블로그 이미지
Leeway is... the freedom that someone has to take the action they want to or to change their plans.
maetel

Notice

Recent Post

Recent Comment

Recent Trackback

Archive

calendar

 « 2025.2  »
일 월 화 수 목 금 토
1
2 3 4 5 6 7 8
9 10 11 12 13 14 15
16 17 18 19 20 21 22
23 24 25 26 27 28
  • total
  • today
  • yesterday

Category

2010ETRIcg: report 0915

2010. 9. 16. 00:40

보호되어 있는 글입니다.
내용을 보시려면 비밀번호를 입력하세요.

2010ETRIcg: report 0916

2010. 9. 15. 21:33

보호되어 있는 글입니다.
내용을 보시려면 비밀번호를 입력하세요.

2010ETRIcg: report 0914

2010. 9. 14. 22:32

보호되어 있는 글입니다.
내용을 보시려면 비밀번호를 입력하세요.

2010ETRIcg: report 0906

2010. 9. 6. 23:28

보호되어 있는 글입니다.
내용을 보시려면 비밀번호를 입력하세요.

2010ETRIcg: report 0904

2010. 9. 4. 05:36

보호되어 있는 글입니다.
내용을 보시려면 비밀번호를 입력하세요.

Masakazu Suzuki "Evapotranspiration Estimates of Forested Watersheds in Japan Using the Short-time Period Water-budget Methods"

2010. 9. 3. 16:29 Computer Vision/2010ETRI-MultiAutoCalib
Suzuki, M. (1985) Evapotranspiration Estimates of Forested Watersheds in Japan Using the Short-time Period Water-budget Methods. Journal of Japanese Forest Society, 67: 115-125. (in Japanese with English summary)
masakazuSuzuki_1985jjfs.pdf


短期水収支法による森林流域からの蒸発散量推定
Evapotranspiration estimates of forested watersheds in Japan using the short-time period water-budget method  [in Japanese]   
鈴木 雅一 (SUZUKI Masakazu)         
京都大学農学部 (Fac.of Agr., Kyoto Univ.)

長期間にわたって水収支観測がなされている日本各地の森 林流域の記録をもとに, 短期水収支法を用いて流域の蒸発散量とその季節変化を求めた。検討には桐生, 川向, 竜の口山, 釜淵, 去川の5試験地, 9流域のそれぞれ10年から40年間の日雨量, 日流出量記録が用いられた。短期水収支法では, 渇水による蒸発散低下は水収支期間内の最小流量に対応して生じ, 蒸発散低下をもたらす限界流量が流域ごとに定められた。渇水による蒸発低下の例を除外して求めた蒸発散量季節変化は, 植生が著しく変わらないとき集計期間が異なってもほぼ同様の結果となった。森林の伐採や山火事によって蒸発散量が減少する傾向は各流域とも同様であるが, その変化が通年にわたり生じた流域とおもに夏期に生じた流域があった。求められた蒸発散量とその季節変化は各流域の気象, 植生を反映する値として, 森林流域の蒸発散量推定式作成の基礎資料になるといえる。

The annual and monthly evapotranspiration were estimated using the method of the short-time period water-budget on 9 water-sheds located in 5 experimental areas in Japan. Daily precipitation and discharge records and a period of 10 to 40 years for each watershed were used in this estimation. The appearance of an evapotranspiration decline because of drought has a relationship with the minimum discharge-rate in a water-budget period. The critical discharge-rate for an evapotranspiration decline can be determined for each watershed. On a watershed without a remarkable change of vegetation, seasonal variations of different averaging periods of years for evapotranspiration under no drought conditions change in the same way. Evapotranspiration decreases after clear-cuttings and forest fires on every watershed where such events have occurred, but the tendency for changes in seasonal variations vary with each watershed.


