CN104333751B - The parallel shooting method for objectively evaluating of stereoscopic camera based on three-dimensional disturbance - Google Patents

The parallel shooting method for objectively evaluating of stereoscopic camera based on three-dimensional disturbance Download PDF

Info

Publication number
CN104333751B
CN104333751B CN201410625203.XA CN201410625203A CN104333751B CN 104333751 B CN104333751 B CN 104333751B CN 201410625203 A CN201410625203 A CN 201410625203A CN 104333751 B CN104333751 B CN 104333751B
Authority
CN
China
Prior art keywords
camera
parallel
shooting
stereoscopic camera
objective
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201410625203.XA
Other languages
Chinese (zh)
Other versions
CN104333751A (en
Inventor
杨嘉琛
楚容容
刘允
高志群
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Tianjin University
Original Assignee
Tianjin University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Tianjin University filed Critical Tianjin University
Priority to CN201410625203.XA priority Critical patent/CN104333751B/en
Publication of CN104333751A publication Critical patent/CN104333751A/en
Application granted granted Critical
Publication of CN104333751B publication Critical patent/CN104333751B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Landscapes

  • Testing, Inspecting, Measuring Of Stereoscopic Televisions And Televisions (AREA)

Abstract

The present invention relates to a kind of stereoscopic camera based on three-dimensional disturbance parallel shooting method for objectively evaluating: three-dimensional shape distortion μ when choosing stereoscopic camera parallel shooting pas an objective evaluation index of stereo pairs; Set up L bwith d ccorrelation k p, with k pas the comprehensive interactional objective evaluation index analyzing shooting distance and camera spacing; Stereoscopic camera is utilized to take stereo pairs in the situation of be arrangeding in parallel; Extract camera shooting distance L when stereoscopic camera is parallel to be taken b, camera spacing d cwith camera focus f; Calculate the objective evaluation index μ of stereo-picture pand k p, form a three-dimensional perception ε of final index p, further by ε pbe combined with camera focus f; Obtaining final objective evaluation index-ρ is the objective comprehensive evaluation index of parallel camera, and value more hour, illustrates that the parallel shooting quality of this stereoscopic camera is better.The present invention well can be applied to the quality evaluation of stereo-picture, evaluation result and subjective evaluation result correlation very strong.

