CN105163111A - Method for evaluating visual comfort of three-dimensional image based on classification of scene modes - Google Patents

Abstract

The invention discloses a method for evaluating the visual comfort of a three-dimensional image based on classification of scene modes. The method comprises the following steps: determining ten kinds of scene modes of the three-dimensional image according to depth positions, where a foreground object and a background area in a natural scene are; then, determining the scene mode, which each three-dimensional image belongs to, in a three-dimensional image database, and establishing a visual comfort evaluation mode in each scene mode; and firstly determining the scene mode, which the three-dimensional image to be evaluated belongs to, then, selecting the visual comfort evaluation mode in the scene mode, calculating a visual comfort evaluation predication value of the three-dimensional image to be evaluated, and finally, correcting the visual comfort evaluation predication value of the three-dimensional image to be evaluated. The method disclosed by the invention has the advantages that: perception of a human visual system to the visual comfort of the three-dimensional image can be precisely reflected through the visual comfort evaluation modes respectively established based on ten kinds of scenes; and thus, the consistency between an objective evaluation result and human eye subjective perception can be effectively improved.

Description

A kind of stereo image vision comfort level evaluation method based on scene mode classification

Technical field

The present invention relates to a kind of stereo-picture Quality of experience evaluation method, especially relate to a kind of stereo image vision comfort level evaluation method based on scene mode classification.

Background technology

Along with the fast development of Stereoscopic Video Presentation technology and high-quality stereoscopic video content acquiring technology, visual experience quality (the QoE of three-dimensional video-frequency, qualityofexperience) be a major issue in three-dimensional video-frequency system, and visual comfort (VC, visualcomfort) is one of key factor of the visual experience quality affecting three-dimensional video-frequency.At present, the visual experience matter quantifier elimination of stereoscopic video/stereo-picture mainly concentrates on the impact of research contents distortion for visual experience quality, and this respect has created many achievements in research.The research of stereo image vision comfort level is the physiological discomfort caused under stereoscopic image content non-distortion prerequisite, the visual comfort objective evaluation model obtained for improve beholder visual experience quality, instruct the making of 3D content and post-processed to have a very important role.

Existing stereo image vision comfort level evaluation method is mainly by extracting the left and right visual point image of stereo-picture and the feature of anaglyph, such as parallax, gradient etc., the method then in Using statistics or the algorithm of machine learning set up objective models.But cause the various factors of discomfort on physiology not study yet thoroughly in the content of current stereoscopic image, cause the consistency between objective evaluation result and human eye subjective perception poor.For obtaining the good visual comfort evaluation method of evaluation effect, need to segment all kinds of natural scene, and on this basis the various characteristics of image of Accurate Analysis to the influence degree of visual comfort.

Summary of the invention

Technical problem to be solved by this invention is to provide a kind of stereo image vision comfort level evaluation method based on scene mode classification, and it can improve the consistency between objective evaluation result and human eye subjective perception effectively.

The present invention solves the problems of the technologies described above adopted technical scheme: a kind of stereo image vision comfort level evaluation method based on scene mode classification, is characterized in that comprising the following steps:

1. determine 10 kinds of scene modes of the natural scene shown by three-dimensional display, the 1st kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is convex be in that non-comfort zone, background region are recessed is in non-comfort zone in screen; 2nd kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is convex be in that non-comfort zone, background region are convex is in non-comfort zone in screen; 3rd kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is recessed be in that non-comfort zone, background region are recessed is in non-comfort zone in screen; 4th kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is convex be in that non-comfort zone, background region are recessed is in comfort zone in screen; 5th kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is convex be in that non-comfort zone, background region are convex is in comfort zone in screen; 6th kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is convex be in that comfort zone, background region are recessed is in non-comfort zone in screen; 7th kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is recessed be in that comfort zone, background region are recessed is in non-comfort zone in screen; 8th kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is convex be in that comfort zone, background region are recessed is in comfort zone in screen; 9th kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is convex be in that comfort zone, background region are convex is in comfort zone in screen; 10th kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is recessed be in that comfort zone, background region are recessed is in comfort zone in screen;

2. every width stereo-picture is selected to have a stereoscopic image data storehouse of the mean subjective scoring average of visual comfort; Then according to the parallactic angle of foreground target in the right viewpoint parallax gray level image of the every width stereo-picture in stereoscopic image data storehouse and the parallactic angle of background region, the scene mode belonging to every width stereo-picture in stereoscopic image data storehouse is determined; Again according to the parallactic angle of foreground target in the respective right viewpoint parallax gray level image of all stereo-pictures belonging to often kind of scene mode in stereoscopic image data storehouse and the parallactic angle of background region, set up the visual comfort evaluation model under often kind of scene mode, the visual comfort evaluation model under n-th kind of scene mode is described below: SMMO _n=(4.2028-u (Q _n))-v (Q _n) × D _a(Q _n)+0.1912 × ln (W _a)-0.0208 × D _a(Q _n) × ln (W _a), wherein, 1≤n≤10, SMMO _nrepresent the output of the visual comfort evaluation model under n-th kind of scene mode, u (Q _n) and v (Q _n) be constant, D _a(Q _n) represent that a width to be entered belongs to the global disparity angle of the right viewpoint parallax gray level image of the stereo-picture to be evaluated of n-th kind of scene mode, D _a(Q _n)=Q _n× | f _a|+(1-Q _n) × | b _a|, Q _nrepresent the weight under n-th kind of scene mode, f _arepresent that a width to be entered belongs to the parallactic angle of the foreground target in the right viewpoint parallax gray level image of the stereo-picture to be evaluated of n-th kind of scene mode, b _arepresent that a width to be entered belongs to the parallactic angle of the background region in the right viewpoint parallax gray level image of the stereo-picture to be evaluated of n-th kind of scene mode, W _arepresent that a width to be entered belongs to the width angle of the foreground target in the right viewpoint parallax gray level image of the stereo-picture to be evaluated of n-th kind of scene mode, symbol " || " is the symbol that takes absolute value;

3. the right viewpoint parallax gray level image of stereo-picture to be evaluated is designated as { d _r(x, y) }, wherein, 1≤x≤W, 1≤y≤H, W represents the width of stereo-picture to be evaluated, it is consistent with the width of the every width stereo-picture in stereoscopic image data storehouse, H represents the height of stereo-picture to be evaluated, and it is consistent with the height of the every width stereo-picture in stereoscopic image data storehouse, d _r(x, y) represents { d _r(x, y) } in coordinate position be the pixel value of the pixel of (x, y); Then adopt and step 2. in identical mode, determine the scene mode belonging to stereo-picture to be evaluated; Scene mode again belonging to stereo-picture to be evaluated, chooses the visual comfort evaluation model under this scene mode; Then according to the visual comfort evaluation model under this scene mode and { d _r(x, y) } in the parallactic angle, { d of foreground target _r(x, y) } in the parallactic angle of background region and { d _r(x, y) } in the width angle of foreground target, calculate the visual comfort evaluation and foreca value of stereo-picture to be evaluated, be designated as smmo, suppose that stereo-picture to be evaluated belongs to n-th kind of scene mode, then smmo=(4.2028-u (Q _n))-v (Q _n) × D _a' (Q _n)+0.1912 × ln (W _a')-0.0208 × D _a' (Q _n) × ln (W _a'), wherein, D _a' (Q _n) represent { d _r(x, y) } global disparity angle, D _a' (Q _n)=Q _n× | f _a' |+(1-Q _n) × | b _a' |, f _a' represent { d _r(x, y) } in the parallactic angle of foreground target, b _a' represent { d _r(x, y) } in the parallactic angle of background region, W _a' represent { d _r(x, y) } in the width angle of foreground target;

4. the visual comfort evaluation and foreca value smmo of stereo-picture to be evaluated is revised, the revised visual comfort evaluation and foreca value of stereo-picture to be evaluated is designated as smm, wherein, max () is for getting max function, and P represents the tortuosity attenuation coefficient of the visual comfort evaluation model under the scene mode belonging to stereo-picture to be evaluated, S _r' represent { d _r(x, y) } in the average prospect line hop count of foreground target, S _c' represent { d _r(x, y) } in the average prospect alignment hop count of foreground target, T _frepresent the parallactic angle threshold value of setting, T _rrepresent the prospect line hop count threshold value of setting, T _crepresent the prospect alignment hop count threshold value of setting.

Described step is middle Q 2. _nacquisition process be:

-1 2., total width number of supposing the stereo-picture belonging to n-th kind of scene mode in stereoscopic image data storehouse is M', wherein, and M' >=1;

2.-2, q is made _ninitial value be 0.1, make Q _ninitial value be 0;

2. the weight value-3, calculated under n-th kind of scene mode is q _ntime stereoscopic image data storehouse in belong to the global disparity angle of the right viewpoint parallax gray level image of every width stereo-picture of n-th kind of scene mode, be q by the weight value under n-th kind of scene mode _ntime stereoscopic image data storehouse in belong to the right viewpoint parallax gray level image of the m' width stereo-picture of n-th kind of scene mode global disparity angle be designated as D _{m', a}(q _n), D _{m', a}(q _n)=q _n× | f _{m', a}|+(1-q _n) × | b _{m', a}|, wherein, 1≤m'≤M', f _{m', a}and b _{m', a}belong to the parallactic angle of foreground target in the right viewpoint parallax gray level image of the m' width stereo-picture of n-th kind of scene mode and the parallactic angle of background region in corresponding expression stereoscopic image data storehouse, symbol " || " is the symbol that takes absolute value;

2. the weight value-4, calculated under n-th kind of scene mode is q _ntime stereoscopic image data storehouse in belong to n-th kind of scene mode every width stereo-picture eliminate the comfort level objective evaluation value of parallax linear effect, be q by the weight value under n-th kind of scene mode _ntime stereoscopic image data storehouse in belong to n-th kind of scene mode the m' width stereo-picture comfort level objective evaluation value that eliminates parallax linear effect be designated as VC _h(D _{m', a}(q _n), W _a), VC _h(D _{m', a}(q _n), W _a)=4.2028+0.1912 × ln (W _a)-0.0208 × D _{m', a}(q _n) × ln (W _a); Then the weight value calculated under n-th kind of scene mode is q _ntime stereoscopic image data storehouse in belong to the comfort level objective evaluation value of the parallax linear effect of every width stereo-picture of n-th kind of scene mode, be q by the weight value under n-th kind of scene mode _ntime stereoscopic image data storehouse in belong to the parallax linear effect of the m' width stereo-picture of n-th kind of scene mode comfort level objective evaluation value be designated as ERR _m', ERR _m'=VC _h(D _{m', a}(q _n), W _a)-MOS _m', wherein, MOS _m'represent that the weight value under n-th kind of scene mode is q _ntime stereoscopic image data storehouse in belong to the comfort level subjective assessment value of the m' width stereo-picture of n-th kind of scene mode; Adopting least square method again, is q to the weight value under n-th kind of scene mode _ntime stereoscopic image data storehouse in belong to n-th kind of scene mode the respective global disparity angle of right viewpoint parallax gray level image of all stereo-pictures and the comfort level objective evaluation value of respective parallax linear effect carry out linear fit, the weight value obtained under n-th kind of scene mode is q _ntime stereoscopic image data storehouse in belong to a fitting a straight line equation corresponding to all stereo-pictures of n-th kind of scene mode: ERR=u (q _n)+v (q _n) × D _a(q _n), wherein, ERR is for representing the dependent variable of parallax to the linear effect of objective comfort level, u (q _n) and v (q _n) be ERR=u (q _n)+v (q _n) × D _a(q _n) in constant, D _a(q _n) be for representing that the weight value under n-th kind of scene mode is q _ntime stereoscopic image data storehouse in belong to the independent variable at the global disparity angle of the right viewpoint parallax gray level image of arbitrary width stereo-picture of n-th kind of scene mode;

2. be-5, q according to the weight value under n-th kind of scene mode _ntime stereoscopic image data storehouse in belong to a fitting a straight line equation corresponding to all stereo-pictures of n-th kind of scene mode, the weight value obtained under n-th kind of scene mode is q _ntime stereoscopic image data storehouse in belong to the comfort level objective evaluation equation of all stereo-pictures of n-th kind of scene mode: SMMO _n(q _n)=VC _h(D _a(q _n), W _a)-ERR=(4.2028-u (q _n))-v (q _n) × D _a(q _n)+0.1912 × ln (W _a)-0.0208 × D _a(q _n) × ln (W _a), wherein, SMMO _n(q _n) for having superposed the comfort level objective evaluation value of parallax linear effect and parallax non-linear effects, VC _h(D _a(q _n), W _a) represent that the weight value under n-th kind of scene mode is q _ntime stereoscopic image data storehouse in belong to n-th kind of scene mode arbitrary width stereo-picture eliminate the comfort level objective evaluation value of parallax linear effect; The weight value of then measuring under n-th kind of scene mode is q _ntime stereoscopic image data storehouse in belong to the comfort level objective evaluation value of all stereo-pictures of n-th kind of scene mode and the fitting degree of corresponding comfort level subjective assessment value;

2.-6, q is made _n=q _n+ 0.1, then return step 2.-3 continuation execution, until q _nterminate when equaling 1.1, corresponding fitting degree when the weight value obtained under n-th kind of scene mode is respectively 0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8,0.9,1.0, then by weight value assignment corresponding for best fit degree to Q _n, wherein, q _n=q _n"=" in+0.1 is assignment;

Described step is middle u (Q 2. _n) value be Monomial coefficient in fitting a straight line equation corresponding to best fit degree; Described step is middle v (Q 2. _n) value be constant in fitting a straight line equation corresponding to best fit degree.