저작자표시 비영리 동일조건

'Computer Vision > 2010ETRI-MultiAutoCalib' 카테고리의 다른 글

2010ETRIcg: report 0906  (0) 2010.09.06
2010ETRIcg: report 0904  (0) 2010.09.04
OpenCV: cvFindContours( )  (0) 2010.09.03
Ferenc Kahlesz & Cornelius Lilge & Reinhard Klein "Easy–to–Use Calibration of Multiple–Camera Setups"  (0) 2010.07.20
case study: 리프로젝션 에러 실험 결과 조사  (0) 2010.07.20
posted by maetel

OpenCV: cvFindContours( )

2010. 9. 3. 16:23 Computer Vision/2010ETRI-MultiAutoCalib
OpenCV: cvFindContours( )

cvFindContours()
int cvFindContours(CvArr* image, CvMemStorage* storage, CvSeq** first_contour, int header_size=sizeof(CvContour), int mode=CV_RETR_LIST, int method=CV_CHAIN_APPROX_SIMPLE, CvPoint offset=cvPoint(0, 0))

Finds the contours in a binary image.

Parameters:
  • image – The source, an 8-bit single channel image. Non-zero pixels are treated as 1’s, zero pixels remain 0’s - the image is treated as binary . To get such a binary image from grayscale, one may use Threshold , AdaptiveThreshold or Canny . The function modifies the source image’s content
  • storage – Container of the retrieved contours
  • first_contour – Output parameter, will contain the pointer to the first outer contour
  • header_size – Size of the sequence header, \ge \texttt{sizeof(CvChain)} if \texttt{method} =\texttt{CV\_CHAIN\_CODE} , and \ge \texttt{sizeof(CvContour)} otherwise
  • mode –

    Retrieval mode

    • CV_RETR_EXTERNAL retrives only the extreme outer contours
    • CV_RETR_LIST retrieves all of the contours and puts them in the list
    • CV_RETR_CCOMP retrieves all of the contours and organizes them into a two-level hierarchy: on the top level are the external boundaries of the components, on the second level are the boundaries of the holes
    • CV_RETR_TREE retrieves all of the contours and reconstructs the full hierarchy of nested contours
  • method –

    Approximation method (for all the modes, except CV_LINK_RUNS , which uses built-in approximation)

    • CV_CHAIN_CODE outputs contours in the Freeman chain code. All other methods output polygons (sequences of vertices)
    • CV_CHAIN_APPROX_NONE translates all of the points from the chain code into points
    • CV_CHAIN_APPROX_SIMPLE compresses horizontal, vertical, and diagonal segments and leaves only their end points
    • CV_CHAIN_APPROX_TC89_L1,CV_CHAIN_APPROX_TC89_KCOS applies one of the flavors of the Teh-Chin chain approximation algorithm.
    • CV_LINK_RUNS uses a completely different contour retrieval algorithm by linking horizontal segments of 1’s. Only the CV_RETR_LIST retrieval mode can be used with this method.
  • offset – Offset, by which every contour point is shifted. This is useful if the contours are extracted from the image ROI and then they should be analyzed in the whole image context



Learning OpenCV: Chater 8. Contours: Contour Finding
: 234p
"the concept of a contour tree"


Suzuki, M. (1985) Evapotranspiration Estimates of Forested Watersheds in Japan Using the Short-time Period Water-budget Methods. Journal of Japanese Forest Society, 67: 115-125. (in Japanese with English summary)


저작자표시 비영리 동일조건

'Computer Vision > 2010ETRI-MultiAutoCalib' 카테고리의 다른 글

2010ETRIcg: report 0904  (0) 2010.09.04
Masakazu Suzuki "Evapotranspiration Estimates of Forested Watersheds in Japan Using the Short-time Period Water-budget Methods"  (0) 2010.09.03
Ferenc Kahlesz & Cornelius Lilge & Reinhard Klein "Easy–to–Use Calibration of Multiple–Camera Setups"  (0) 2010.07.20
case study: 리프로젝션 에러 실험 결과 조사  (0) 2010.07.20
J. Weng, P. Cohen, and M. Herniou "Camera calibration with distortion models and accuracy evaluation"  (0) 2010.07.16
posted by maetel

Ferenc Kahlesz & Cornelius Lilge & Reinhard Klein "Easy–to–Use Calibration of Multiple–Camera Setups"