Description

The parallel shooting method for objectively evaluating of stereoscopic camera based on three-dimensional disturbance
Technical field
The invention belongs to three-dimensional image acquisition technical field, be specifically related to the parallel shooting evaluation method of a kind of new stereoscopic camera.
Background technology
As everyone knows, image technique relates to each large fields such as amusement, industry, military affairs, is the indispensable important technology of scientific technological advance and world's progress.Nowadays, traditional two dimensional image can not be satisfied the demand, and new stereo-picture technology is used widely.Compared to two dimensional image, stereo-picture presses close to the true picture that human eye captures more, adds depth perception and the sense of reality, can reappear real scene more exactly.But, stereo-picture is in the process gathering, store, transmit, compress and show, due to the error of algorithm and equipment itself, inevitably can produce the distortion of each middle various kinds, as trapezoidal distortion, poly-loose conflict, theater effect, cardboard effect etc., so how to weigh out the impact of these distortion stereoscopic image quality exactly, become the emphasis topic of 3D (threedimensional, 3D) area research.In stereo-picture assessment technique research field, most people is placed on research emphasis in stereo image information loss degree, and the distortion level namely in stereo-picture a series of end processing sequences after collection, few people have the evaluation to original three-dimensional image in mind.In stereo-picture acquisition system, main application be the method for the left and right visual pattern superposition of simulation human eye system, in this process, relate to many parameters, as distance between axles, angle of visibility etc. in camera focus, shooting distance, camera, therefore, how reasonably to use these parameters to form stereoscopic camera QA system, just become research emphasis of the present invention to instruct stereoscopic shooting.
Stereo image quality evaluation method can be divided into subjective evaluation method and method for objectively evaluating.Subjective evaluation method is dominated by people, and namely the information that comprises of human eye subjective perception stereo-picture, judges the quality of stereo-picture with this; Method for objectively evaluating refers to some feature in the process produced according to stereo-picture itself or stereo-picture, thus calculates and derive and the evaluation points that draws and formula, can formation algorithm or model, and then by programming realization.Quality due to stereo-picture is all finally judged by human eye, so subjective evaluation method is the most reliable method.But subjective assessment complex steps, expend time in length, and the individual difference of subject (people) is large, easily there is fluctuation in subjective scores, and method for objectively evaluating efficiency is higher, operability is better, the more important thing is and can realize automatic Evaluation system, so the range of application of subjective evaluation method is extensive not as method for objectively evaluating.If obtain evaluation index comparatively accurately, subjective assessment is absolutely necessary, therefore, the present invention is on the basis proposing the parallel shooting criterion of stereoscopic camera, also deeper research is carried out further in conjunction with existing three-dimensional disturbance, utilize subjective experiment to verify, thus draw the shooting standard with universality.
Summary of the invention
Technical problem to be solved by this invention is: propose the parallel shooting method for objectively evaluating of a kind of stereoscopic camera, effectively can evaluate the stereo image quality of the parallel shooting of stereoscopic camera, and can provide shooting criterion for stereoscopic shooting amateurs.Technical solution of the present invention is as follows:
The parallel shooting quality method for objectively evaluating of a kind of stereoscopic camera, involved symbol and implication as shown in the table:
The method comprises and comprising the following steps:
1) set subjective assessment value as MOS, determine the evaluation standards of grading of stereoscopic camera shooting quality;
2) according to subjective experiment result, three-dimensional shape distortion μ when choosing stereoscopic camera parallel shooting pan objective evaluation index as stereo pairs:
μ p = L s · tan ( β / 2 ) f · tan ( α / 2 )
3) according to subjective experiment result, L is set up bwith d ccorrelation k p, with k pas the comprehensive interactional objective evaluation index analyzing shooting distance and camera spacing:
k p=d c-0.000054·L b 1.6
4) stereoscopic camera is utilized to take stereo pairs in the situation of be arrangeding in parallel;
5) according to captured stereo pairs, camera shooting distance L when stereoscopic camera is parallel to be taken is extracted b, camera spacing d cwith camera focus f;
6) the objective evaluation index μ of stereo-picture is calculated pand k p, form a three-dimensional perception ε of final index p, further by ε pbe combined with camera focus f;
ϵ p = μ p · f · k p 100 · d e = L s · tan ( α / 2 ) · ( d c - 0.000054 · L b 1.6 ) 100 · d e · tan ( β / 2 )
ε p=ρ·f -1.9-20
7) final objective evaluation index is obtained---ρ is the objective comprehensive evaluation index of parallel camera, and parallel camera objective comprehensive evaluation index ρ value more hour, illustrates that the parallel shooting quality of this stereoscopic camera is better:
ρ = ( L s · tan ( α / 2 ) · ( d c - 0.000054 · L b 1.6 ) 100 · d e · tan ( β / 2 ) + 20 ) · f 1.9 .
The present invention's advantage is compared with prior art: the present invention utilizes double vision point stereo pairs to extract the inner parameter of stereoscopic camera, external parameter and scenario parameters, by subjective detection test stereoscopic image to making subjective assessment, based on the feature of the parallel shooting of stereoscopic camera and the requirement to acquisition parameters, the three-dimensional disturbance of abundant research is on the impact of three-dimensional shooting effect, set up the parallel shooting criterion of a kind of new stereoscopic camera, the stereoeffect of stereo-picture and picture quality when the method can evaluate the parallel shooting of stereoscopic camera effectively, and take criterion accordingly for the amateur of stereoscopic shooting provides.
Accompanying drawing explanation
Describe embodiments of the present invention in detail by referring to accompanying drawing, above feature and advantage of the present invention will be easy to become obvious for those skilled in the art, wherein:
Fig. 1 is the parallel schematic diagram of stereoscopic camera, and (a) takes schematic diagram, and (b) watches schematic perspective view;
Fig. 2 is stereoscopic camera parallel shooting quality objective evaluation model;
Fig. 3 is L bwith d ccorrelation schematic diagram;
Fig. 4 is ε pwith f correlation schematic diagram.
Embodiment
In further detail embodiments of the present invention are described hereinafter with reference to accompanying drawing now.But, specifically can implement the present invention by different forms and the present invention should be interpreted as be limited to execution mode set forth herein.More precisely, provide these execution modes to be comprehensive and complete in order to expose, and idea of the present invention will be passed on to those skilled in the art all sidedly.
The present invention only studies for the parallel shooting of stereoscopic camera, in conjunction with the feature (as Fig. 1) when the shooting of parallel camera and stereo-picture viewing and existing shooting criterion, probe into and extend multiple stereoscopic camera acquisition parameters to the impact of picture quality, and establish corresponding model.Simultaneously in conjunction with the basis of the restriction of the scenario parameters of parallel shooting and the analysis and research of three-dimensional disturbance, propose the parallel shooting criterion of new a kind of stereoscopic camera.
