Method for displaying stereoscopic image

• US 8,120,648 B2
• Filed: 08/28/2007
• Issued: 02/21/2012
• Est. Priority Date: 12/29/2006
• Status: Active Grant

• Current Assignee: Quanta Computer
• Original Assignee: Quanta Computer
• Inventors: Liu, Yun-Cheng, Hsieh, Tung-Lin, Chiang, Ching-Chun
• Primary Examiner(s): Hamza, Faruk
• Application Number:
  US11/892,887
• Publication Number:
  US 20080158347A1
• Time in Patent Office: 1,638 Days
• Field of Search: 709/42, 709/51, 348/42, 348/51
• US Class Current: 348/51
• CPC Class Codes: H04N 13/122 Improving the 3D impression of stereoscopic images by modifying image signal contents, e.g. by filtering or adding monoscopic depth cues H04N13/128 takes precedence
• CPC Class Codes: H04N 2013/0081 Depth or disparity estimation from stereoscopic image signals

Abstract

The invention discloses a method for displaying a stereoscopic image, wherein the stereoscopic image includes a first-eye image and a second-eye image. The method is to firstly calculate a shift of the first-eye image with respect to the second-eye image. Then, the method moves the first-eye image for a length of a shift. Afterward, the method outputs the moved first-eye image and the second-eye image to display the stereoscopic image. Thereby, the parallax of the stereoscopic image is adjusted, so as to enhance the synthesizing effect of the stereoscopic image created.

Claims

1. A method for displaying a stereoscopic image, the stereoscopic image comprising a first-eye image and a second-eye image, the method comprising steps of:

(a) dividing the first-eye image into a plurality of first search frames, the first search frames comprising N effective search frames, N being a natural number;

(b) searching for a target frame in the second-eye image according to each of the effective search frames in the first-eye image;

(c) calculating a parallax according to the position of each of the effective search frames in the first-eye image and the position of the corresponding target frame in the second-eye image;

(d) calculating a shift of the first-eye image with respect to the second-eye image according to the parallaxes;

(e) moving the first-eye image for a length of the shift; and

(f) outputting the moved first-eye image and the second-eye image to display the stereoscopic image;

wherein before the step (a), the method further comprises performing a high-pass filtering to the first-eye image and the second-eye image;

the step (b) further comprises (b1) searching for the corresponding target frame in the second-eye image according to an ith effective search frame, i being an integer ranging from 1 to N;

the step (c) further comprises;

(c1) judging whether the target frame is effective, if YES, calculating a the corresponding parallax according to the position of the ith effective search frame in the first-eye image and the position of the corresponding target frame in the second-eye image; and

(c2) performing the steps (b1) and (c1) to each of the N effective search frames to obtain M parallaxes, M being an integer ranging from 1 to N; and

the step (d) further comprises (d1) calculating the shift according to the M parallaxes.

2. The method of claim 1, wherein the first-eye image is a right-eye image and the second-eye image is a left-eye image.

3. The method of claim 1, wherein the first-eye image is a left-eye image and the second-eye image is a right-eye image.

4. The method of claim 1, wherein each of the first search frames individually comprises a plurality of first pixels which each individually has a first gray scale, step (a) further comprises steps of:

(a1) summing up the first gray scales of each of the first search frames to obtain a first gray scale sum; and

(a2) judging whether the first gray scale sum of each of the first search frames is larger than a first threshold, if YES, marking the first search frame as the effective search frame.

5. The method of claim 4, wherein step (b1) further comprises steps of:

(b11) defining T second search frames in the second-eye image, each of the T second search frames comprising a plurality of second pixels which each has a second gray scale, T being a natural number;

(b12) respectively calculating a difference between each of the first gray scales of the ith effective search frame and each of the second gray scales of a jth second search frame, and summing up the differences to obtain a difference sum, j being an integer ranging from 1 to T;

(b13) performing step (b12) to each of the T second search frames to obtain T difference sums; and

(b14) selecting the second search frame corresponding to the minimum difference sum among the T difference sums as the target frame.

6. The method of claim 5, wherein step (c1) further comprises step of:

judging whether the difference sum corresponding to the target frame is smaller than a second threshold, if YES, marking the target frame as effective.

7. The method of claim 1, wherein step (c2) further comprises a step of normalizing the M parallaxes.

8. The method of claim 1, wherein step (d1) further comprises a step of calculating an average of the M parallaxes as the shift.

9. The method of claim 1, wherein step (d1) further comprises a step of selecting a median from the M parallaxes as the shift.

10. The method of claim 1, wherein step (d1) further comprises steps of:

(d11) generating a distribution curve according to the M parallaxes;

(d12) selecting a maximum parallax and a minimum parallax according to the distribution curve and a third threshold; and

(d13) calculating an average of the maximum parallax and the minimum parallax as the shift.