2010. 7. 20. 22:36 Computer Vision/2010ETRI-MultiAutoCalib
Easy–to–Use Calibration of Multiple–Camera Setups
ferencKahlesz_2007icvs.pdf
Ferenc Kahlesz, Cornelius Lilge & Reinhard Klein
University of Bonn, Institute of Computer Science II, Computer Graphics Group
http://biecoll.ub.uni-bielefeld.de
Proceedings of the ICVS(International Conference on Computer Vision Systems)  Workshop on Camera Calibration Methods for Computer Vision Systems - CCMVS2007 Published in 2007 by Applied Computer Science Group, Bielefeld University, Germany


저작자표시 비영리 동일조건

'Computer Vision > 2010ETRI-MultiAutoCalib' 카테고리의 다른 글

Masakazu Suzuki "Evapotranspiration Estimates of Forested Watersheds in Japan Using the Short-time Period Water-budget Methods"  (0) 2010.09.03
OpenCV: cvFindContours( )  (0) 2010.09.03
case study: 리프로젝션 에러 실험 결과 조사  (0) 2010.07.20
J. Weng, P. Cohen, and M. Herniou "Camera calibration with distortion models and accuracy evaluation"  (0) 2010.07.16
DLR Camera Calibration Toolbox  (0) 2010.07.16
posted by maetel

case study: 리프로젝션 에러 실험 결과 조사

2010. 7. 20. 15:07 Computer Vision/2010ETRI-MultiAutoCalib
1. 실험실 환경에서의 결과 
In the case of experiments in a laboratory:  

(1) Yasutaka Furukawa & Jean Ponce, "Accurate Camera Calibration from Multi-View Stereo and Bundle Adjustment", CVPR 2008 
(* 영상의 크기와 리프로젝션 에러 사이의 관계를 수학적으로 설명)
For example, the robot arm (Stanford spherical gantry) used in the multi-view stereo evaluation of [18] has an accuracy of 0.01◦ for a 1m radius sphere observing an object about 15cm in diameter, which yields approximately 1.0[m]×0.01×π/180 = 0.175[mm] errors near an object. Even with the low-resolution 640×480 cameras used in [18], where a pixel covers roughly 0.25mm on the surface of an object, this error corresponds to 0.175/0.25 = 0.7pixels, which is not negligible. If one used a highresolution 4000× 3000 camera, the positioning error would increase to 0.7×4000/640 = 4.4pixels.  

The mean reprojection error decreases from 2-3pixels before refinement to about 0.25 to 0.5 pixels for most datasets after six iterations.  

In practice, we have found PMVS to be robust to errors in camera parameters as long as the image resolution matches the corresponding reprojection errors—that is, when features to be matched are roughly within two pixels of the corresponding 3D points.  
  




(2) Ivo Ihrke & Lukas Ahrenberg & Marcus Magnor, “External camera calibration for synchronized multi-video systems”, Journal of WSCG(the 12th International Conference in Central Europe on Computer Graphics, Visualization and Computer Vision), 2004                    




(3) Cheng-I Chen & Yong-Sheng Chen "Central Catadioptric Camera Calibration using Planar Objects", ICVS 2007





(4) Seitz, S. M., Curless, B., Diebel, J., Scharstein, D., and Szeliski, R. 2006. A Comparison and Evaluation of Multi-View Stereo Reconstruction Algorithms, CVPR 2006
: S. Seitz et al. Multi-view stereo evaluation web page at http://vision.middlebury.edu/mview/

Calibration accuracy on these datasets appears to be on the order of a pixel (a pixel spans about 1/4mm on the object). It is difficult to quantify the calibration accuracy because we don't have point correspondences in all views. However, when the images are reprojected onto the laser scanned mesh reconstruction and averaged, 1-2 pixel wide features are clearly visible. Some images and regions of the object are a bit out of focus, due in part to the limited depth of field afforded by the imaging configuration. This is probably a blessing in disguise, as it helps compensate for the lack of sub-pixel registration. While there is certainly room for improvement, we expect that this degree of accuracy will enable good results for most algorithms. Note that the images have been corrected to remove radial distortion.




2. 일반/극한 환경 (실내/실외)에서의 결과 
In the case of experiments in an exterior, interior or extreme circumstance:   



3. 리프로젝션 에러 최소화 관련 방법과 용어 및 문헌
methods or related terms to be used to minimize reprojection errors  

(1) sampson error

(2) LM(Levenberg-marquardt) algorithm
M. Lourakis. levmar: Levenberg-marquardt nonlinear least squares algorithms in C/C++. [web page] http://www.ics.forth.gr/~lourakis/levmar, July 2004.

(3) bundle adjustment

(4) iterative gradient descent (in Bouget's toolbox)


ref.
1) Gregorij Kurillo &  Zeyu Li & Ruzena Bajcsy, "Framework for Hierarchical Calibration of Multi-camera Systems for Teleimmersion", ICIT 2009


Figure 1 shows typical error distribution as obtained on the four cameras within the cluster. After the global optimization on all four cameras, the combined error distribution resembles Gaussian distribution with the mean value of 0.153 pixels and standard deviation of 0.091 pixels. Note that the reprojection error on the color camera (#4) is higher than on the grayscale cameras. The maximal error for this set of images was 0.8 though only a small number of points had errors in that range.

Bundle adjustment minimizes the following reprojection error:


Figure 4 shows the reprojection error on all cameras. The cameras whose position and orientation were obtained by indirect transformation path with the reference camera had no significantly different reprojection errors as compared to the cameras calibrated directly with the reference camera. The mean reprojection error between all the cameras was 0.3633 pixels with the standard deviation of 0.0486 pixels.

2) Kai Ide & Steffen Siering & Thomas Sikora, "Automating Multi-Camera Self-Calibration", WACV 2009

In practice the reprojection error is usually larger at image boundaries as radial distortions due to the optical system used in the cameras become more apparent. Somewhat surprisingly this is also true for the optics of the projector even though its high quality object lens should show no signs of radial distortion when projecting rectangular images onto a screen.

The reprojection error for our setup consisting of four cameras and one projector is shown shown in fig. 9. The total mean reprojection error is 0.29 pixels with a standard deviation of 0.20 pixels. It is important to mention that this result, despite being satisfactory, is not necessarily in itself meaningful as one has to compare this value with the distribution of points within the image space that have survived the RANSAC verification. An even spread throughout the image plane is desirable, a property that has been demonstrated in fig. 8. As we decrease radius and spacing to r = s = 1 we get results similar to the unmodified calibration sequence, yielding a relatively large reprojection error of 0.8 pixels in average.

저작자표시 비영리 동일조건

'Computer Vision > 2010ETRI-MultiAutoCalib' 카테고리의 다른 글

OpenCV: cvFindContours( )  (0) 2010.09.03
Ferenc Kahlesz & Cornelius Lilge & Reinhard Klein "Easy–to–Use Calibration of Multiple–Camera Setups"  (0) 2010.07.20
J. Weng, P. Cohen, and M. Herniou "Camera calibration with distortion models and accuracy evaluation"  (0) 2010.07.16
DLR Camera Calibration Toolbox  (0) 2010.07.16
Test: reprojection error (Quickcam Pro4000)  (0) 2010.07.07
posted by maetel

J. Weng, P. Cohen, and M. Herniou "Camera calibration with distortion models and accuracy evaluation"

2010. 7. 16. 16:52 Computer Vision/2010ETRI-MultiAutoCalib
J. Weng, P. Cohen, and M. Herniou. "Camera calibration with distortion models and accuracy evaluation." In IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI), 14(10): 965-980, 1992.
juyangWeng_1992pami.pdf


저작자표시 비영리 동일조건

'Computer Vision > 2010ETRI-MultiAutoCalib' 카테고리의 다른 글

Ferenc Kahlesz & Cornelius Lilge & Reinhard Klein "Easy–to–Use Calibration of Multiple–Camera Setups"  (0) 2010.07.20
case study: 리프로젝션 에러 실험 결과 조사  (0) 2010.07.20
DLR Camera Calibration Toolbox  (0) 2010.07.16
Test: reprojection error (Quickcam Pro4000)  (0) 2010.07.07
Tomas Svoboda <Multi-Camera Self-Calibration>  (0) 2010.07.07
posted by maetel

DLR Camera Calibration Toolbox

2010. 7. 16. 16:42 Computer Vision/2010ETRI-MultiAutoCalib
The DLR Camera Calibration Toolbox
http://www.dlr.de/rm/desktopdefault.aspx/tabid-3925/


Institute of Robotics and Mechatronics, German Aerospace Center (DLR)



> 특징
스테레오 (두 대 이상의 카메라)
자동 캘리브레이션(카메라 렌즈 왜곡, 내부/외부 파라미터)
Hand-Eye Calibration
수동 조작 가능


[1] K. H. Strobl and G. Hirzinger. "Optimal Hand-Eye Calibration." In Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2006), Beijing, China, pp. 4647-4653, October 2006.

[2] J. Weng, P. Cohen, and M. Herniou. "Camera calibration with distortion models and accuracy evaluation." In IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI), 14(10): 965-980, 1992.

[3] K. H. Strobl and G. Hirzinger. "More Accurate Camera and Hand-Eye Calibrations with Unknown Grid Pattern Dimensions." In Proceedings of the IEEE International Conference on Robotics and Automation (ICRA 2008), Pasadena, California, USA, May 2008, in press.


저작자표시 비영리 동일조건

'Computer Vision > 2010ETRI-MultiAutoCalib' 카테고리의 다른 글

Ferenc Kahlesz & Cornelius Lilge & Reinhard Klein "Easy–to–Use Calibration of Multiple–Camera Setups"  (0) 2010.07.20
case study: 리프로젝션 에러 실험 결과 조사  (0) 2010.07.20
J. Weng, P. Cohen, and M. Herniou "Camera calibration with distortion models and accuracy evaluation"  (0) 2010.07.16
Test: reprojection error (Quickcam Pro4000)  (0) 2010.07.07
Tomas Svoboda <Multi-Camera Self-Calibration>  (0) 2010.07.07
posted by maetel

Test: reprojection error (Quickcam Pro4000)

2010. 7. 7. 18:05 Computer Vision/2010ETRI-MultiAutoCalib



frame # 129  ---------------------------
# of found lines = 10 vertical, 8 horizontal
vertical lines:
horizontal lines:
p.size = 80
corner
35.6611, -6.40545  33.0973, 24.1258  32.2603, 34.0929  29.1089, 71.6216  26.9599, 97.2142  24.1377, 130.823  21.5627, 161.487  18.0768, 203  86.259, -2.68977  84.6836, 27.3713  84.1602, 37.3582  82.2186, 74.405  80.9047, 99.4752  79.1713, 132.552  77.5933, 162.661  75.4792, 203  108.14, -1.08291  106.889, 28.7684  106.47, 38.7618  104.926, 75.595  103.885, 100.438  102.508, 133.286  101.257, 163.156  99.5866, 203  155.983, 2.43046  155.675, 31.8377  155.57, 41.8509  155.189, 78.2292  154.934, 102.578  154.596, 134.923  154.288, 164.267  153.883, 203  168.044, 3.31615  168.351, 32.6352  168.456, 42.6616  168.836, 78.9444  169.09, 103.171  169.427, 135.389  169.733, 164.591  170.135, 203  202.166, 5.82193  202.773, 34.8009  202.983, 44.8339  203.736, 80.7734  204.236, 104.644  204.904, 136.504  205.508, 165.34  206.297, 203  216.478, 6.8729  217.688, 35.7392  218.109, 45.7855  219.61, 81.6053  220.604, 105.33  221.933, 137.039  223.135, 165.709  224.698, 203  246.815, 9.1007  248.312, 37.6659  248.838, 47.7188  250.7, 83.2347  251.927, 106.643  253.572, 138.033  255.057, 166.378  256.976, 203  274.553, 11.1377  276.632, 39.4477  277.371, 49.514  279.96, 84.7681  281.658, 107.889  283.942, 138.987  286.001, 167.026  288.643, 203  312.163, 13.8996  315.103, 41.8681  316.163, 51.9545  319.831, 86.8577  322.22, 109.589  325.447, 140.292  328.35, 167.913  332.038, 203 