For the parallel shooting of stereoscopic camera, the present invention is based on three-dimensional disturbance and establish a kind of new stereoscopic camera parallel shooting criterion model with subjective assessment, as shown in Figure 2.First, the acquisition parameters of stereoscopic camera is obtained from stereo image pair, by parameters input to parallel camera shooting quality evaluation module, existing shooting criterion, the evaluation index of effectively avoiding the rule of three-dimensional disturbance and the present invention to propose are combined, and export corresponding objective evaluation value, by the contrast of this objective evaluation value and subjective assessment value MOS, can learn that shooting criterion that the present invention proposes is applicable to the evaluation to three-dimensional camera parallel shooting quality.Concrete implementation step is as follows:
(1) stereoscopic camera acquisition parameters is obtained from stereogram image source
Stereoscopic camera parameter extraction acceptance of the bid earnest method mainly at world coordinate system, set up transformational relation model between camera coordinates system and image coordinate system.Have the method for multiple extracting parameter in prior art, the present invention adopts vanishing point method to extract shooting distance L b, camera focus f and camera spacing d c.The specific implementation of method can see two sections of documents below:
[1]JA.deFranca,MR.Stemmer,MBD.Franca,Anewrobustalgorithmicformulti-cameracalibrationwitha1Dobjectunder[J].PatternRecognition,45(10),3636-3647,2012.
[2]E.Guillou,D.Meneveaux,E.Maisel,andK.Bouatouch,Usingvanishingpointsforcameracalibrationandcoarse3Dreconstructionfromasingleimage[J].VisualComputer,16(7),396-410,2000.
The present invention extracts stereoscopic camera inside and outside parameter by hyperspace sampling analytic method founding mathematical models and optimal estimation parameter model two steps, and the area-of-interest exacting method of combining image visual stimulus, and extract scenario parameters with this.The method adopts interactively mode, top-down information is converted to the different weighted values of input picture low-level image feature, simultaneously in conjunction with Itti-Koch model and Stentiford model, realizes the area-of-interest exacting method of combining image visual stimulus.
(2) the present invention is based on forefathers to part camera parameter and the research of stereoeffect relation, according to achievement in research and conclusions such as Percival comfort zone, A Enshacikaya stereo formula and Bercovitz formula, in conjunction with human eye corresponding circle of sensation comfort zone, by subjectivity detection experiment, correlation between the parallel acquisition parameters of research stereoscopic camera, based on three-dimensional disturbance and subjective assessment analysis and summary, the parallel shooting criterion of a kind of new stereoscopic camera of proposition.The symbol related in formula and implication as shown in table 1.
Symbol in table 1 formula and implication thereof
Because objective evaluation model proposed by the invention is set up on a large amount of subjective experiment basis, therefore, the present invention carries out subjective experiment by according to Pyatyi evaluation criterion (table 2) general at present.And the output of objective evaluation model method subjective evaluation result and the present invention proposed contrasts, and verifies exploitativeness of the present invention with this.
Table 2 stereoscopic camera close-up photography quality evaluation standards of grading
Will from following three aspect researchs this shooting criterion:
(2.1) three-dimensional shape distortion
Weigh stereo-picture distortion level directly and effective way is width magnification ratio and degree of depth magnification ratio [3].Width magnification ratio (M w) refer to the ratio of wide (or high) of 3D (three-dimensional, the 3D) object of reproduction and original 3D object wide (or high).In like manner, degree of depth magnification ratio (M d) refer to the ratio of the degree of depth of the 3D object of reproduction and the degree of depth of original 3D object.The people such as Zilly propose three-dimensional shape distortion be a kind of simple means of interpretation to show the distortion degree of stereo-picture, three-dimensional shape distortion can be expressed as the ratio of width magnification ratio and degree of depth magnification ratio, shown in (1).
μ = M w M d = L s · d c L c · ξ - L b · ( ξ - d e )
Wherein, ξ = a 1 · f · d c L c - - - ( 1 )
A in formula 1refer to screen magnifying rate, the ratio of the width of the virtual screen felt when namely the width of screen and observer watch image, formula is as follows:
a 1 = tan ( β / 2 ) tan ( α / 2 ) - - - ( 2 )
By the design standard (as shown in Figure 1) of parallel camera, camera skew (Sensorshift) is the side-play amount of the convergence from the center of image to lens axis:
S = f · d c 2 · L c - - - ( 3 )
Under parallel camera configuration, when undistorted stereo scene is presented near stereoscopic screen, experimentally condition, now S=d e/ (2a 1), then formula (3) can be reduced to formula (4), and subjective results is as shown in table 3.
μ p = L s a 1 · f = L s · tan ( β / 2 ) f · tan ( α / 2 ) - - - ( 4 )
Table 3 μ pwith the corresponding relation of MOS
[3]ZillyF,KlugerJ,KauffP.Productionrulesforstereoacquisition[J].ProceedingsoftheIEEE,2011,99(4):590-606.
(2.2) shooting formula---the L of the present invention's proposition b, d crelational expression
So far, the basis shooting formula about stereoscopic camera shooting has a lot, but seldom can clearly express shooting distance L bwith camera spacing d caffect the corresponding relation of shooting quality.By a series of subjective and objective experiment, first by L bas fixed value, choose d cscope is from 10mm to 400mm, and stepping is 10mm; Secondly, L is changed bvalue, d cdo equally and change.According to subjectivity detection experiment, we obtain L bwith d ccorrelation schematic diagram, as shown in Figure 3.
So, L bwith d ccorrelation formula such as formula shown in (5).
d c=0.000054·L b 1.6+k p(5)
In figure 3, k prepresent constant, have after result of the test matching: k p" d in=63.2 representative graphs 3 cvsL b(1) ", the line of demarcation of representative " MOS=5 " and " MOS=4 ", k p" d in=102.5 representative graphs 3 cvsL b(2) ", the line of demarcation of representative " MOS=4 " and " MOS=3 ", k p" d in=135.7 representative graphs 3 cvsL b(3) ", the line of demarcation of representative " MOS=3 " and " MOS=2 ", k p" d in=159.1 representative graphs 3 cvsL b(4) line of demarcation of " MOS=2 " and " MOS=1 " ", is represented.
Formula (5) is converted to following form:
k p=d c-0.000054·L b 1.6(6)
Subjective experiment result is as shown in table 4.
Table 4k pwith the corresponding relation of MOS
(3) in order to the parameter in stereoscopic camera camera system can be controlled, two parts research contents in step (2) is arranged and combines, form a three-dimensional perception ε of final index p(as shown in Equation 7), μ pfor three-dimensional shape distortion, k pfor the L that the present invention proposes b, d crelational expression.
ϵ p = μ p · f · k p 100 · d e = L s · tan ( α / 2 ) · ( d c - 0.000054 · L b 1.6 ) 100 · d e · tan ( β / 2 ) - - - ( 7 )
In order to enrich index ε further p, the present invention has carried out again the research about camera focus f, and experimental result as shown in Figure 4.Can obtain thus, index ε pformula (8) can be had to represent with the relational expression of f.
ε p=ρ·f -1.9-20(8)
Evaluate for parallel camera shooting quality, ρ is the objective comprehensive evaluation index of parallel camera (as shown in Equation 9).
ρ = ( L s · tan ( α / 2 ) · ( d c - 0.000054 · L b 1.6 ) 100 · d e · tan ( β / 2 ) + 20 ) · f 1.9 - - - ( 9 )
From formula 9 and Fig. 4, the value of the objective indicator ρ finally proposed as the present invention more hour, illustrates that the parallel shooting quality of this stereoscopic camera is better.
The objective evaluation model of foundation proposed by the invention not only achieves the evaluation to three-dimensional camera parallel shooting quality, but also may be used for instructing the parallel shooting of stereoscopic camera low coverage, for amateur provides the foundation that rationally arranges acquisition parameters.
The content be not described in detail in specification of the present invention belongs to the known prior art of professional and technical personnel in the field.