The obtain manner of the foreground target in the right viewpoint parallax gray level image of every width stereo-picture of described step 2. in neutral body image data base and background region and described step 3. in stereo-picture to be evaluated right viewpoint parallax gray level image in foreground target identical with the obtain manner of background region, using the every width stereo-picture in stereoscopic image data storehouse and stereo-picture to be evaluated all as pending stereo-picture, foreground target in the right viewpoint parallax gray level image of then pending stereo-picture and the acquisition process of background region are: the grey level histogram obtaining the right viewpoint parallax gray level image of pending stereo-picture, then maximum variance between clusters is adopted, and the grey level histogram of right viewpoint parallax gray level image according to pending stereo-picture, obtain the intensity slicing threshold value of the right viewpoint parallax gray level image of pending stereo-picture, be designated as P _seg, again pixel value in the right viewpoint parallax gray level image of pending stereo-picture is more than or equal to P _segpixel be defined as foreground pixel point, and pixel value in the right viewpoint parallax gray level image of pending stereo-picture is less than P _segpixel be defined as background pixel, the last foreground target be made up of all foreground pixel points in the right viewpoint parallax gray level image of pending stereo-picture, forms the background region in the right viewpoint parallax gray level image of pending stereo-picture by all background pixels.

The intensity slicing threshold value P of the right viewpoint parallax gray level image of described pending stereo-picture _segacquisition process be:

X1, set up intensity slicing threshold value find target function, be designated as T, T=w _f× (μ-μ _f) ²+ w _b× (μ-μ _b) ², wherein, w _frepresent that total number of the foreground pixel point in the right viewpoint parallax gray level image of pending stereo-picture accounts for the ratio of total number of the pixel in the right viewpoint parallax gray level image of pending stereo-picture, μ _frepresent the average of the pixel value of all foreground pixel points in the right viewpoint parallax gray level image of pending stereo-picture, w _brepresent that total number of the background pixel in the right viewpoint parallax gray level image of pending stereo-picture accounts for the ratio of total number of the pixel in the right viewpoint parallax gray level image of pending stereo-picture, μ _brepresent the average of the pixel value of all background pixels in the right viewpoint parallax gray level image of pending stereo-picture, μ represents the average of the pixel value of all pixels in the right viewpoint parallax gray level image of pending stereo-picture, μ=w _f× μ _f+ w _b× μ _b, Hist _dg () represents the grey level histogram { Hist of the right viewpoint parallax gray level image of pending stereo-picture _d(g) } in pixel value be the number of the pixel of g, 0≤g _min≤ g≤g _max≤ 255, g _minrepresent the grey level histogram { Hist of the right viewpoint parallax gray level image of pending stereo-picture _d(g) } in the number of the pixel minimum pixel value that is greater than 0, g _maxrepresent the grey level histogram { Hist of the right viewpoint parallax gray level image of pending stereo-picture _d(g) } in the number of the pixel max pixel value that is greater than 0;

X2, at interval [g _min, g _max] interior traversal t, the value of t when making T maximum is defined as the intensity slicing threshold value P of the right viewpoint parallax gray level image of pending stereo-picture _seg.

Described step 2. in determine to determine that the mode of the scene mode belonging to stereo-picture to be evaluated is identical during the mode of the scene mode belonging to every width stereo-picture in stereoscopic image data storehouse and described step are 3., using the every width stereo-picture in stereoscopic image data storehouse and stereo-picture to be evaluated all as pending stereo-picture, then determine that the detailed process of the scene mode belonging to pending stereo-picture is:

Z1, for the foreground target in the right viewpoint parallax gray level image of pending stereo-picture, if f _a ^*>1 °, then determine that foreground target is convex and be in non-comfort zone in screen; If 0 °≤f _a ^*<1 °, then to determine that foreground target is convex and be in comfort zone in screen; If-1 ° of <f _a ^*<0 °, then to determine that foreground target is recessed and be in comfort zone in screen; If f _a ^*<-1 °, then determine that foreground target is recessed and be in non-comfort zone in screen;

For the background region in the right viewpoint parallax gray level image of pending stereo-picture, if b _a ^*>1 °, then determine that background region is convex and be in non-comfort zone in screen; If 0 °≤b _a ^*<1 °, then to determine that background region is convex and be in comfort zone in screen; If-1 ° of < b _a ^*<0 °, then to determine that background region is recessed and be in comfort zone in screen; If b _a ^*<-1 °, then determine that background region is recessed and be in non-comfort zone in screen;

Wherein, f _a ^*represent the parallactic angle of the foreground target in the right viewpoint parallax gray level image of pending stereo-picture, f _a ^*=f-k, b _a ^*represent the parallactic angle of the background region in the right viewpoint parallax gray level image of pending stereo-picture, b _a ^*=b-k, f represent the convergent angle of human eye binocular viewing foreground target, b represents the convergent angle of human eye binocular viewing background region, k represents the adjustment angle of human eye binocular, arctan () is tan of negating, p represents the interpupillary distance of human eye binocular, L represents the width of display, N represents the horizontal resolution of display, h represents the distance of human eye to display, F represents the mean parallax amplitude of the foreground target in the right viewpoint parallax gray level image of pending stereo-picture, and F is in units of pixel $F=\frac{\underset{x=1}{\overset{W}{\&Sigma;}}\underset{y=1}{\overset{H}{\&Sigma;}}\mathrm{\&epsiv;}\left({d_R}^{*}\right(x,y)-P_{seg})\&times;{d_R}^{*}(x,y)}{\underset{x=1}{\overset{W}{\&Sigma;}}\underset{y=1}{\overset{H}{\&Sigma;}}\mathrm{\&epsiv;}\left({d_R}^{*}\right(x,y)-P_{seg})},\mathrm{\&epsiv;}\left({d_R}^{*}\right(x,y)-P_{seg})=\left\{\begin{array}{c}1,{d_R}^{*}(x,y)-P_{seg}\&GreaterEqual;0\\ 0,{d_R}^{*}(x,y)-P_{seg}<0\end{array}\right.,$ B represents the mean parallax amplitude of the background region in the right viewpoint parallax gray level image of pending stereo-picture, B in units of pixel, $B=\frac{\underset{x=1}{\overset{W}{\&Sigma;}}\underset{y=1}{\overset{H}{\&Sigma;}}\mathrm{\&epsiv;}(P_{seg}-{d_R}^{*}(x,y\left)\right)\&times;{d_R}^{*}(x,y)}{\underset{x=1}{\overset{W}{\&Sigma;}}\underset{y=1}{\overset{H}{\&Sigma;}}\mathrm{\&epsiv;}(P_{seg}-{d_R}^{*}(x,y\left)\right)},\mathrm{\&epsiv;}(P_{seg}-{d_R}^{*}(x,y\left)\right)=\left\{\begin{array}{c}1,P_{seg}-{d_R}^{*}(x,y)\&GreaterEqual;0\\ 0,P_{seg}-{d_R}^{*}(x,y)<0\end{array}\right.,$ D _r ^*(x, y) represents that in the right viewpoint parallax gray level image of pending stereo-picture, coordinate position is the pixel value of the pixel of (x, y);

Z2, according to the foreground target in the right viewpoint parallax gray level image of pending stereo-picture relative to the concavity and convexity of screen and whether be in comfort zone, background region relative to screen concavity and convexity and whether be in comfort zone, determine the scene mode belonging to pending stereo-picture, be specially: if foreground target is convex be in that non-comfort zone, background region are recessed is in non-comfort zone in screen in screen, then determine that pending stereo-picture belongs to the 1st kind of scene mode; If foreground target is convex be in that non-comfort zone, background region are convex is in non-comfort zone in screen in screen, then determine that pending stereo-picture belongs to the 2nd kind of scene mode; If foreground target is recessed be in that non-comfort zone, background region are recessed is in non-comfort zone in screen in screen, then determine that pending stereo-picture belongs to the 3rd kind of scene mode; If foreground target is convex be in that non-comfort zone, background region are recessed is in comfort zone in screen in screen, then determine that pending stereo-picture belongs to the 4th kind of scene mode; If foreground target is convex be in that non-comfort zone, background region are convex is in comfort zone in screen in screen, then determine that pending stereo-picture belongs to the 5th kind of scene mode; If foreground target is convex be in that comfort zone, background region are recessed is in non-comfort zone in screen in screen, then determine that pending stereo-picture belongs to the 6th kind of scene mode; If foreground target is recessed be in that comfort zone, background region are recessed is in non-comfort zone in screen in screen, then determine that pending stereo-picture belongs to the 7th kind of scene mode; If foreground target is convex be in that comfort zone, background region are recessed is in comfort zone in screen in screen, then determine that pending stereo-picture belongs to the 8th kind of scene mode; If foreground target is convex be in that comfort zone, background region are convex is in comfort zone in screen in screen, then determine that pending stereo-picture belongs to the 9th kind of scene mode; If foreground target is recessed be in that comfort zone, background region are recessed is in comfort zone in screen in screen, then determine that pending stereo-picture belongs to the 10th kind of scene mode.

Described step 3. in { d _r(x, y) } in the width angle of foreground target wherein, arctan () is tan of negating, and h represents the distance of human eye to display, W _frontrepresent { d _r(x, y) } in the mean breadth of foreground target, W _frontacquisition process be:

A1, to { d _r(x, y) } carry out binary conversion treatment, obtain { d _r(x, y) } binary image, be designated as { BI (x, y) }, the pixel value being the pixel of (x, y) by coordinate position in { BI (x, y) } is designated as BI (x, y), as { d _r(x, y) } in coordinate position be the pixel of (x, y) when belonging to foreground target, make BI (x, y)=1, as { d _r(x, y) } in coordinate position be the pixel of (x, y) when belonging to background region, make BI (x, y)=0;

A2, to { BI (x, y) } line scanning is carried out, for { BI (x, y) row } is capable, scan to the right from the 1st pixel that row is capable, when to scan the 1st pixel value be the pixel of 1, using the original position of this pixel as the 1st section of prospect width line segment, and the row coordinate of this pixel is designated as x ₁, continuing scanning to the right and obtain the 1st section of prospect width line segment until scanning till pixel value is the pixel of 0, is that the row coordinate of the pixel of 0 is designated as x by this pixel value ₂, the length of the 1st section of prospect width line segment is designated as WL ₁, WL ₁=x ₂-x ₁; Continue scan to the right, in the mode identical with acquisition the 1st section of prospect width line segment, obtain row capable in all prospect width line segments; Wherein, the initial value of row is 1,1≤row≤H, 1≤x ₁<x ₂≤ W;

All prospect width line segments that a3, all prospect width line segments length in { BI (x, y) } being less than 0.002W and length are greater than 0.995W are removed; Then sort to remaining all prospect width line segments by length order from small to large, the prospect width line segment getting middle 80% forms prospect width line segment aggregate, is designated as { WL' _n, wherein, WL' _nrepresent { WL' _nin the length of n-th section of prospect width line segment, 1≤n≤N', N' represents { WL' _nin total hop count of prospect width line segment of comprising;

A4, calculating { WL' _nin the mean value of length of all prospect width line segments, be designated as W _front', then by W _front' as { d _r(x, y) } in the mean breadth W of foreground target _fronteven, W _front=W _front'.

Described step 4. in { d _r(x, y) } in the average prospect line hop count of foreground target { d _r(x, y) } in the average prospect alignment hop count of foreground target wherein, 1≤row≤H, 1≤col≤W, RLQ _rowrepresent { d _r(x, y) } in row capable in the quantity of prospect line section, $\mathrm{\&epsiv;}\left({\mathrm{RLQ}}_{row}\right)=\{\begin{array}{c}1,{\mathrm{RLQ}}_{row}>0\\ 0,{\mathrm{RLQ}}_{row}\&le;0\end{array},$ CLQ _colrepresent { d _r(x, y) } in col row in the quantity of prospect alignment section, $\mathrm{\&epsiv;}\left({\mathrm{CLQ}}_{col}\right)=\left\{\begin{array}{c}1,CL{Q}_{col}>0\\ 0,CL{Q}_{col}\&le;0\end{array}\right.;$

Wherein, RLQ _rowand CLQ _colacquisition process be:

B1, to { d _r(x, y) } carry out binary conversion treatment, obtain { d _r(x, y) } binary image, be designated as { BI (x, y) }, the pixel value being the pixel of (x, y) by coordinate position in { BI (x, y) } is designated as BI (x, y), as { d _r(x, y) } in coordinate position be the pixel of (x, y) when belonging to foreground target, make BI (x, y)=1, as { d _r(x, y) } in coordinate position be the pixel of (x, y) when belonging to background region, make BI (x, y)=0;

B2, to { BI (x, y) 2 dilation operations of mathematical morphology, 4 erosion operations, 2 dilation operations } are carried out successively, obtain { BI'(x, y) }, wherein, BI'(x, y) represent { BI'(x, y) }, coordinate position is the pixel value of the pixel of (x, y);

The quantity of effective prospect line section in often row in b3, statistics { BI'(x, y) }, by capable for the row in { BI'(x, y) } effectively the quantity of prospect line section be designated as RLQ _row', RLQ _row' acquisition process be: b3-1, make RLQ _row' initial value be 0; B3-2, from { BI'(x, y) the 1st pixel that the row } is capable starts to scan to the right, when to scan the 1st pixel value be the pixel of 1, using the original position of this pixel as the 1st section of prospect line section, and the row coordinate of this pixel is designated as x ₁', continuing scanning to the right and obtain the 1st section of prospect line section until scanning till pixel value is the pixel of 0, is that the row coordinate of the pixel of 0 is designated as x by this pixel value ₂', if the length x of the 1st section of prospect line section ₂'-x ₁' be greater than 0.005W, then determine that the 1st section of prospect line section is effective prospect line section, and make RLQ _row'=RLQ _row'+1, if the length x of the 1st section of prospect line section ₂'-x ₁' be less than or equal to 0.005W, then determine that the 1st section of prospect line section is invalid prospect line section; Continue to scan to the right, to determine whether the 1st section of prospect line section is the mode that effective prospect line section is identical with acquisition the 1st section of prospect line section, obtain row capable in all effective prospect line section, and statistics obtain row capable in the amount R LQ of effective prospect line section _row'; Wherein, the initial value of row is 1,1≤row≤H, 1≤x ₁' <x ₂'≤W, RLQ _row'=RLQ _row"=" in '+1 is assignment;

Equally, the quantity of effective prospect alignment section in the often row in statistics { BI'(x, y) }, by capable for the col in { BI'(x, y) } effectively the quantity of prospect alignment section be designated as CLQ _col', CLQ _col' acquisition process be: b3-1), make CLQ _col' initial value be 0; B3-2), from { BI'(x, 1st pixel of the col row y) } starts downward scanning, when to scan the 1st pixel value be the pixel of 1, using the original position of this pixel as the 1st section of prospect alignment section, and the row-coordinate of this pixel is designated as y ₁', continuing scanning downwards and obtain the 1st section of prospect alignment section until scanning till pixel value is the pixel of 0, is that the row-coordinate of the pixel of 0 is designated as y by this pixel value ₂', if the length y of the 1st section of prospect alignment section ₂'-y ₁' be greater than 0.005H, then determine that the 1st section of prospect alignment section is effective prospect alignment section, and make CLQ _col'=CLQ _col'+1, if the length y of the 1st section of prostatitis alignment section ₂'-y ₁' be less than or equal to 0.005H, then determine that the 1st section of prospect alignment section is invalid prospect alignment section; Continue scanning downwards, to determine whether the 1st section of prospect alignment section is the mode that effective prospect alignment section is identical with acquisition the 1st section of prospect alignment section, obtain col capable in all effective prospect alignment section, and statistics obtains the quantity CLQ of effective prospect alignment section in col row _col'; Wherein, the initial value of col is 1,1≤col≤W, 1≤y ₁' <y ₂'≤H, CLQ _col'=CLQ _col"=" in '+1 is assignment;

B4, make RLQ _row=RLQ _row', make CLQ _col=CLQ _col'.

Described step 4. in get P=1.6, T _f=2.0, T _r=2, T _c=1.5.

Compared with prior art, the invention has the advantages that: foreground target and the depth location residing for background region in the natural scene of the inventive method according to three-dimensional display, determine 10 kinds of scene modes of stereo-picture, this can fully demonstrate the degree of depth factor of object in stereo-picture and the non-linear relation of visual comfort, thus the visual comfort evaluation model set up respectively based on 10 kinds of scene modes can reflect the perception of the visual comfort of human visual system's stereoscopic image more accurately, thus the consistency that can effectively improve between objective evaluation result and human eye subjective perception.

Accompanying drawing explanation

Fig. 1 be the inventive method totally realize block diagram;

Fig. 2 is the geometric representation of f, b, k, p, h, F, B;

Fig. 3 is W ' _ageometric representation;

Fig. 4 a is the weight value under the 4th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 4th kind of scene mode when being 0.1 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 4 b is the weight value under the 4th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 4th kind of scene mode when being 0.2 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 4 c is the weight value under the 4th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 4th kind of scene mode when being 0.3 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 4 d is the weight value under the 4th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 4th kind of scene mode when being 0.4 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 4 e is the weight value under the 4th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 4th kind of scene mode when being 0.5 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 4 f is the weight value under the 4th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 4th kind of scene mode when being 0.6 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 4 g is the weight value under the 4th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 4th kind of scene mode when being 0.7 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 4 h is the weight value under the 4th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 4th kind of scene mode when being 0.8 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 4 i is the weight value under the 4th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 4th kind of scene mode when being 0.9 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 4 j is the weight value under the 4th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 4th kind of scene mode when being 1.0 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 5 is that the weight value under the 4th kind of scene mode belongs to fitting degree corresponding to all stereo-pictures of the 4th kind of scene mode when being respectively 0.1 ~ 1.0 (fitting degree is weighed with MAE in stereoscopic image data storehouse, when weight value is 0.7, MAE is minimum, and the weight value that namely best fit degree is corresponding is 0.7);

Fig. 6 a is the weight value under the 5th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 5th kind of scene mode when being 0.1 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 6 b is the weight value under the 5th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 5th kind of scene mode when being 0.2 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 6 c is the weight value under the 5th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 5th kind of scene mode when being 0.3 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 6 d is the weight value under the 5th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 5th kind of scene mode when being 0.4 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 6 e is the weight value under the 5th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 5th kind of scene mode when being 0.5 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 6 f is the weight value under the 5th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 5th kind of scene mode when being 0.6 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 6 g is the weight value under the 5th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 5th kind of scene mode when being 0.7 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 6 h is the weight value under the 5th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 5th kind of scene mode when being 0.8 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 6 i is the weight value under the 5th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 5th kind of scene mode when being 0.9 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 6 j is the weight value under the 5th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 5th kind of scene mode when being 1.0 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 7 is that the weight value under the 5th kind of scene mode belongs to fitting degree corresponding to all stereo-pictures of the 5th kind of scene mode when being respectively 0.1 ~ 1.0 (fitting degree is weighed with MAE in stereoscopic image data storehouse, when weight value is 0.6, MAE is minimum, and the weight value that namely best fit degree is corresponding is 0.6);

Fig. 8 a is the weight value under the 8th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 8th kind of scene mode when being 0.1 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 8 b is the weight value under the 8th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 8th kind of scene mode when being 0.2 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 8 c is the weight value under the 8th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 8th kind of scene mode when being 0.3 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 8 d is the weight value under the 8th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 8th kind of scene mode when being 0.4 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 8 e is the weight value under the 8th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 8th kind of scene mode when being 0.5 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 8 f is the weight value under the 8th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 8th kind of scene mode when being 0.6 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 8 g is the weight value under the 8th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 8th kind of scene mode when being 0.7 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 8 h is the weight value under the 8th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 8th kind of scene mode when being 0.8 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 8 i is the weight value under the 8th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 8th kind of scene mode when being 0.9 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 8 j is the weight value under the 8th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 8th kind of scene mode when being 1.0 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Fig. 9 is that the weight value under the 8th kind of scene mode belongs to fitting degree corresponding to all stereo-pictures of the 8th kind of scene mode when being respectively 0.1 ~ 1.0 (fitting degree is weighed with MAE in stereoscopic image data storehouse, when weight value is 0.6, MAE is minimum, and the weight value that namely best fit degree is corresponding is 0.6);

Figure 10 a is the weight value under the 9th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 9th kind of scene mode when being 0.1 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Figure 10 b is the weight value under the 9th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 9th kind of scene mode when being 0.2 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Figure 10 c is the weight value under the 9th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 9th kind of scene mode when being 0.3 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Figure 10 d is the weight value under the 9th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 9th kind of scene mode when being 0.4 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Figure 10 e is the weight value under the 9th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 9th kind of scene mode when being 0.5 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Figure 10 f is the weight value under the 9th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 9th kind of scene mode when being 0.6 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Figure 10 g is the weight value under the 9th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 9th kind of scene mode when being 0.7 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Figure 10 h is the weight value under the 9th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 9th kind of scene mode when being 0.8 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Figure 10 i is the weight value under the 9th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 9th kind of scene mode when being 0.9 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Figure 10 j is the weight value under the 9th kind of scene mode belongs to Linear Fit Chart corresponding to all stereo-pictures of the 9th kind of scene mode when being 1.0 in stereoscopic image data storehouse, abscissa is global disparity angle, and ordinate is the objective comfortable angle value of parallax linear effect;

Figure 11 is that the weight value under the 9th kind of scene mode belongs to fitting degree corresponding to all stereo-pictures of the 9th kind of scene mode when being respectively 0.1 ~ 1.0 (fitting degree is weighed with MAE in stereoscopic image data storehouse, when weight value is 0.5, MAE is minimum, and the weight value that namely best fit degree is corresponding is 0.5).

Embodiment

Below in conjunction with accompanying drawing embodiment, the present invention is described in further detail.

A kind of stereo image vision comfort level evaluation method based on scene mode classification that the present invention proposes, it totally realizes block diagram as shown in Figure 1, and it comprises the following steps:

1. determine 10 kinds of scene modes of the natural scene shown by three-dimensional display, the 1st kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is convex be in that non-comfort zone, background region are recessed is in non-comfort zone in screen; 2nd kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is convex be in that non-comfort zone, background region are convex is in non-comfort zone in screen; 3rd kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is recessed be in that non-comfort zone, background region are recessed is in non-comfort zone in screen; 4th kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is convex be in that non-comfort zone, background region are recessed is in comfort zone in screen; 5th kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is convex be in that non-comfort zone, background region are convex is in comfort zone in screen; 6th kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is convex be in that comfort zone, background region are recessed is in non-comfort zone in screen; 7th kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is recessed be in that comfort zone, background region are recessed is in non-comfort zone in screen; 8th kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is convex be in that comfort zone, background region are recessed is in comfort zone in screen; 9th kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is convex be in that comfort zone, background region are convex is in comfort zone in screen; 10th kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is recessed be in that comfort zone, background region are recessed is in comfort zone in screen.

