提高測量精度的方法

1、Improving the accuracy ofphase-shifting techniques金 成 英2016-01-08Page 2 Contents Conclusions4 The steps of the proposed method2 Experimental evaluation3 1 The principle of the proposed method Page 3Enter your subtitle or main authors name hereInsert the title of your presentation herePrincipleDigita

2、l fringe projection system Reference:Song Zhang,”Rencent progress on real-time 3D shape measurement using digital fringe projection techniques.”Page 4Phase-shifting algorithms( ,)cos( ,)(1)kkIx yIIx yThe intensities of the phase-shifting image at every pixel are given by , x yEq. (1) can be rewritte

3、n as ( ,)cos,cos()sin,sin(2)kkkIx yIIx yIx y012,( ,) cos()( ,) sin()(3)kkkIx yx yx yx yorPage 5)cos(),(, 10kIyxIyx)sin(),( 2kIyx21210)sin(),()cos(),(),(,NkkkkyxyxyxyxIEwhereand0121001211012122( , )( , )( , )cos()( , )sin()02( , )( , )( , )cos()( , )sin() cos02( , ),cos,sinsin0NkkkkNkkkkkNkkkkkEIx yx

4、 yx yx yEIx yx yx yx yEIx yx yx yx y Page 6NkkkNkkkNkkNkkNkkkNkkNkkkNkkNkkNkkNkkIIIyxyxyxN1112101211112111sincos),(),(),(sinsincossinsincoscoscossincos),()(),(),(),()(1kkkkyxBAyxyxByxA2;1,kk NkN)sinsin(tan),(),(tan,111121NkkkNkkkIIyxyxyxReference:”O(jiān)ptical shop testing.”Page 7Camera-projector calibra

5、tion00, ,1,y ,z ,10001smA R t MTTwwwmu vMxuAvCamera calibrationReference:Zhengyou Zhang,”A flexible new technique for camera calibration.”Page 8Projection calibrationReference:Zhang and Huang,”Novel method for structured light system calibration.” The key to projection calibration is to establish th

6、e correspondence between the camera pixel and the projection pixel. Page 9Reference: Z.Li et al.,”Accurate calibration method for a structured light system.” Page 10Correspond between the CCD image and the DMD image.Image in the first row from left to right are horizontal CCDfringe images,CCD center

7、line image,and DMD fringe image.The second row shows the corresponding images for verticalFringe image.11100000sin,tan()sin(,)( ,)( ,),NkkkNkkkNnnIx yIxyx yx yN Calculate the absolute phase: Convert the absolute phase into theprojector image coordinate: Thus the corresponding projector pixel can be

8、obtained.,2222hvpx ypx yHWxy,x yPage 11Enter your subtitle or main authors name hereCCD image and its corresponding DMD image: (a)CCD image, (b) horizontal absolute phase map, (c) vertical absolute phase map, (d) DMD image. Page 12StepsFundamental stages for the camera-projector calibrationand three

9、-dimensional reconstruction.Page 13Reference:SUSAN-a new approach to low level image processing. 600000000001,10,2,0I rI rtrif I rI rtc r rif I rI rtc r ren rc r rgn rif n rgR rotherwise The SUSAN principle for feature detectionPage 14Enter your subtitle or main authors name hereThe steps of interpo

10、lation of feature points: 1.An image of the object is captured onto the scene.2.An initial set of corners is detected using a SUSAN (smallest univalue segment assimilating nucleus ) corner detector.3.A refinement algorithm is applied to the initial set of corners,and obtaina set of corners at subpix

11、el accuracy.4.The corresponding projector positions are computed for the set of detectedcorners using phase interpolation.Page 15Enter your subtitle or main authors name hereExperimental Evaluation Its not difficult to find that linear interpolation of the detected corners inthe camera gives the bes

12、t result for the calibration of the projector,whereasnearest interpolation of the detected corners in the camera for projector calibration gives the worst result. Page 16Page 17Page 18 The proposed method using linear interpolation of feature points has better accuracy than the traditional method,wh

13、ereas using nearest-neighbor interpolation of feature points has the worst accuracy overall the cases.Page 19Enter your subtitle or main authors name hereConclusions The proposed method improves the accuracy of traditional phase-shifting technique. It is based on the detection of a set of featurepoints in the image space at subpixel accuracy,and phase interpolation in order to generate itscorrespondin