Claims (1)

1. based on the parallel shooting method for objectively evaluating of stereoscopic camera of three-dimensional disturbance, involved symbol and implication as shown in the table:
The method comprises the following steps:
1) set subjective assessment value as MOS, determine the evaluation standards of grading of stereoscopic camera shooting quality;
2) according to objective and subjective experiment result, when choosing stereoscopic camera parallel shooting, three-dimensional shape distorts μ pan objective evaluation index as stereo pairs:
3) according to subjective experiment result, L is set up bwith d ccorrelation k p, with k pas the comprehensive interactional objective evaluation index analyzing shooting distance and camera spacing:
k p=d c-0.000054·L b 1.6
4) stereoscopic camera is utilized to take stereo pairs in the situation of be arrangeding in parallel;
5) according to captured stereo pairs, camera shooting distance L when stereoscopic camera is parallel to be taken is extracted b, camera spacing d cwith camera focus f;
6) the objective evaluation index μ of stereo-picture is calculated pand k p, form a three-dimensional perception ε of final index p, further by ε pbe combined with camera focus f;
ε p=ρ·f -1.9-20
7) final objective evaluation index is obtained---ρ is the objective comprehensive evaluation index of parallel camera, and parallel camera objective comprehensive evaluation index ρ value more hour, illustrates that the parallel shooting quality of this stereoscopic camera is better:
CN201410625203.XA 2014-11-08 2014-11-08 The parallel shooting method for objectively evaluating of stereoscopic camera based on three-dimensional disturbance Active CN104333751B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201410625203.XA CN104333751B (en) 2014-11-08 2014-11-08 The parallel shooting method for objectively evaluating of stereoscopic camera based on three-dimensional disturbance