2. every width stereo-picture is selected to have a stereoscopic image data storehouse of the mean subjective scoring average of visual comfort; Then according to the parallactic angle of foreground target in the right viewpoint parallax gray level image of the every width stereo-picture in stereoscopic image data storehouse and the parallactic angle of background region, the scene mode belonging to every width stereo-picture in stereoscopic image data storehouse is determined; Again according to the parallactic angle of foreground target in the respective right viewpoint parallax gray level image of all stereo-pictures belonging to often kind of scene mode in stereoscopic image data storehouse and the parallactic angle of background region, set up the visual comfort evaluation model under often kind of scene mode, the visual comfort evaluation model under n-th kind of scene mode is described below: SMMO _n=(4.2028-u (Q _n))-v (Q _n) × D _a(Q _n)+0.1912 × ln (W _a)-0.0208 × D _a(Q _n) × ln (W _a), wherein, 1≤n≤10, SMMO _nrepresent the output of the visual comfort evaluation model under n-th kind of scene mode, u (Q _n) and v (Q _n) be constant, u (Q _n) and v (Q _n) value by the u (q in step 2.-4 _n) and v (q _n) determine, D _a(Q _n) represent that a width to be entered belongs to the global disparity angle of the right viewpoint parallax gray level image of the stereo-picture to be evaluated of n-th kind of scene mode, D _a(Q _n)=Q _n× | f _a|+(1-Q _n) × | b _a|, Q _nrepresent the weight under n-th kind of scene mode, f _arepresent that a width to be entered belongs to the parallactic angle of the foreground target in the right viewpoint parallax gray level image of the stereo-picture to be evaluated of n-th kind of scene mode, b _arepresent that a width to be entered belongs to the parallactic angle of the background region in the right viewpoint parallax gray level image of the stereo-picture to be evaluated of n-th kind of scene mode, W _arepresent that a width to be entered belongs to the width angle of the foreground target in the right viewpoint parallax gray level image of the stereo-picture to be evaluated of n-th kind of scene mode, symbol " || " is the symbol that takes absolute value.

In this particular embodiment, step 2. middle Q _nacquisition process be:

-1 2., total width number of supposing the stereo-picture belonging to n-th kind of scene mode in stereoscopic image data storehouse is M', wherein, and M' >=1.

2.-2, q is made _ninitial value be 0.1, make Q _ninitial value be 0.

2. the weight value-3, calculated under n-th kind of scene mode is q _ntime stereoscopic image data storehouse in belong to the global disparity angle of the right viewpoint parallax gray level image of every width stereo-picture of n-th kind of scene mode, be q by the weight value under n-th kind of scene mode _ntime stereoscopic image data storehouse in belong to the right viewpoint parallax gray level image of the m' width stereo-picture of n-th kind of scene mode global disparity angle be designated as D _{m', a}(q _n), D _{m', a}(q _n)=q _n× | f _{m', a}|+(1-q _n) × | b _{m', a}|, wherein, 1≤m'≤M', f _{m', a}and b _{m', a}belong to the parallactic angle of foreground target in the right viewpoint parallax gray level image of the m' width stereo-picture of n-th kind of scene mode and the parallactic angle of background region in corresponding expression stereoscopic image data storehouse, symbol " || " is the symbol that takes absolute value.

2. the weight value-4, calculated under n-th kind of scene mode is q _ntime stereoscopic image data storehouse in belong to n-th kind of scene mode every width stereo-picture eliminate the comfort level objective evaluation value of parallax linear effect, be q by the weight value under n-th kind of scene mode _ntime stereoscopic image data storehouse in belong to n-th kind of scene mode the m' width stereo-picture comfort level objective evaluation value that eliminates parallax linear effect be designated as VC _h(D _{m', a}(q _n), W _a), VC _h(D _{m', a}(q _n), W _a)=4.2028+0.1912 × ln (W _a)-0.0208 × D _{m', a}(q _n) × ln (W _a); Then the weight value calculated under n-th kind of scene mode is q _ntime stereoscopic image data storehouse in belong to the comfort level objective evaluation value of the parallax linear effect of every width stereo-picture of n-th kind of scene mode, be q by the weight value under n-th kind of scene mode _ntime stereoscopic image data storehouse in belong to the parallax linear effect of the m' width stereo-picture of n-th kind of scene mode comfort level objective evaluation value be designated as ERR _m', ERR _m'=VC _h(D _{m', a}(q _n), W _a)-MOS _m', wherein, MOS _m'represent that the weight value under n-th kind of scene mode is q _ntime stereoscopic image data storehouse in belong to the comfort level subjective assessment value of the m' width stereo-picture of n-th kind of scene mode, MOS _m'value known; Adopting existing least square method again, is q to the weight value under n-th kind of scene mode _ntime stereoscopic image data storehouse in belong to n-th kind of scene mode the respective global disparity angle of right viewpoint parallax gray level image of all stereo-pictures and the comfort level objective evaluation value of respective parallax linear effect carry out linear fit, the weight value obtained under n-th kind of scene mode is q _ntime stereoscopic image data storehouse in belong to a fitting a straight line equation corresponding to all stereo-pictures of n-th kind of scene mode: ERR=u (q _n)+v (q _n) × D _a(q _n), wherein, ERR is for representing the dependent variable of parallax to the linear effect of objective comfort level, u (q _n) and v (q _n) be ERR=u (q _n)+v (q _n) × D _a(q _n) middle constant, u (q _n) and v (q _n) value be matching result out, D _a(q _n) be for representing that the weight value under n-th kind of scene mode is q _ntime stereoscopic image data storehouse in belong to the independent variable at the global disparity angle of the right viewpoint parallax gray level image of arbitrary width stereo-picture of n-th kind of scene mode.

2. be-5, q according to the weight value under n-th kind of scene mode _ntime stereoscopic image data storehouse in belong to a fitting a straight line equation corresponding to all stereo-pictures of n-th kind of scene mode, the weight value obtained under n-th kind of scene mode is q _ntime stereoscopic image data storehouse in belong to the comfort level objective evaluation equation of all stereo-pictures of n-th kind of scene mode: SMMO _n(q _n)=VC _h(D _a(q _n), W _a)-ERR=(4.2028-u (q _n))-v (q _n) × D _a(q _n)+0.1912 × ln (W _a)-0.0208 × D _a(q _n) × ln (W _a), wherein, SMMO _n(q _n) for having superposed the comfort level objective evaluation value of parallax linear effect and parallax non-linear effects, VC _h(D _a(q _n), W _a) represent that the weight value under n-th kind of scene mode is q _ntime stereoscopic image data storehouse in belong to n-th kind of scene mode arbitrary width stereo-picture eliminate the comfort level objective evaluation value of parallax linear effect; Then the weight value adopting existing average absolute value error (MAE) to measure under n-th kind of scene mode is q _ntime stereoscopic image data storehouse in belong to the comfort level objective evaluation value of all stereo-pictures of n-th kind of scene mode and the fitting degree of corresponding comfort level subjective assessment value.

2.-6, q is made _n=q _n+ 0.1, then return step 2.-3 continuation execution, until q _nterminate when equaling 1.1, corresponding fitting degree when the weight value obtained under n-th kind of scene mode is respectively 0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8,0.9,1.0, then by weight value assignment corresponding for best fit degree to Q _n, wherein, q _n=q _n"=" in+0.1 is assignment.

In this particular embodiment, step 2. middle u (Q _n) value be Monomial coefficient in fitting a straight line equation corresponding to best fit degree, v (Q _n) value be constant in fitting a straight line equation corresponding to best fit degree.

At this, for the stereoscopic image data storehouse that Korea Advanced Institute of Science and Technology image and video system laboratory (IVYLAB) provide, this stereoscopic image data storehouse comprises the right viewpoint parallax gray level image of 120 width stereo-pictures and correspondence, this stereoscopic image data storehouse contains the indoor and outdoors stereo-picture of various scene depth, and gives the mean subjective scoring average of the visual comfort of every width stereo-picture, calculate the parallactic angle of foreground target in the right viewpoint parallax gray level image of the every width stereo-picture in this stereoscopic image data storehouse and the parallactic angle of background region, obtain foreground target and background region separately relative to the concavity and convexity of screen, and whether be positioned at the characteristic in euphorosia district, can find that the stereo-picture in this stereoscopic image data storehouse is only limitted to the 4th kind, 5th kind, 8th kind and the 9th kind of scene mode, therefore with the 4th kind, 5th kind, 8th kind and the 9th kind of scene mode are that example sets up the 4th kind, 5th kind, visual comfort evaluation model under 8th kind and the 9th kind of scene mode.

The Linear Fit Chart that all stereo-pictures of the 4th kind of scene mode are corresponding is belonged in stereoscopic image data storehouse when Fig. 4 a to Fig. 4 j correspondence weight value given under the 4th kind of scene mode is respectively 0.1 ~ 1.0, (fitting degree is weighed with MAE to belong to fitting degree corresponding to all stereo-pictures of the 4th kind of scene mode when Fig. 5 weight value given under the 4th kind of scene mode is respectively 0.1 ~ 1.0 in stereoscopic image data storehouse, when weight value is 0.7, MAE is minimum, and the weight value that namely best fit degree is corresponding is 0.7); The Linear Fit Chart that all stereo-pictures of the 5th kind of scene mode are corresponding is belonged in stereoscopic image data storehouse when Fig. 6 a to Fig. 6 j correspondence weight value given under the 5th kind of scene mode is respectively 0.1 ~ 1.0, (fitting degree is weighed with MAE to belong to fitting degree corresponding to all stereo-pictures of the 5th kind of scene mode when Fig. 7 weight value given under the 5th kind of scene mode is respectively 0.1 ~ 1.0 in stereoscopic image data storehouse, when weight value is 0.6, MAE is minimum, and the weight value that namely best fit degree is corresponding is 0.6); The Linear Fit Chart that all stereo-pictures of the 8th kind of scene mode are corresponding is belonged in stereoscopic image data storehouse when Fig. 8 a to Fig. 8 j correspondence weight value given under the 8th kind of scene mode is respectively 0.1 ~ 1.0, (fitting degree is weighed with MAE to belong to fitting degree corresponding to all stereo-pictures of the 8th kind of scene mode when Fig. 9 weight value given under the 8th kind of scene mode is respectively 0.1 ~ 1.0 in stereoscopic image data storehouse, when weight value is 0.6, MAE is minimum, and the weight value that namely best fit degree is corresponding is 0.6); The Linear Fit Chart that all stereo-pictures of the 9th kind of scene mode are corresponding is belonged in stereoscopic image data storehouse when Figure 10 a to Figure 10 j correspondence weight value given under the 9th kind of scene mode is respectively 0.1 ~ 1.0, (fitting degree is weighed with MAE to belong to fitting degree corresponding to all stereo-pictures of the 9th kind of scene mode when Figure 11 weight value given under the 9th kind of scene mode is respectively 0.1 ~ 1.0 in stereoscopic image data storehouse, when weight value is 0.5, MAE is minimum, and the weight value that namely best fit degree is corresponding is 0.5).

Be respectively according to the above weight calculated under the 4th kind, the 5th kind, the 8th kind and the 9th kind of scene mode: Q ₄=0.7, Q ₅=0.6, Q ₈=0.6, Q ₉visual comfort evaluation model under=0.5,4th kind, the 5th kind, the 8th kind and the 9th kind of scene mode is respectively:

SMMO ₄＝4.3938-0.6652×D _a(Q ₄)+0.1912×ln(W _a)-0.0208×D _a(Q ₄)×ln(W _a)

SMMO ₅＝4.2326-0.7120×D _a(Q ₅)+0.1912×ln(W _a)-0.0208×D _a(Q ₅)×ln(W _a)

。

SMMO ₈＝4.5232-0.8918×D _a(Q ₈)+0.1912×ln(W _a)-0.0208×D _a(Q ₈)×ln(W _a)

SMMO ₉＝5.4614-2.6905×D _a(Q ₉)+0.1912×ln(W _a)-0.0208×D _a(Q ₉)×ln(W _a)

3. the left viewpoint coloured image of stereo-picture to be evaluated is designated as { I _l(x, y) }, the right viewpoint coloured image of stereo-picture to be evaluated is designated as { I _r(x, y) }, the right viewpoint parallax gray level image of stereo-picture to be evaluated is designated as { d _r(x, y) }, wherein, 1≤x≤W, 1≤y≤H, W represents the width of stereo-picture to be evaluated, it is consistent with the width of the every width stereo-picture in stereoscopic image data storehouse, H represents the height of stereo-picture to be evaluated, and it is consistent with the height of the every width stereo-picture in stereoscopic image data storehouse, I _l(x, y) represents { I _l(x, y) } in coordinate position be the pixel value of the pixel of (x, y), I _r(x, y) represents { I _r(x, y) } in coordinate position be the pixel value of the pixel of (x, y), d _r(x, y) represents { d _r(x, y) } in coordinate position be the pixel value of the pixel of (x, y); Then adopt and step 2. in identical mode, determine the scene mode belonging to stereo-picture to be evaluated; Scene mode again belonging to stereo-picture to be evaluated, chooses the visual comfort evaluation model under this scene mode; Then according to the visual comfort evaluation model under this scene mode and { d _r(x, y) } in the parallactic angle, { d of foreground target _r(x, y) } in the parallactic angle of background region and { d _r(x, y) } in the width angle of foreground target, calculate the visual comfort evaluation and foreca value of stereo-picture to be evaluated, be designated as smmo, suppose that stereo-picture to be evaluated belongs to n-th kind of scene mode, then smmo=(4.2028-u (Q _n))-v (Q _n) × D _a' (Q _n)+0.1912 × ln (W _a')-0.0208 × D _a' (Q _n) × ln (W _a'), wherein, D _a' (Q _n) represent { d _r(x, y) } global disparity angle, D _a' (Q _n)=Q _n× | f _a' |+(1-Q _n) × | b _a' |, f _a' represent { d _r(x, y) } in the parallactic angle of foreground target, b _a' represent { d _r(x, y) } in the parallactic angle of background region, W _a' represent { d _r(x, y) } in the width angle of foreground target.

