image content based on said total block structure energy by said processing unit;

determining an optimal scaling factor of said input image content by said processing unit; and

warping said input image content by using a new coordinate matrices and uniformly scaling said input image content by said processing to output a target image resolution.

2. The method of claim 1, further comprising a step of solving the least-squares solution of a sparse linear system by said processing unit before warping said input image content.

3. The method of claim 1, wherein said total block structure energy may be determined by an intensity gradient, block structure energy and a saliency map.

4. The method of claim 1, wherein said compressibility rate (r_x and r_y) of said input image content may be obtained by the following equations:

5. The method of claim 4, wherein said optimal scaling factor (S_x*, S_y*) is obtained by

S_x*=β

·

S_y

S_y*=β

+1−

ω

·

(r_x·

β

−

β

−

1−

r_y)/(1+ω

)(β

²+1)

wherein β

is equal to W_r·

H_o/W_o·

H_r, and (S_x, S_y) denotes a resizing factor.

6. A method for retargeting a video, comprising:

determining a total block structure energy and a compressibility rate of an input video content of each frame by a processing unit;

determining an optimal scaling factor of said input video content by said processing unit; and

warping said input video content by using a new coordinate matrices and uniformly scaling said input video content by said processing unit to output a target video resolution.

7. The method of claim 6, further comprising a step of solving the least-squares solution of a sparse linear system with constraints and smoothing said optimal scaling factors by said processing unit before warping said input video content.

8. The method of claim 7, wherein said constraints comprise,

X_x,y^t−

X_x,y^t+1=0

Y_x,y^t−

Y_x,y^t+1=0

wherein X and Y denote said new coordinate matrices, and x and y denote an original coordinate matrices.

9. The method of claim 7, further comprising a step of retaining said new coordinate matrices as the initial solution for the next frame by said processing unit before warping said input video content.

10. The method of claim 6, wherein said total block structure energy may be determined by an intensity gradient, block structure energy and a saliency map.

11. The method of claim 6, wherein said compressibility rate (r_x and r_y) of said input video content may be obtained by the following equations:

12. The method of claim 6, wherein said optimal scaling factor (S_x*, S_y*) is obtained by

S_x*=β

·

S_y

S_y*=β

+1−

ω

·

(r_x·

β

−

β

−

1−

r_y)/(1+ω

)(β

²+1)

wherein β

is equal to W_r·

H_o/W_o·

H_r, and (S_x, S_y) denotes a resizing factor.