please
CRimage.size = 80
index matching:    image# 0, 0     to world# 3, 5
index matching:    image# 0, 1     to world# 3, 6
index matching:    image# 0, 2     to world# 3, 7
index matching:    image# 0, 3     to world# 3, 8
index matching:    image# 0, 4     to world# 3, 9
index matching:    image# 0, 5     to world# 3, 10
index matching:    image# 1, 0     to world# 4, 5
index matching:    image# 1, 1     to world# 4, 6
index matching:    image# 1, 2     to world# 4, 7
index matching:    image# 1, 3     to world# 4, 8
index matching:    image# 1, 4     to world# 4, 9
index matching:    image# 1, 5     to world# 4, 10
index matching:    image# 2, 0     to world# 5, 5
index matching:    image# 2, 1     to world# 5, 6
index matching:    image# 2, 2     to world# 5, 7
index matching:    image# 2, 3     to world# 5, 8
index matching:    image# 2, 4     to world# 5, 9
index matching:    image# 2, 5     to world# 5, 10
index matching:    image# 3, 0     to world# 6, 5
index matching:    image# 3, 1     to world# 6, 6
index matching:    image# 3, 2     to world# 6, 7
index matching:    image# 3, 3     to world# 6, 8
index matching:    image# 3, 4     to world# 6, 9
index matching:    image# 3, 5     to world# 6, 10
index matching:    image# 4, 0     to world# 7, 5
index matching:    image# 4, 1     to world# 7, 6
index matching:    image# 4, 2     to world# 7, 7
index matching:    image# 4, 3     to world# 7, 8
index matching:    image# 4, 4     to world# 7, 9
index matching:    image# 4, 5     to world# 7, 10
index matching:    image# 5, 0     to world# 8, 5
index matching:    image# 5, 1     to world# 8, 6
index matching:    image# 5, 2     to world# 8, 7
index matching:    image# 5, 3     to world# 8, 8
index matching:    image# 5, 4     to world# 8, 9
index matching:    image# 5, 5     to world# 8, 10
index matching:    image# 6, 0     to world# 9, 5
index matching:    image# 6, 1     to world# 9, 6
index matching:    image# 6, 2     to world# 9, 7
index matching:    image# 6, 3     to world# 9, 8
index matching:    image# 6, 4     to world# 9, 9
index matching:    image# 6, 5     to world# 9, 10
index matching:    image# 7, 0     to world# 10, 5
index matching:    image# 7, 1     to world# 10, 6
index matching:    image# 7, 2     to world# 10, 7
index matching:    image# 7, 3     to world# 10, 8
index matching:    image# 7, 4     to world# 10, 9
index matching:    image# 7, 5     to world# 10, 10
index matching:    image# 8, 0     to world# 11, 5
index matching:    image# 8, 1     to world# 11, 6
index matching:    image# 8, 2     to world# 11, 7
index matching:    image# 8, 3     to world# 11, 8
index matching:    image# 8, 4     to world# 11, 9
index matching:    image# 8, 5     to world# 11, 10
index matching:    image# 9, 0     to world# 12, 5
index matching:    image# 9, 1     to world# 12, 6
index matching:    image# 9, 2     to world# 12, 7
index matching:    image# 9, 3     to world# 12, 8
index matching:    image# 9, 4     to world# 12, 9
index matching:    image# 9, 5     to world# 12, 10
coordinate matching:    image 35.5294, -6.40545     to world 62.07, 147.129, 0.0
coordinate matching:    image 33.4637, 24.1207     to world 62.07, 182.066, 0.0
coordinate matching:    image 32.4883, 34.7344     to world 62.07, 193.937, 0.0
coordinate matching:    image 29.3473, 71.7697     to world 62.07, 233.937, 0.0
coordinate matching:    image 27.0674, 97.0156     to world 62.07, 259.855, 0.0
coordinate matching:    image 24.235, 130.495     to world 62.07, 293.265, 0.0
coordinate matching:    image 86.259, -2.68977     to world 102.07, 147.129, 0.0
coordinate matching:    image 84.5506, 27.4862     to world 102.07, 182.066, 0.0
coordinate matching:    image 84.0488, 37.9824     to world 102.07, 193.937, 0.0
coordinate matching:    image 82.2594, 74.6152     to world 102.07, 233.937, 0.0
coordinate matching:    image 80.7738, 99.5121     to world 102.07, 259.855, 0.0
coordinate matching:    image 79.3161, 132.494     to world 102.07, 293.265, 0.0
coordinate matching:    image 108.14, -1.08291     to world 119.45, 147.129, 0.0
coordinate matching:    image 106.447, 28.8103     to world 119.45, 182.066, 0.0
coordinate matching:    image 105.952, 39.3494     to world 119.45, 193.937, 0.0
coordinate matching:    image 104.521, 75.7666     to world 119.45, 233.937, 0.0
coordinate matching:    image 103.606, 100.493     to world 119.45, 259.855, 0.0
coordinate matching:    image 102.511, 133.31     to world 119.45, 293.265, 0.0
coordinate matching:    image 155.327, 2.50148     to world 159.45, 147.129, 0.0
coordinate matching:    image 155.143, 32.0653     to world 159.45, 182.066, 0.0
coordinate matching:    image 154.91, 42.5242     to world 159.45, 193.937, 0.0
coordinate matching:    image 154.682, 78.4145     to world 159.45, 233.937, 0.0
coordinate matching:    image 154.596, 102.636     to world 159.45, 259.855, 0.0
coordinate matching:    image 154.507, 134.865     to world 159.45, 293.265, 0.0
coordinate matching:    image 169.293, 3.464     to world 171.309, 147.129, 0.0
coordinate matching:    image 169.238, 33.0578     to world 171.309, 182.066, 0.0
coordinate matching:    image 169.302, 43.3455     to world 171.309, 193.937, 0.0
coordinate matching:    image 169.333, 79.1385     to world 171.309, 233.937, 0.0
coordinate matching:    image 169.423, 103.342     to world 171.309, 259.855, 0.0
coordinate matching:    image 169.546, 135.378     to world 171.309, 293.265, 0.0
coordinate matching:    image 202.289, 5.97719     to world 199.965, 147.129, 0.0
coordinate matching:    image 202.842, 35.3023     to world 199.965, 182.066, 0.0
coordinate matching:    image 203.138, 45.4776     to world 199.965, 193.937, 0.0
coordinate matching:    image 203.843, 80.9247     to world 199.965, 233.937, 0.0
coordinate matching:    image 204.481, 104.772     to world 199.965, 259.855, 0.0
coordinate matching:    image 205.419, 136.449     to world 199.965, 293.265, 0.0
coordinate matching:    image 217.636, 7.27727     to world 213.626, 147.129, 0.0
coordinate matching:    image 218.523, 36.3636     to world 213.626, 182.066, 0.0
coordinate matching:    image 218.858, 46.4402     to world 213.626, 193.937, 0.0
coordinate matching:    image 220.052, 81.6308     to world 213.626, 233.937, 0.0
coordinate matching:    image 220.875, 105.426     to world 213.626, 259.855, 0.0
coordinate matching:    image 222.084, 136.772     to world 213.626, 293.265, 0.0
coordinate matching:    image 246.503, 9.52037     to world 239.562, 147.129, 0.0
coordinate matching:    image 247.889, 38.3327     to world 239.562, 182.066, 0.0
coordinate matching:    image 248.441, 48.3851     to world 239.562, 193.937, 0.0
coordinate matching:    image 250.452, 82.8983     to world 239.562, 233.937, 0.0
coordinate matching:    image 251.677, 106.428     to world 239.562, 259.855, 0.0
coordinate matching:    image 253.478, 137.564     to world 239.562, 293.265, 0.0
coordinate matching:    image 274.335, 11.6118     to world 265.097, 147.129, 0.0
coordinate matching:    image 276.296, 40.1247     to world 265.097, 182.066, 0.0
coordinate matching:    image 276.934, 50.0298     to world 265.097, 193.937, 0.0
coordinate matching:    image 279.607, 84.2214     to world 265.097, 233.937, 0.0
coordinate matching:    image 281.52, 107.359     to world 265.097, 259.855, 0.0
coordinate matching:    image 283.694, 138.233     to world 265.097, 293.265, 0.0
coordinate matching:    image 312.421, 14.5857     to world 300.95, 147.129, 0.0
coordinate matching:    image 314.962, 42.5479     to world 300.95, 182.066, 0.0
coordinate matching:    image 316.064, 52.2651     to world 300.95, 193.937, 0.0
coordinate matching:    image 320.518, 85.6258     to world 300.95, 233.937, 0.0
coordinate matching:    image 322.22, 109.589     to world 300.95, 259.855, 0.0
coordinate matching:    image 325.447, 138.885     to world 300.95, 293.265, 0.0