In this particular embodiment, step 3. in { d _r(x, y) } in the width angle of foreground target wherein, arctan () is tan of negating, and h represents the distance of human eye to display, W _frontrepresent { d _r(x, y) } in the mean breadth of foreground target, W _frontacquisition process be:

A1, to { d _r(x, y) } carry out binary conversion treatment, obtain { d _r(x, y) } binary image, be designated as { BI (x, y) }, the pixel value being the pixel of (x, y) by coordinate position in { BI (x, y) } is designated as BI (x, y), as { d _r(x, y) } in coordinate position be the pixel of (x, y) when belonging to foreground target, make BI (x, y)=1, as { d _r(x, y) } in coordinate position be the pixel of (x, y) when belonging to background region, make BI (x, y)=0.

A2, to { BI (x, y) } line scanning is carried out, for { BI (x, y) row } is capable, scan to the right from the 1st pixel that row is capable, when to scan the 1st pixel value be the pixel of 1, using the original position of this pixel as the 1st section of prospect width line segment, and the row coordinate of this pixel is designated as x ₁, continuing scanning to the right and obtain the 1st section of prospect width line segment until scanning till pixel value is the pixel of 0, is that the row coordinate of the pixel of 0 is designated as x by this pixel value ₂, the length of the 1st section of prospect width line segment is designated as WL ₁, WL ₁=x ₂-x ₁; Continue scan to the right, in the mode identical with acquisition the 1st section of prospect width line segment, obtain row capable in all prospect width line segments; Wherein, the initial value of row is 1,1≤row≤H, 1≤x ₁<x ₂≤ W.

A3, remove { BI (x, y) the prospect width line segment of length exception }, namely prospect width line segment short especially and prospect width line segment long is especially removed, be about to length in { BI (x, y) } be less than all prospect width line segments that all prospect width line segments of 0.002W and length is greater than 0.995W and remove; Then sort to remaining all prospect width line segments by length order from small to large, the prospect width line segment getting middle 80% forms prospect width line segment aggregate, is designated as { WL' _n, wherein, WL' _nrepresent { WL' _nin the length of n-th section of prospect width line segment, 1≤n≤N', N' represents { WL' _nin total hop count of prospect width line segment of comprising.

A4, calculating { WL' _nin the mean value of length of all prospect width line segments, be designated as W _front', then by W _front' as { d _r(x, y) } in the mean breadth W of foreground target _fronteven, W _front=W _front'.

Fig. 3 gives W _a' geometric representation.

4. the visual comfort evaluation and foreca value smmo of stereo-picture to be evaluated is revised, the revised visual comfort evaluation and foreca value of stereo-picture to be evaluated is designated as smm, wherein, max () is for getting max function, and P represents the tortuosity attenuation coefficient of the visual comfort evaluation model under the scene mode belonging to stereo-picture to be evaluated, S _r' represent { d _r(x, y) } in the average prospect line hop count of foreground target, S _c' represent { d _r(x, y) } in the average prospect alignment hop count of foreground target, T _frepresent the parallactic angle threshold value of setting, T _rrepresent the prospect line hop count threshold value of setting, T _crepresent the prospect alignment hop count threshold value of setting, get P=1.6, T in the present embodiment _f=2.0, T _r=2, T _c=1.5.

In this particular embodiment, step 4. in { d _r(x, y) } in the average prospect line hop count of foreground target ${S_r}^{\&prime;}=\frac{\underset{row=1}{\overset{H}{\&Sigma;}}{\mathrm{RLQ}}_{row}}{\underset{row=1}{\overset{H}{\&Sigma;}}\mathrm{\&epsiv;}\left({\mathrm{RLQ}}_{row}\right)},$ { d _r(x, y) } in the average prospect alignment hop count of foreground target ${S_c}^{\&prime;}=\frac{\underset{col=1}{\overset{W}{\&Sigma;}}{\mathrm{CLQ}}_{col}}{\underset{col=1}{\overset{W}{\&Sigma;}}\mathrm{\&epsiv;}\left({\mathrm{CLQ}}_{col}\right)},$ Wherein, S _r' and S _c' be { d _r(x, y) } in the tortuosity parameter of foreground target, 1≤row≤H, 1≤col≤W, RLQ _rowrepresent { d _r(x, y) } in row capable in the quantity of prospect line section, $\mathrm{\&epsiv;}\left({\mathrm{RLQ}}_{row}\right)=\{\begin{array}{c}1,{\mathrm{RLQ}}_{row}>0\\ 0,{\mathrm{RLQ}}_{row}\&le;0\end{array},$ CLQ _colrepresent { d _r(x, y) } in col row in the quantity of prospect alignment section, $\mathrm{\&epsiv;}\left({\mathrm{CLQ}}_{col}\right)=\left\{\begin{array}{c}1,CL{Q}_{col}>0\\ 0,CL{Q}_{col}\&le;0\end{array}\right..$

Wherein, RLQ _rowand CLQ _colacquisition process be:

B1, to { d _r(x, y) } carry out binary conversion treatment, obtain { d _r(x, y) } binary image, be designated as { BI (x, y) }, the pixel value being the pixel of (x, y) by coordinate position in { BI (x, y) } is designated as BI (x, y), as { d _r(x, y) } in coordinate position be the pixel of (x, y) when belonging to foreground target, make BI (x, y)=1, as { d _r(x, y) } in coordinate position be the pixel of (x, y) when belonging to background region, make BI (x, y)=0.

B2, to { BI (x, y) 2 dilation operations of mathematical morphology, 4 erosion operations, 2 dilation operations } are carried out successively, obtain { BI'(x, y) }, wherein, BI'(x, y) represent { BI'(x, y) }, coordinate position is the pixel value of the pixel of (x, y).

The quantity of effective prospect line section in often row in b3, statistics { BI'(x, y) }, by capable for the row in { BI'(x, y) } effectively the quantity of prospect line section be designated as RLQ _row', RLQ _row' acquisition process be: b3-1, make RLQ _row' initial value be 0; B3-2, from { BI'(x, y) the 1st pixel that the row } is capable starts to scan to the right, when to scan the 1st pixel value be the pixel of 1, using the original position of this pixel as the 1st section of prospect line section, and the row coordinate of this pixel is designated as x ₁', continuing scanning to the right and obtain the 1st section of prospect line section until scanning till pixel value is the pixel of 0, is that the row coordinate of the pixel of 0 is designated as x by this pixel value ₂', if the length x of the 1st section of prospect line section ₂'-x ₁' be greater than 0.005W, then determine that the 1st section of prospect line section is effective prospect line section, and make RLQ _row'=RLQ _row'+1, if the length x of the 1st section of prospect line section ₂'-x ₁' be less than or equal to 0.005W, then determine that the 1st section of prospect line section is invalid prospect line section; Continue to scan to the right, to determine whether the 1st section of prospect line section is the mode that effective prospect line section is identical with acquisition the 1st section of prospect line section, obtain row capable in all effective prospect line section, and statistics obtain row capable in the amount R LQ of effective prospect line section _row'; Wherein, the initial value of row is 1,1≤row≤H, 1≤x ₁' <x ₂'≤W, RLQ _row'=RLQ _row"=" in '+1 is assignment.

Equally, the quantity of effective prospect alignment section in the often row in statistics { BI'(x, y) }, by capable for the col in { BI'(x, y) } effectively the quantity of prospect alignment section be designated as CLQ _col', CLQ _col' acquisition process be: b3-1), make CLQ _col' initial value be 0; B3-2), from { BI'(x, 1st pixel of the col row y) } starts downward scanning, when to scan the 1st pixel value be the pixel of 1, using the original position of this pixel as the 1st section of prospect alignment section, and the row-coordinate of this pixel is designated as y ₁', continuing scanning downwards and obtain the 1st section of prospect alignment section until scanning till pixel value is the pixel of 0, is that the row-coordinate of the pixel of 0 is designated as y by this pixel value ₂', if the length y of the 1st section of prospect alignment section ₂'-y ₁' be greater than 0.005H, then determine that the 1st section of prospect alignment section is effective prospect alignment section, and make CLQ _col'=CLQ _col'+1, if the length y of the 1st section of prostatitis alignment section ₂'-y ₁' be less than or equal to 0.005H, then determine that the 1st section of prospect alignment section is invalid prospect alignment section; Continue scanning downwards, to determine whether the 1st section of prospect alignment section is the mode that effective prospect alignment section is identical with acquisition the 1st section of prospect alignment section, obtain col capable in all effective prospect alignment section, and statistics obtains the quantity CLQ of effective prospect alignment section in col row _col'; Wherein, the initial value of col is 1,1≤col≤W, 1≤y ₁' <y ₂'≤H, CLQ _col'=CLQ _col"=" in '+1 is assignment.

B4, make RLQ _row=RLQ _row', make CLQ _col=CLQ _col'.

In above-mentioned, the obtain manner of the foreground target in the right viewpoint parallax gray level image of every width stereo-picture of step 2. in neutral body image data base and background region and step 3. in stereo-picture to be evaluated right viewpoint parallax gray level image in foreground target identical with the obtain manner of background region, using the every width stereo-picture in stereoscopic image data storehouse and stereo-picture to be evaluated all as pending stereo-picture, foreground target in the right viewpoint parallax gray level image of then pending stereo-picture and the acquisition process of background region are: the grey level histogram obtaining the right viewpoint parallax gray level image of pending stereo-picture, then existing maximum variance between clusters (OTSU) is adopted, and the grey level histogram of right viewpoint parallax gray level image according to pending stereo-picture, obtain the intensity slicing threshold value of the right viewpoint parallax gray level image of pending stereo-picture, be designated as P _seg, again pixel value in the right viewpoint parallax gray level image of pending stereo-picture is more than or equal to P _segpixel be defined as foreground pixel point, and pixel value in the right viewpoint parallax gray level image of pending stereo-picture is less than P _segpixel be defined as background pixel, the last foreground target be made up of all foreground pixel points in the right viewpoint parallax gray level image of pending stereo-picture, forms the background region in the right viewpoint parallax gray level image of pending stereo-picture by all background pixels.

At this, the intensity slicing threshold value P of the right viewpoint parallax gray level image of pending stereo-picture _segacquisition process be:

X1, set up intensity slicing threshold value find target function, be designated as T, T=w _f× (μ-μ _f) ²+ w _b× (μ-μ _b) ², wherein, w _frepresent that total number of the foreground pixel point in the right viewpoint parallax gray level image of pending stereo-picture accounts for the ratio of total number of the pixel in the right viewpoint parallax gray level image of pending stereo-picture, μ _frepresent the average of the pixel value of all foreground pixel points in the right viewpoint parallax gray level image of pending stereo-picture, w _brepresent that total number of the background pixel in the right viewpoint parallax gray level image of pending stereo-picture accounts for the ratio of total number of the pixel in the right viewpoint parallax gray level image of pending stereo-picture, μ _brepresent the average of the pixel value of all background pixels in the right viewpoint parallax gray level image of pending stereo-picture, μ represents the average of the pixel value of all pixels in the right viewpoint parallax gray level image of pending stereo-picture, μ=w _f× μ _f+ w _b× μ _b, Hist _dg () represents the grey level histogram { Hist of the right viewpoint parallax gray level image of pending stereo-picture _d(g) } in pixel value be the number of the pixel of g, 0≤g _min≤ g≤g _max≤ 255, g _minrepresent the grey level histogram { Hist of the right viewpoint parallax gray level image of pending stereo-picture _d(g) } in the number of the pixel minimum pixel value that is greater than 0, g _maxrepresent the grey level histogram { Hist of the right viewpoint parallax gray level image of pending stereo-picture _d(g) } in the number of the pixel max pixel value that is greater than 0; Grey level histogram { the Hist of the right viewpoint parallax gray level image of pending stereo-picture _d(g) } in abscissa represent the number that gray scale, the ordinate of 0 ~ 255 represent the number of pixel, each pillar represents the pixel of different grey-scale.

