CN102523467A - Simple method for converting 2D video into 3D - Google Patents

Abstract

The invention relates to a simple method for converting a 2D video into 3D, characterized by comprising the following steps of: (1) firstly dividing a 2D video source into two paths, one of which is an odd column of pixels and the other of which is an even column of pixels, and storing the pixels in a memory; (2) integrally moving the odd column of pixels and the even column of pixels in the first frame leftward for 5-10 units; (3) integrally moving the odd column of pixels and the even column of pixels in the second frame rightward for 5-10 units; and (4) forming a 3D effect with a separate left eye and a right eye by a pair of shutter type 3D spectacles, and sequentially repeating the step 2 to step 4. The invention provides a simple method for converting 2D video into a 3D image without adding any complex algorithm. The method provided by the invention has the advantages of generating a feeling that one frame inclines leftward while the other frame inclines rightward and a double image is generated if the two frames are combined together, and forming a 3D vision feint by the coordination of the left and right switches of the spectacles.

Description

A kind of 2D video changes the simple and easy method of 3D

Technical field

The present invention relates to a kind of 2D video changes the simple and easy method of 3D.

Background technology

Change in the Video processing of 3D at 2D, if want good effect, its algorithm is relatively very complicated, and depth map must have; Just so-called disparity map, degree of depth map generalization quality have determined whether the effect of 2D commentaries on classics 3D is outstanding, also just because of this; Processing about depth map also is ever-changing having his own strong points, but nothing more than several first gray scale map generalizations that it is noted that are a little louder arranged; At present nearly all algorithm that changes 3D about 2D all has gray-scale map to generate this, just original coloured image is become gray level image, and it two is two-dimentional Gauss's smothing filterings; Purpose is further denoising, obtains level and smooth depth map, has better kept edge effect.At last treated depth map and original image are synthesized and formed corresponding with it disparity map, thereby cooperate the 3D glasses to form 3D effect.

But have to unify here at 2, the one, the 2-d gaussian filters calculating parameter $G(u,v)=\frac{1}{2\mathrm{\&pi;}{\mathrm{\&sigma;}}^2}{e}^{-({\mathrm{<figure-callout\; id="2"\; label="u"\; filenames="HDA0000118347540000011.png"\; state="\{\{state\}\}">u</figure-callout>}}^2+v^2)/\left(2{\mathrm{\&sigma;}}^2\right)}$ U in the formula, v are pixel coordinate, and δ is the standard deviation of normal distribution; Determine filtering strength, specifically getting how many values, u here; How v confirms each have their own view and idea, can not be definite fully, and the 2nd, how depth map of confirming and former figure make up; How to merge to also is a uncertain factor together, and all there is different views in every family.How concrete effect can only go to observe with human eye is experienced the criterion that neither one is rigid.

Summary of the invention

The purpose of this invention is to provide a kind of 2D video changes the simple and easy method of 3D, can avoid the use of realization 2D video commentaries on classics 3D under the condition of complicated algorithm.

A kind of 2D video changes the simple and easy method of 3D, and its special feature is, comprises the steps:

(1) at first divide two-way with 2D video film source, strange row pixel one tunnel, even row pixel one road deposits in the memory simultaneously;

(2) strange row pixel and the even row pixel data integral body with first frame is moved to the left 5-10 unit;

(3) with the strange row pixel of second frame and even row pixel data integral body 5-10 the unit that move right;

(4) cooperate fast gate-type 3D glasses to form the 3D effect that right and left eyes separates, repeating step 2 gets final product to step 4 successively.

Wherein memory is the DDR memory.

Wherein be moved to the left 8 units in the step (2), 8 units move right in the step (3).

The invention provides and a kind ofly need not add any complicated algorithm, just can realize the 2D video is changeed the simple algorithm of 3D rendering.Adopt the inventive method to take back for people's the frame of feeling, a frame takes over, and lumps together to have ghost image, cooperates the left and right sides switch of glasses, thereby forms the 3D visual artifacts.

Description of drawings

Fig. 1 handles block diagram for 2D of the present invention changes the 3D video image.

Embodiment

Embodiment 1:

With full HD (resolution is 1920x1080) TV is example; First frame data that at first will be cached to DDR are read; Then preceding 8 row are total to the individual data of 1080x8 and throw away (not showing), the 9th pixel that deposits DDR in is presented at inside first pixel of screen, by that analogy; The 9th pixel reciprocal repeated to be presented at last 9 pixels the inside of screen; So be equivalent to whole image integral body and be moved to the left 8 pixels, preceding 8 row pixels are lost, and the value of back 8 row pixels is replenished with the pixel of the 9th row reciprocal.This is the processing mode to first two field picture.

Integral body 8 pixels that move right when from DDR, reading second frame; Be about to preceding 8 pixels the inside that first pixel repeats to be presented at screen; The 9th the locations of pixels back that second pixel is presented at screen shows normally that successively back reciprocal 8 row pixels are all thrown away, so the whole image integral body 8 row pixels that moved right; Preceding 8 row pixels repeat to show and replenish that back 8 row pixels are thrown away.Thereby take back for people's the frame of feeling, a frame takes over, and lumps together to have ghost image, cooperates the left and right sides switch of glasses, thereby forms the 3D visual artifacts.

If reaching 1920x1080, the screen physical resolution still can not use the method yet; Promptly can remove the individual pixel of (5～8) x (n); N represents the vertical resolution of screen, and pixel forms the video effect that 2D changes 3D thereby the N row are sacrificed in every utilization, all belongs to the category of this patent.

Claims (3)

1. the simple and easy method of a 2D video commentaries on classics 3D is characterized in that, comprises the steps:

(1) at first divide two-way with 2D video film source, strange row pixel one tunnel, even row pixel one road deposits in the memory simultaneously;

(2) strange row pixel and the even row pixel data integral body with first frame is moved to the left 5-10 unit;

(3) with the strange row pixel of second frame and even row pixel data integral body 5-10 the unit that move right;

(4) cooperate fast gate-type 3D glasses to form the 3D effect that right and left eyes separates, repeating step 2 gets final product to step 4 successively.

2. a kind of 2D video as claimed in claim 1 changes the simple and easy method of 3D, and it is characterized in that: wherein memory is the DDR memory.

3. according to claim 1 or claim 2 a kind of 2D video changes the simple and easy method of 3D, and it is characterized in that: wherein be moved to the left 8 units in the step (2), 8 units move right in the step (3).

Images