camera matrix
fx=301.669 0 cx=149.066
0 fy=237.063 cy=167.683
0 0 1

lens distortion
k1 = 0.0280614
k2 = -0.0278541
p1 = -0.00314248
p2 = 0.00323144

rotation vector
-0.155056  -0.115457  0.0134497

translation vector
-152.854  -325.802  262.059

check reprojection?
reprojection errors
0.769114    0.25348    0.163043    0.217367    0.0561251    0.188387    0.106855    0.103549    0.00969562    0.162657    0.168114    0.0333034    0.609429    0.146523    0.0704871    0.181848    0.0884141    0.106862    0.0507463    0.14039    0.0975367    0.131104    0.0561415    0.086923    0.194587    0.0698258    0.116085    0.080983    0.104469    0.163957    0.221519    0.0241836    0.0841894    0.076401    0.154674    0.239551    0.0717952    0.0691027    0.0618768    0.0627396    0.193959    0.0950587    0.0905574    0.201376    0.209811    0.123518    0.055315    0.0485843    0.0836709    0.204236    0.272494    0.163189    0.1419    0.137937    0.124958    0.370733    0.284505    0.849666    0.627085    0.48934   
error mean = 0.176032     std = 0.107796

저작자표시 비영리 동일조건

'Computer Vision > 2010ETRI-MultiAutoCalib' 카테고리의 다른 글

Ferenc Kahlesz & Cornelius Lilge & Reinhard Klein "Easy–to–Use Calibration of Multiple–Camera Setups"  (0) 2010.07.20
case study: 리프로젝션 에러 실험 결과 조사  (0) 2010.07.20
J. Weng, P. Cohen, and M. Herniou "Camera calibration with distortion models and accuracy evaluation"  (0) 2010.07.16
DLR Camera Calibration Toolbox  (0) 2010.07.16
Tomas Svoboda <Multi-Camera Self-Calibration>  (0) 2010.07.07
posted by maetel

Tomas Svoboda <Multi-Camera Self-Calibration>

2010. 7. 7. 14:04 Computer Vision/2010ETRI-MultiAutoCalib
http://cmp.felk.cvut.cz/~svoboda/SelfCal/

Center for Machine Perception, Department of Cybernetics, Czech Technical University in Prague
저작자표시 비영리 동일조건

'Computer Vision > 2010ETRI-MultiAutoCalib' 카테고리의 다른 글

Ferenc Kahlesz & Cornelius Lilge & Reinhard Klein "Easy–to–Use Calibration of Multiple–Camera Setups"  (0) 2010.07.20
case study: 리프로젝션 에러 실험 결과 조사  (0) 2010.07.20
J. Weng, P. Cohen, and M. Herniou "Camera calibration with distortion models and accuracy evaluation"  (0) 2010.07.16
DLR Camera Calibration Toolbox  (0) 2010.07.16
Test: reprojection error (Quickcam Pro4000)  (0) 2010.07.07
posted by maetel

티스토리툴바