X2, at interval [g _min, g _max] interior traversal t, the value of t when making T maximum is defined as the intensity slicing threshold value P of the right viewpoint parallax gray level image of pending stereo-picture _seg.

In above-mentioned, step 2. in determine to determine that the mode of the scene mode belonging to stereo-picture to be evaluated is identical during the mode of the scene mode belonging to every width stereo-picture in stereoscopic image data storehouse and step are 3., using the every width stereo-picture in stereoscopic image data storehouse and stereo-picture to be evaluated all as pending stereo-picture, then determine that the detailed process of the scene mode belonging to pending stereo-picture is:

Z1, for the foreground target in the right viewpoint parallax gray level image of pending stereo-picture, if f _a ^*>1 °, then determine that foreground target is convex and be in non-comfort zone in screen; If 0 °≤f _a ^*<1 °, then to determine that foreground target is convex and be in comfort zone in screen; If-1 ° of <f _a ^*<0 °, then to determine that foreground target is recessed and be in comfort zone in screen; If f _a ^*<-1 °, then determine that foreground target is recessed and be in non-comfort zone in screen.

For the background region in the right viewpoint parallax gray level image of pending stereo-picture, if b _a ^*>1 °, then determine that background region is convex and be in non-comfort zone in screen; If 0 °≤b _a ^*<1 °, then to determine that background region is convex and be in comfort zone in screen; If-1 ° of <b _a ^*<0 °, then to determine that background region is recessed and be in comfort zone in screen; If b _a ^*<-1 °, then determine that background region is recessed and be in non-comfort zone in screen.

Wherein, f _a ^*represent the parallactic angle of the foreground target in the right viewpoint parallax gray level image of pending stereo-picture, f _a ^*=f-k, b _a ^*represent the parallactic angle of the background region in the right viewpoint parallax gray level image of pending stereo-picture, b _a ^*=b-k, f represent the convergent angle of human eye binocular viewing foreground target, b represents the convergent angle of human eye binocular viewing background region, k represents the adjustment angle of human eye binocular, arctan () is tan of negating, p represents the interpupillary distance of human eye binocular, L represents the width of display, N represents the horizontal resolution of display, h represents the distance of human eye to display, F represents the mean parallax amplitude of the foreground target in the right viewpoint parallax gray level image of pending stereo-picture, and F is in units of pixel $F=\frac{\underset{x=1}{\overset{W}{\&Sigma;}}\underset{y=1}{\overset{H}{\&Sigma;}}\mathrm{\&epsiv;}\left({d_R}^{*}\right(x,y)-P_{seg})\&times;{d_R}^{*}(x,y)}{\underset{x=1}{\overset{W}{\&Sigma;}}\underset{y=1}{\overset{H}{\&Sigma;}}\mathrm{\&epsiv;}\left({d_R}^{*}\right(x,y)-P_{seg})},\mathrm{\&epsiv;}\left({d_R}^{*}\right(x,y)-P_{seg})=\left\{\begin{array}{c}1,{d_R}^{*}(x,y)-P_{seg}\&GreaterEqual;0\\ 0,{d_R}^{*}(x,y)-P_{seg}<0\end{array}\right.,$ B represents the mean parallax amplitude of the background region in the right viewpoint parallax gray level image of pending stereo-picture, B in units of pixel, $B=\frac{\underset{x=1}{\overset{W}{\&Sigma;}}\underset{y=1}{\overset{H}{\&Sigma;}}\mathrm{\&epsiv;}(P_{seg}-{d_R}^{*}(x,y\left)\right)\&times;{d_R}^{*}(x,y)}{\underset{x=1}{\overset{W}{\&Sigma;}}\underset{y=1}{\overset{H}{\&Sigma;}}\mathrm{\&epsiv;}(P_{seg}-{d_R}^{*}(x,y\left)\right)},\mathrm{\&epsiv;}(P_{seg}-{d_R}^{*}(x,y\left)\right)=\left\{\begin{array}{c}1,P_{seg}-{d_R}^{*}(x,y)\&GreaterEqual;0\\ 0,P_{seg}-{d_R}^{*}(x,y)<0\end{array}\right.,$ D _r ^*(x, y) represents that in the right viewpoint parallax gray level image of pending stereo-picture, coordinate position is the pixel value of the pixel of (x, y), and Fig. 2 gives the geometric representation of f, b, k, p, h, F, B.

Z2, according to the foreground target in the right viewpoint parallax gray level image of pending stereo-picture relative to the concavity and convexity of screen and whether be in comfort zone, background region relative to screen concavity and convexity and whether be in comfort zone, determine the scene mode belonging to pending stereo-picture, be specially: if foreground target is convex be in that non-comfort zone, background region are recessed is in non-comfort zone, i.e. f in screen in screen _a ^*>1 ° and b _a ^*<-1 °, then determine that pending stereo-picture belongs to the 1st kind of scene mode; If foreground target is convex be in that non-comfort zone, background region are convex is in non-comfort zone, i.e. f in screen in screen _a ^*>1 ° and b _a ^*>1 °, then determine that pending stereo-picture belongs to the 2nd kind of scene mode; If foreground target is recessed be in that non-comfort zone, background region are recessed is in non-comfort zone, i.e. f in screen in screen _a ^*<-1 ° and b _a ^*<-1 °, then determine that pending stereo-picture belongs to the 3rd kind of scene mode; If foreground target is convex be in that non-comfort zone, background region are recessed is in comfort zone, i.e. f in screen in screen _a ^*>1 ° and-1 °≤b _a ^*<0 °, then determine that pending stereo-picture belongs to the 4th kind of scene mode; If foreground target is convex be in that non-comfort zone, background region are convex is in comfort zone, i.e. f in screen in screen _a ^*>1 ° and 0 °≤b _a ^*≤ 1 °, then determine that pending stereo-picture belongs to the 5th kind of scene mode; If foreground target is convex be in that comfort zone, background region are recessed is in non-comfort zone, i.e. 0 °≤f in screen in screen _a ^*≤ 1 ° and b _a ^*<-1 °, then determine that pending stereo-picture belongs to the 6th kind of scene mode; If foreground target is recessed be in that comfort zone, background region are recessed is in non-comfort zone, namely-1 °≤f in screen in screen _a ^*<0 ° and b _a ^*<-1 °, then determine that pending stereo-picture belongs to the 7th kind of scene mode; If foreground target is convex be in that comfort zone, background region are recessed is in comfort zone, i.e. 0 °≤f in screen in screen _a ^*≤ 1 ° and-1 °≤b _a ^*<0 °, then determine that pending stereo-picture belongs to the 8th kind of scene mode; If foreground target is convex be in that comfort zone, background region are convex is in comfort zone, i.e. 0 °≤f in screen in screen _a ^*≤ 1 ° and 0 °≤b _a ^*≤ 1 °, then determine that pending stereo-picture belongs to the 9th kind of scene mode; If foreground target is recessed be in that comfort zone, background region are recessed is in comfort zone, namely-1 °≤f in screen in screen _a ^*<0 ° and-1 °≤b _a ^*<0 °, then determine that pending stereo-picture belongs to the 10th kind of scene mode.

In the present embodiment, utilize 5 of evaluate image quality evaluating method conventional objective parameters as evaluation index, namely Pearson correlation coefficient (the Pearsonlinearcorrelationcoefficient under nonlinear regression condition, PLCC), Spearman coefficient correlation (Spearmanrankordercorrelationcoefficient, SROCC), Kendall coefficient correlation (Kendallrank-ordercorrelationcoefficient, KROCC), average absolute value error (MeanAbsoluteError, MAE), root-mean-square error (rootmeansquarederror, RMSE).PLCC reflects the correlation of objective evaluation predicted value and subjective assessment value; SROCC and KROCC reflects monotonicity, the consistency of objective evaluation predicted value and subjective assessment value; MAE and RMSE reflects the accuracy of objective evaluation predicted value.The inventive method is utilized to obtain 120 width stereo-pictures final visual comfort evaluation and foreca value separately, five parameter Logistic function nonlinear fittings are done to the final visual comfort evaluation and foreca value of this 120 width stereo-picture, PLCC, SROCC and KROCC value is higher, and the correlation that the evaluation result of the less explanation the inventive method of MAE and RMSE value and mean subjective are marked between average is better.Table 1 gives the evaluation index value of the inventive method, illustrates that the result of objective evaluation result and the human eye subjective perception utilizing the inventive method to obtain is more consistent, describes the validity of the inventive method well.

Table 1 evaluation index value

Evaluation index	PLCC	SROCC	KROCC	MAE	RMSE
							0.919	0.909	0.748	0.237	0.318

Claims (8)

1., based on a stereo image vision comfort level evaluation method for scene mode classification, it is characterized in that comprising the following steps:

1. determine 10 kinds of scene modes of the natural scene shown by three-dimensional display, the 1st kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is convex be in that non-comfort zone, background region are recessed is in non-comfort zone in screen; 2nd kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is convex be in that non-comfort zone, background region are convex is in non-comfort zone in screen; 3rd kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is recessed be in that non-comfort zone, background region are recessed is in non-comfort zone in screen; 4th kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is convex be in that non-comfort zone, background region are recessed is in comfort zone in screen; 5th kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is convex be in that non-comfort zone, background region are convex is in comfort zone in screen; 6th kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is convex be in that comfort zone, background region are recessed is in non-comfort zone in screen; 7th kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is recessed be in that comfort zone, background region are recessed is in non-comfort zone in screen; 8th kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is convex be in that comfort zone, background region are recessed is in comfort zone in screen; 9th kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is convex be in that comfort zone, background region are convex is in comfort zone in screen; 10th kind of scene mode in screen, to be that the foreground target in the right viewpoint parallax gray level image of stereo-picture is recessed be in that comfort zone, background region are recessed is in comfort zone in screen;

2. every width stereo-picture is selected to have a stereoscopic image data storehouse of the mean subjective scoring average of visual comfort; Then according to the parallactic angle of foreground target in the right viewpoint parallax gray level image of the every width stereo-picture in stereoscopic image data storehouse and the parallactic angle of background region, the scene mode belonging to every width stereo-picture in stereoscopic image data storehouse is determined; Again according to the parallactic angle of foreground target in the respective right viewpoint parallax gray level image of all stereo-pictures belonging to often kind of scene mode in stereoscopic image data storehouse and the parallactic angle of background region, set up the visual comfort evaluation model under often kind of scene mode, the visual comfort evaluation model under n-th kind of scene mode is described below: SMMO _n=(4.2028-u (Q _n))-v (Q _n) × D _a(Q _n)+0.1912 × ln (W _a)-0.0208 × D _a(Q _n) × ln (W _a), wherein, 1≤n≤10, SMMO _nrepresent the output of the visual comfort evaluation model under n-th kind of scene mode, u (Q _n) and v (Q _n) be constant, D _a(Q _n) represent that a width to be entered belongs to the global disparity angle of the right viewpoint parallax gray level image of the stereo-picture to be evaluated of n-th kind of scene mode, D _a(Q _n)=Q _n× | f _a|+(1-Q _n) × | b _a|, Q _nrepresent the weight under n-th kind of scene mode, f _arepresent that a width to be entered belongs to the parallactic angle of the foreground target in the right viewpoint parallax gray level image of the stereo-picture to be evaluated of n-th kind of scene mode, b _arepresent that a width to be entered belongs to the parallactic angle of the background region in the right viewpoint parallax gray level image of the stereo-picture to be evaluated of n-th kind of scene mode, W _arepresent that a width to be entered belongs to the width angle of the foreground target in the right viewpoint parallax gray level image of the stereo-picture to be evaluated of n-th kind of scene mode, symbol " || " is the symbol that takes absolute value;