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201410625203.XA CN104333751B (en) 2014-11-08 2014-11-08 The parallel shooting method for objectively evaluating of stereoscopic camera based on three-dimensional disturbance

Publications (2)

Publication Number Publication Date
CN104333751A CN104333751A (en) 2015-02-04
CN104333751B true CN104333751B (en) 2016-04-13

Family

ID=52408390

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201410625203.XA Active CN104333751B (en) 2014-11-08 2014-11-08 The parallel shooting method for objectively evaluating of stereoscopic camera based on three-dimensional disturbance

Country Status (1)

Country Link
CN (1) CN104333751B (en)

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102209257B (en) * 2011-06-17 2013-11-20 宁波大学 Stereo image quality objective evaluation method
US9686526B2 (en) * 2011-10-04 2017-06-20 Telefonaktiebolaget L M Ericsson (Publ) Objective 3D video quality assessment model
CN104010188B (en) * 2014-05-26 2015-09-30 天津大学 The parallel shooting quality method for objectively evaluating of a kind of stereoscopic camera long distance
CN103997642B (en) * 2014-05-26 2016-01-06 天津大学 A kind of stereoscopic camera long distance assembles shooting quality method for objectively evaluating

Also Published As

Publication number Publication date
CN104333751A (en) 2015-02-04

Similar Documents

Publication Publication Date Title
CN107578403A (en) The stereo image quality evaluation method of binocular view fusion is instructed based on gradient information
CN104811693B (en) A kind of stereo image vision comfort level method for objectively evaluating
CN107578404A (en) The complete of view-based access control model notable feature extraction refers to objective evaluation method for quality of stereo images
CN107635136B (en) View-based access control model perception and binocular competition are without reference stereo image quality evaluation method
CN105654142B (en) Based on natural scene statistics without reference stereo image quality evaluation method
CN109167996A (en) It is a kind of based on convolutional neural networks without reference stereo image quality evaluation method
CN103873854B (en) The defining method of a kind of stereo-picture subjective assessment subject's quantity and experimental data
CN105407349A (en) No-reference objective three-dimensional image quality evaluation method based on binocular visual perception
CN109523513A (en) Based on the sparse stereo image quality evaluation method for rebuilding color fusion image
CN103618891B (en) Objective evaluation method of stereo camera microspur convergence shooting quality
CN108257125A (en) A kind of depth image quality based on natural scene statistics is without with reference to evaluation method
CN104361583B (en) A kind of method determining asymmetric distortion three-dimensional image objective quality
CN108337504A (en) A kind of method and device of evaluation video quality
CN104902268A (en) Non-reference three-dimensional image objective quality evaluation method based on local ternary pattern
CN104853182A (en) Amplitude and phase based stereo image quality objective evaluation method
CN103472668A (en) Stereo imaging device and method
CN103997642B (en) A kind of stereoscopic camera long distance assembles shooting quality method for objectively evaluating
CN105898279B (en) A kind of objective evaluation method for quality of stereo images
CN103780903B (en) A kind of stereoscopic camera low coverage assembles shooting quality method for objectively evaluating
CN103914835A (en) Non-reference quality evaluation method for fuzzy distortion three-dimensional images
CN103596001B (en) Method for objectively evaluating quality of parallel macrophotography of stereo camera
CN105491371A (en) Tone mapping image quality evaluation method based on gradient magnitude similarity
CN108377387A (en) Virtual reality method for evaluating video quality based on 3D convolutional neural networks
CN104333751B (en) The parallel shooting method for objectively evaluating of stereoscopic camera based on three-dimensional disturbance
CN109257591A (en) Based on rarefaction representation without reference stereoscopic video quality method for objectively evaluating

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant