Example: bankruptcy

ステレオマッチングによる物体 ... - topic.ad.jp

1 295 ( ) 295-7 Acquisition of an Object Shape Using Stereo Matching , Hirokazu Yoshida , Hiroyuki Kamaya * *National Institude of Technology, Hachinohe College : (Stereo Camera), (Triangulation), OpenCV(Open Source Computer Vision) : 039-1192 16-1 Tel.: 0178-27-7283 E-mail: _____1. 3D 2. x-y (Pxr,Pyr,Pzr) 2 (Px ,Py ,Pz ) h {Pxr=Px hPyr=Py Pzr=Pz (1) p f}

3 Fig.2 System configuration 3-2.ハードウェア ステレオカメラには、安価な Ovrvision1[1]を使用した。このカメラの 性能を以下に示す。

Tags:

  Topics

Information

Domain:

Source:

Link to this page:

Please notify us if you found a problem with this document:

Other abuse

Transcription of ステレオマッチングによる物体 ... - topic.ad.jp

1 1 295 ( ) 295-7 Acquisition of an Object Shape Using Stereo Matching , Hirokazu Yoshida , Hiroyuki Kamaya * *National Institude of Technology, Hachinohe College : (Stereo Camera), (Triangulation), OpenCV(Open Source Computer Vision) : 039-1192 16-1 Tel.: 0178-27-7283 E-mail: _____1. 3D 2. x-y (Pxr,Pyr,Pzr) 2 (Px ,Py ,Pz ) h {Pxr=Px hPyr=Py Pzr=Pz (1) p f p (x ,y ) (xr,yr) {x =z Px Pz =fPx Pz y =z Py Pz =fPy Pz (2) {xr=zrPxrPzr=fPxrPzryr=zrPyrPzr=fPyrPzr (3) (1) (3) {xr=fPx hPz =fPx Pz fhPz =x fhPz yr=fPy Pz =y (4) (4) P =fhx xr (5)}}}}

2 X x x (a) (b) The principle of stereo camera 3. 3-1. 1 1 SONY VAIO (Windows7(64bitOS),Intel CORE i5-2430M, 4GB, ) Xl 3 System configuration 3-2. Ovrvision1[1] 1280 480 640 480 2 60fps 50mm 86 36 135 3-3. Microsoft Visual C++ 2010 Express OpenCV [2] Ovrvision SDK Version [1] 4. 4-1. Y (5) Table 1 [3] Table 1 Parameter f 1 2 p1 p2 3 (u0,v0) R t Chessboard USB 4 4-2.

3 Before paralleled After paralleled 5. 5-1. OpenCV [4] [fx0u00fyv0001]=[ 0 0 1] [fx0u00fyv0001]=[ 0 0 0 0 1] [fx0u00fyv0001]=[ 0 0 0 0 1] 150mm 450mm 50mm 100[mm] Target object and marker 5-2. 1.

4 5 2. 3. 4. 5. x y 6. 5 x (5) 7. Comparison result (red marker) Error rate(red marker) Comparison result (yellow marker) Error rate (yellow marker) 100200300400500150250350450 Distance[mm]True Distance[mm]Theoretical valueExperimental value-10-50510150250350450 Error rate[%]True Distance[mm]100200300400500150250350450 Ditance[mm]True Distance[mm]Theoretical valueExperimental value-10-50510150250350450 Error rate[%]True Distance[mm]6 5-3. RGB {R =R GB =B RY =G BG =G R 6 (R red, B:blue, G:green, Y:yellow) 100 1.}

5 2. 3. 4. 100 5. 6. 7. 8. x y 9. 8 x (5) 10. Comparison result (red marker) Error rate (red marker) 6. 6 Table 2 100200300400500150250350450 Distance[mm]True Distance[mm]Theoretical valueExperimantal value-10-50510150250350450 Error rate[%]Ture Distance[mm]7 Table 2 Acquisition of an Object Shape Color X[mm] Y[mm] Z[mm] Red Yellow Blue Green [ms] Measured object shape 7.

6 [ ] 8. 4 Y [1] .com, (2015 ) [2] , (2015 6 ) [3] (2000) [4] Gray Bradski and Adrian Kaebler OpenCV (2009)


Related search queries