3. the right viewpoint parallax gray level image of stereo-picture to be evaluated is designated as { d _r(x, y) }, wherein, 1≤x≤W, 1≤y≤H, W represents the width of stereo-picture to be evaluated, it is consistent with the width of the every width stereo-picture in stereoscopic image data storehouse, H represents the height of stereo-picture to be evaluated, and it is consistent with the height of the every width stereo-picture in stereoscopic image data storehouse, d _r(x, y) represents { d _r(x, y) } in coordinate position be the pixel value of the pixel of (x, y); Then adopt and step 2. in identical mode, determine the scene mode belonging to stereo-picture to be evaluated; Scene mode again belonging to stereo-picture to be evaluated, chooses the visual comfort evaluation model under this scene mode; Then according to the visual comfort evaluation model under this scene mode and { d _r(x, y) } in the parallactic angle, { d of foreground target _r(x, y) } in the parallactic angle of background region and { d _r(x, y) } in the width angle of foreground target, calculate the visual comfort evaluation and foreca value of stereo-picture to be evaluated, be designated as smmo, suppose that stereo-picture to be evaluated belongs to n-th kind of scene mode, then smmo=(4.2028-u (Q _n))-v (Q _n) × D _a' (Q _n)+0.1912 × ln (W _a')-0.0208 × D _a' (Q _n) × ln (W _a'), wherein, D _a' (Q _n) represent { d _r(x, y) } global disparity angle, D _a' (Q _n)=Q _n× | f _a' |+(1-Q _n) × | b _a' |, f _a' represent { d _r(x, y) } in the parallactic angle of foreground target, b _a' represent { d _r(x, y) } in the parallactic angle of background region, W _a' represent { d _r(x, y) } in the width angle of foreground target;

4. the visual comfort evaluation and foreca value smmo of stereo-picture to be evaluated is revised, the revised visual comfort evaluation and foreca value of stereo-picture to be evaluated is designated as smm, wherein, max () is for getting max function, and P represents the tortuosity attenuation coefficient of the visual comfort evaluation model under the scene mode belonging to stereo-picture to be evaluated, S _r' represent { d _r(x, y) } in the average prospect line hop count of foreground target, S _c' represent { d _r(x, y) } in the average prospect alignment hop count of foreground target, T _frepresent the parallactic angle threshold value of setting, T _rrepresent the prospect line hop count threshold value of setting, T _crepresent the prospect alignment hop count threshold value of setting.

2. a kind of stereo image vision comfort level evaluation method based on scene mode classification according to claim 1, is characterized in that described step 2. middle Q _nacquisition process be:

-1 2., total width number of supposing the stereo-picture belonging to n-th kind of scene mode in stereoscopic image data storehouse is M', wherein, and M' >=1;

2.-2, q is made _ninitial value be 0.1, make Q _ninitial value be 0;

2. the weight value-3, calculated under n-th kind of scene mode is q _ntime stereoscopic image data storehouse in belong to the global disparity angle of the right viewpoint parallax gray level image of every width stereo-picture of n-th kind of scene mode, be q by the weight value under n-th kind of scene mode _ntime stereoscopic image data storehouse in belong to the right viewpoint parallax gray level image of the m' width stereo-picture of n-th kind of scene mode global disparity angle be designated as D _{m', a}(q _n), D _{m', a}(q _n)=q _n× | f _{m', a}|+(1-q _n) × | b _{m', a}|, wherein, 1≤m'≤M', f _{m', a}and b _{m', a}belong to the parallactic angle of foreground target in the right viewpoint parallax gray level image of the m' width stereo-picture of n-th kind of scene mode and the parallactic angle of background region in corresponding expression stereoscopic image data storehouse, symbol " || " is the symbol that takes absolute value;

2. the weight value-4, calculated under n-th kind of scene mode is q _ntime stereoscopic image data storehouse in belong to n-th kind of scene mode every width stereo-picture eliminate the comfort level objective evaluation value of parallax linear effect, be q by the weight value under n-th kind of scene mode _ntime stereoscopic image data storehouse in belong to n-th kind of scene mode the m' width stereo-picture comfort level objective evaluation value that eliminates parallax linear effect be designated as VC _h(D _{m', a}(q _n), W _a), VC _h(D _{m', a}(q _n), W _a)=4.2028+0.1912 × ln (W _a)-0.0208 × D _{m', a}(q _n) × ln (W _a); Then the weight value calculated under n-th kind of scene mode is q _ntime stereoscopic image data storehouse in belong to the comfort level objective evaluation value of the parallax linear effect of every width stereo-picture of n-th kind of scene mode, be q by the weight value under n-th kind of scene mode _ntime stereoscopic image data storehouse in belong to the parallax linear effect of the m' width stereo-picture of n-th kind of scene mode comfort level objective evaluation value be designated as ERR _m', ERR _m'=VC _h(D _{m', a}(q _n), W _a)-MOS _m', wherein, MOS _m'represent that the weight value under n-th kind of scene mode is q _ntime stereoscopic image data storehouse in belong to the comfort level subjective assessment value of the m' width stereo-picture of n-th kind of scene mode; Adopting least square method again, is q to the weight value under n-th kind of scene mode _ntime stereoscopic image data storehouse in belong to n-th kind of scene mode the respective global disparity angle of right viewpoint parallax gray level image of all stereo-pictures and the comfort level objective evaluation value of respective parallax linear effect carry out linear fit, the weight value obtained under n-th kind of scene mode is q _ntime stereoscopic image data storehouse in belong to a fitting a straight line equation corresponding to all stereo-pictures of n-th kind of scene mode: ERR=u (q _n)+v (q _n) × D _a(q _n), wherein, ERR is for representing the dependent variable of parallax to the linear effect of objective comfort level, u (q _n) and v (q _n) be ERR=u (q _n)+v (q _n) × D _a(q _n) in constant, D _a(q _n) be for representing that the weight value under n-th kind of scene mode is q _ntime stereoscopic image data storehouse in belong to the independent variable at the global disparity angle of the right viewpoint parallax gray level image of arbitrary width stereo-picture of n-th kind of scene mode;

2. be-5, q according to the weight value under n-th kind of scene mode _ntime stereoscopic image data storehouse in belong to a fitting a straight line equation corresponding to all stereo-pictures of n-th kind of scene mode, the weight value obtained under n-th kind of scene mode is q _ntime stereoscopic image data storehouse in belong to the comfort level objective evaluation equation of all stereo-pictures of n-th kind of scene mode: SMMO _n(q _n)=VC _h(D _a(q _n), W _a)-ERR=(4.2028-u (q _n))-v (q _n) × D _a(q _n)+0.1912 × ln (W _a)-0.0208 × D _a(q _n) × ln (W _a), wherein, SMMO _n(q _n) for having superposed the comfort level objective evaluation value of parallax linear effect and parallax non-linear effects, VC _h(D _a(q _n), W _a) represent that the weight value under n-th kind of scene mode is q _ntime stereoscopic image data storehouse in belong to n-th kind of scene mode arbitrary width stereo-picture eliminate the comfort level objective evaluation value of parallax linear effect; The weight value of then measuring under n-th kind of scene mode is q _ntime stereoscopic image data storehouse in belong to the comfort level objective evaluation value of all stereo-pictures of n-th kind of scene mode and the fitting degree of corresponding comfort level subjective assessment value;

2.-6, q is made _n=q _n+ 0.1, then return step 2.-3 continuation execution, until q _nterminate when equaling 1.1, corresponding fitting degree when the weight value obtained under n-th kind of scene mode is respectively 0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8,0.9,1.0, then by weight value assignment corresponding for best fit degree to Q _n, wherein, q _n=q _n"=" in+0.1 is assignment;

Described step is middle u (Q 2. _n) value be Monomial coefficient in fitting a straight line equation corresponding to best fit degree; Described step is middle v (Q 2. _n) value be constant in fitting a straight line equation corresponding to best fit degree.

3. a kind of stereo image vision comfort level evaluation method based on scene mode classification according to claim 1 and 2, it is characterized in that the foreground target in the right viewpoint parallax gray level image of stereo-picture to be evaluated during the obtain manner of the foreground target in the right viewpoint parallax gray level image of every width stereo-picture of described step 2. in neutral body image data base and background region and described step are 3. is identical with the obtain manner of background region, using the every width stereo-picture in stereoscopic image data storehouse and stereo-picture to be evaluated all as pending stereo-picture, foreground target in the right viewpoint parallax gray level image of then pending stereo-picture and the acquisition process of background region are: the grey level histogram obtaining the right viewpoint parallax gray level image of pending stereo-picture, then maximum variance between clusters is adopted, and the grey level histogram of right viewpoint parallax gray level image according to pending stereo-picture, obtain the intensity slicing threshold value of the right viewpoint parallax gray level image of pending stereo-picture, be designated as P _seg, again pixel value in the right viewpoint parallax gray level image of pending stereo-picture is more than or equal to P _segpixel be defined as foreground pixel point, and pixel value in the right viewpoint parallax gray level image of pending stereo-picture is less than P _segpixel be defined as background pixel, the last foreground target be made up of all foreground pixel points in the right viewpoint parallax gray level image of pending stereo-picture, forms the background region in the right viewpoint parallax gray level image of pending stereo-picture by all background pixels.

4. a kind of stereo image vision comfort level evaluation method based on scene mode classification according to claim 3, is characterized in that the intensity slicing threshold value P of the right viewpoint parallax gray level image of described pending stereo-picture _segacquisition process be:

X1, set up intensity slicing threshold value find target function, be designated as T, T=w _f× (μ-μ _f) ²+ w _b× (μ-μ _b) ², wherein, w _frepresent that total number of the foreground pixel point in the right viewpoint parallax gray level image of pending stereo-picture accounts for the ratio of total number of the pixel in the right viewpoint parallax gray level image of pending stereo-picture, μ _frepresent the average of the pixel value of all foreground pixel points in the right viewpoint parallax gray level image of pending stereo-picture, w _brepresent that total number of the background pixel in the right viewpoint parallax gray level image of pending stereo-picture accounts for the ratio of total number of the pixel in the right viewpoint parallax gray level image of pending stereo-picture, μ _brepresent the average of the pixel value of all background pixels in the right viewpoint parallax gray level image of pending stereo-picture, μ represents the average of the pixel value of all pixels in the right viewpoint parallax gray level image of pending stereo-picture, μ=w _f× μ _f+ w _b× μ _b, Hist _dg () represents the grey level histogram { Hist of the right viewpoint parallax gray level image of pending stereo-picture _d(g) } in pixel value be the number of the pixel of g, 0≤g _min≤ g≤g _max≤ 255, g _minrepresent the grey level histogram { Hist of the right viewpoint parallax gray level image of pending stereo-picture _d(g) } in the number of the pixel minimum pixel value that is greater than 0, g _maxrepresent the grey level histogram { Hist of the right viewpoint parallax gray level image of pending stereo-picture _d(g) } in the number of the pixel max pixel value that is greater than 0;

X2, at interval [g _min, g _max] interior traversal t, the value of t when making T maximum is defined as the intensity slicing threshold value P of the right viewpoint parallax gray level image of pending stereo-picture _seg.

5. a kind of stereo image vision comfort level evaluation method based on scene mode classification according to claim 3, it is characterized in that determining during described step 2. to determine that the mode of the scene mode belonging to stereo-picture to be evaluated is identical during the mode of the scene mode belonging to every width stereo-picture in stereoscopic image data storehouse and described step are 3., using the every width stereo-picture in stereoscopic image data storehouse and stereo-picture to be evaluated all as pending stereo-picture, then determine that the detailed process of the scene mode belonging to pending stereo-picture is:

Z1, for the foreground target in the right viewpoint parallax gray level image of pending stereo-picture, if f _a ^*>1 °, then determine that foreground target is convex and be in non-comfort zone in screen; If 0 °≤f _a ^*<1 °, then to determine that foreground target is convex and be in comfort zone in screen; If-1 ° of <f _a ^*<0 °, then to determine that foreground target is recessed and be in comfort zone in screen; If f _a ^*<-1 °, then determine that foreground target is recessed and be in non-comfort zone in screen;

For the background region in the right viewpoint parallax gray level image of pending stereo-picture, if b _a ^*>1 °, then determine that background region is convex and be in non-comfort zone in screen; If 0 °≤b _a ^*<1 °, then to determine that background region is convex and be in comfort zone in screen; If-1 ° of <b _a ^*<0 °, then to determine that background region is recessed and be in comfort zone in screen; If b _a ^*<-1 °, then determine that background region is recessed and be in non-comfort zone in screen;

Wherein, f _a ^*represent the parallactic angle of the foreground target in the right viewpoint parallax gray level image of pending stereo-picture, f _a ^*=f-k, b _a ^*represent the parallactic angle of the background region in the right viewpoint parallax gray level image of pending stereo-picture, b _a ^*=b-k, f represent the convergent angle of human eye binocular viewing foreground target, b represents the convergent angle of human eye binocular viewing background region, k represents the adjustment angle of human eye binocular, arctan () is tan of negating, p represents the interpupillary distance of human eye binocular, L represents the width of display, N represents the horizontal resolution of display, h represents the distance of human eye to display, F represents the mean parallax amplitude of the foreground target in the right viewpoint parallax gray level image of pending stereo-picture, and F is in units of pixel $F=\frac{\underset{x=1}{\overset{W}{\&Sigma;}}\underset{y=1}{\overset{H}{\&Sigma;}}\mathrm{\&epsiv;}\left({d_R}^{*}\right(x,y)-P_{seg})\&times;{d_R}^{*}(x,y)}{\underset{x=1}{\overset{W}{\&Sigma;}}\underset{y=1}{\overset{H}{\&Sigma;}}\mathrm{\&epsiv;}\left({d_R}^{*}\right(x,y)-P_{seg})},\mathrm{\&epsiv;}\left({d_R}^{*}\right(x,y)-P_{seg})=\left\{\begin{array}{c}1,{d_R}^{*}(x,y)-P_{seg}\&GreaterEqual;0\\ 0,{d_R}^{*}(x,y)-P_{seg}<0\end{array}\right.,$ B represents the mean parallax amplitude of the background region in the right viewpoint parallax gray level image of pending stereo-picture, B in units of pixel, $B=\frac{\underset{x=1}{\overset{W}{\&Sigma;}}\underset{y=1}{\overset{H}{\&Sigma;}}\mathrm{\&epsiv;}(P_{seg}-{d_R}^{*}(x,y\left)\right)\&times;{d_R}^{*}(x,y)}{\underset{x=1}{\overset{W}{\&Sigma;}}\underset{y=1}{\overset{H}{\&Sigma;}}\mathrm{\&epsiv;}(P_{seg}-{d_R}^{*}(x,y\left)\right)},\mathrm{\&epsiv;}(P_{seg}-{d_R}^{*}(x,y\left)\right)=\left\{\begin{array}{c}1,P_{seg}-{d_R}^{*}(x,y)\&GreaterEqual;0\\ 0,P_{seg}-{d_R}^{*}(x,y)<0\end{array}\right.,$ D _r ^*(x, y) represents that in the right viewpoint parallax gray level image of pending stereo-picture, coordinate position is the pixel value of the pixel of (x, y);

Z2, according to the foreground target in the right viewpoint parallax gray level image of pending stereo-picture relative to the concavity and convexity of screen and whether be in comfort zone, background region relative to screen concavity and convexity and whether be in comfort zone, determine the scene mode belonging to pending stereo-picture, be specially: if foreground target is convex be in that non-comfort zone, background region are recessed is in non-comfort zone in screen in screen, then determine that pending stereo-picture belongs to the 1st kind of scene mode; If foreground target is convex be in that non-comfort zone, background region are convex is in non-comfort zone in screen in screen, then determine that pending stereo-picture belongs to the 2nd kind of scene mode; If foreground target is recessed be in that non-comfort zone, background region are recessed is in non-comfort zone in screen in screen, then determine that pending stereo-picture belongs to the 3rd kind of scene mode; If foreground target is convex be in that non-comfort zone, background region are recessed is in comfort zone in screen in screen, then determine that pending stereo-picture belongs to the 4th kind of scene mode; If foreground target is convex be in that non-comfort zone, background region are convex is in comfort zone in screen in screen, then determine that pending stereo-picture belongs to the 5th kind of scene mode; If foreground target is convex be in that comfort zone, background region are recessed is in non-comfort zone in screen in screen, then determine that pending stereo-picture belongs to the 6th kind of scene mode; If foreground target is recessed be in that comfort zone, background region are recessed is in non-comfort zone in screen in screen, then determine that pending stereo-picture belongs to the 7th kind of scene mode; If foreground target is convex be in that comfort zone, background region are recessed is in comfort zone in screen in screen, then determine that pending stereo-picture belongs to the 8th kind of scene mode; If foreground target is convex be in that comfort zone, background region are convex is in comfort zone in screen in screen, then determine that pending stereo-picture belongs to the 9th kind of scene mode; If foreground target is recessed be in that comfort zone, background region are recessed is in comfort zone in screen in screen, then determine that pending stereo-picture belongs to the 10th kind of scene mode.

6. a kind of stereo image vision comfort level evaluation method based on scene mode classification according to claim 5, is characterized in that { d during described step 3. _r(x, y) } in the width angle of foreground target wherein, arctan () is tan of negating, and h represents the distance of human eye to display, W _frontrepresent { d _r(x, y) } in the mean breadth of foreground target, W _frontacquisition process be:

A1, to { d _r(x, y) } carry out binary conversion treatment, obtain { d _r(x, y) } binary image, be designated as { BI (x, y) }, the pixel value being the pixel of (x, y) by coordinate position in { BI (x, y) } is designated as BI (x, y), as { d _r(x, y) } in coordinate position be the pixel of (x, y) when belonging to foreground target, make BI (x, y)=1, as { d _r(x, y) } in coordinate position be the pixel of (x, y) when belonging to background region, make BI (x, y)=0;

A2, to { BI (x, y) } line scanning is carried out, for { BI (x, y) row } is capable, scan to the right from the 1st pixel that row is capable, when to scan the 1st pixel value be the pixel of 1, using the original position of this pixel as the 1st section of prospect width line segment, and the row coordinate of this pixel is designated as x ₁, continuing scanning to the right and obtain the 1st section of prospect width line segment until scanning till pixel value is the pixel of 0, is that the row coordinate of the pixel of 0 is designated as x by this pixel value ₂, the length of the 1st section of prospect width line segment is designated as WL ₁, WL ₁=x ₂-x ₁; Continue scan to the right, in the mode identical with acquisition the 1st section of prospect width line segment, obtain row capable in all prospect width line segments; Wherein, the initial value of row is 1,1≤row≤H, 1≤x ₁<x ₂≤ W;

All prospect width line segments that a3, all prospect width line segments length in { BI (x, y) } being less than 0.002W and length are greater than 0.995W are removed; Then sort to remaining all prospect width line segments by length order from small to large, the prospect width line segment getting middle 80% forms prospect width line segment aggregate, is designated as { WL' _n, wherein, WL' _nrepresent { WL' _nin the length of n-th section of prospect width line segment, 1≤n≤N', N' represents { WL' _nin total hop count of prospect width line segment of comprising;

A4, calculating { WL' _nin the mean value of length of all prospect width line segments, be designated as W _front', then by W _front' as { d _r(x, y) } in the mean breadth W of foreground target _fronteven, W _front=W _front'.

7. a kind of stereo image vision comfort level evaluation method based on scene mode classification according to claim 6, is characterized in that { d during described step 4. _r(x, y) } in the average prospect line hop count of foreground target { d _r(x, y) } in the average prospect alignment hop count of foreground target wherein, 1≤row≤H, 1≤col≤W, RLQ _rowrepresent { d _r(x, y) } in row capable in the quantity of prospect line section, $\mathrm{\&epsiv;}\left({\mathrm{RLQ}}_{row}\right)=\{\begin{array}{c}1,{\mathrm{RLQ}}_{row}>0\\ 0,{\mathrm{RLQ}}_{row}\&le;0\end{array},$ CLQ _colrepresent { d _r(x, y) } in col row in the quantity of prospect alignment section, $\mathrm{\&epsiv;}\left({\mathrm{CLQ}}_{col}\right)=\left\{\begin{array}{c}1,CL{Q}_{col}>0\\ 0,CL{Q}_{col}\&le;0\end{array}\right.;$

Wherein, RLQ _rowand CLQ _colacquisition process be:

B1, to { d _r(x, y) } carry out binary conversion treatment, obtain { d _r(x, y) } binary image, be designated as { BI (x, y) }, the pixel value being the pixel of (x, y) by coordinate position in { BI (x, y) } is designated as BI (x, y), as { d _r(x, y) } in coordinate position be the pixel of (x, y) when belonging to foreground target, make BI (x, y)=1, as { d _r(x, y) } in coordinate position be the pixel of (x, y) when belonging to background region, make BI (x, y)=0;

B2, to { BI (x, y) 2 dilation operations of mathematical morphology, 4 erosion operations, 2 dilation operations } are carried out successively, obtain { BI'(x, y) }, wherein, BI'(x, y) represent { BI'(x, y) }, coordinate position is the pixel value of the pixel of (x, y);

The quantity of effective prospect line section in often row in b3, statistics { BI'(x, y) }, by capable for the row in { BI'(x, y) } effectively the quantity of prospect line section be designated as RLQ _row', RLQ _row' acquisition process be: b3-1, make RLQ _row' initial value be 0; B3-2, from { BI'(x, y) the 1st pixel that the row } is capable starts to scan to the right, when to scan the 1st pixel value be the pixel of 1, using the original position of this pixel as the 1st section of prospect line section, and the row coordinate of this pixel is designated as x ₁', continuing scanning to the right and obtain the 1st section of prospect line section until scanning till pixel value is the pixel of 0, is that the row coordinate of the pixel of 0 is designated as x by this pixel value ₂', if the length x of the 1st section of prospect line section ₂'-x ₁' be greater than 0.005W, then determine that the 1st section of prospect line section is effective prospect line section, and make RLQ _row'=RLQ _row'+1, if the length x of the 1st section of prospect line section ₂'-x ₁' be less than or equal to 0.005W, then determine that the 1st section of prospect line section is invalid prospect line section; Continue to scan to the right, to determine whether the 1st section of prospect line section is the mode that effective prospect line section is identical with acquisition the 1st section of prospect line section, obtain row capable in all effective prospect line section, and statistics obtain row capable in the amount R LQ of effective prospect line section _row'; Wherein, the initial value of row is 1,1≤row≤H, 1≤x ₁' <x ₂'≤W, RLQ _row'=RLQ _row"=" in '+1 is assignment;

Equally, the quantity of effective prospect alignment section in the often row in statistics { BI'(x, y) }, by capable for the col in { BI'(x, y) } effectively the quantity of prospect alignment section be designated as CLQ _col', CLQ _col' acquisition process be: b3-1), make CLQ _col' initial value be 0; B3-2), from { BI'(x, 1st pixel of the col row y) } starts downward scanning, when to scan the 1st pixel value be the pixel of 1, using the original position of this pixel as the 1st section of prospect alignment section, and the row-coordinate of this pixel is designated as y ₁', continuing scanning downwards and obtain the 1st section of prospect alignment section until scanning till pixel value is the pixel of 0, is that the row-coordinate of the pixel of 0 is designated as y by this pixel value ₂', if the length y of the 1st section of prospect alignment section ₂'-y ₁' be greater than 0.005H, then determine that the 1st section of prospect alignment section is effective prospect alignment section, and make CLQ _col'=CLQ _col'+1, if the length y of the 1st section of prostatitis alignment section ₂'-y ₁' be less than or equal to 0.005H, then determine that the 1st section of prospect alignment section is invalid prospect alignment section; Continue scanning downwards, to determine whether the 1st section of prospect alignment section is the mode that effective prospect alignment section is identical with acquisition the 1st section of prospect alignment section, obtain col capable in all effective prospect alignment section, and statistics obtains the quantity CLQ of effective prospect alignment section in col row _col'; Wherein, the initial value of col is 1,1≤col≤W, 1≤y ₁' <y ₂'≤H, CLQ _col'=CLQ _col"=" in '+1 is assignment;

B4, make RLQ _row=RLQ _row', make CLQ _col=CLQ _col'.

8. a kind of stereo image vision comfort level evaluation method based on scene mode classification according to claim 7, is characterized in that getting P=1.6, T during described step 4. _f=2.0, T _r=2, T _c=1.5.