一种基于近似点光源模型的光度立体视觉系统标定方法
|
摘要
文章研究了一种由近似点光源组成的光度立体视觉系统的标定方法。由于光源自身的非均匀发光特性,传统的平行光和点光源模型不再适用,为实现精确的三维法向重建,必须对光源及光照场进行精确的标定。论文首先根据光源自身的发光特性建立其辐射度模型,进而提出了一种两步标定策略,对光源位置参数及主光轴参数进行标定。以相机作为参照系,提出了一种多球标定方法,用于计算各个光源的位置参数,该方法利用相机观察到的多个球面高光点图像坐标,结合球体轮廓线及球体半径实现了光源位置的精确计算。为实现光源主光轴参数的估算,提出了一种基于参考平面的标定方法,通过分析近似点光源照射下的平面亮度分布,结合等亮度线拟合策略实现了光源主光轴方向参数的计算。基于系统标定参数以及光源的发光模型,就可以对每个场景点的入射光条件进行精确建模,从而实现精确的法向计算。在实验部分,分别对标准几何物体和自由曲面物体进行三维重建实验,并与传统的平行光及点光源模型进行了对比。结果显示,所提出的标定方法能够在非均匀光照条件下获得精确的三维法向重建结果。
In this paper,a new calibration method for a specially designed photometric stereo system was presented.Different from conventional photometric stereo systems which adopt parallel or distant point light sources,the quasi-point light sources are used in the proposed system,which permits the system with compact size and short working distance.Subject to the non-isotropic radiance property of the quasi-pointlight source,the lighting field should be precisely calibrated so as to determine the intensity and direction for each incident light ray.The proposed calibration method was performed in two steps.The first step was to calculate light source position,where a multi-sphere-based method was introduced.By extracting the highlight image points formed from the light source on the specular sphere,and utilizing the sphere contour and radius,it shows that the light source coordinate under camera reference frame can be precisely estimated.In second step,a reference plane-based method was proposed to calculate the principle optical axis of light source.It proves that,the principle optical axis should cross the brightness point in the image of reference plane,and the brightness point can be estimated by fitting the iso-luminance curve in the image.With the calibration parameters,and considering the radiance model of light source,lighting condition for each scene point can be determined.The experiments were conducted with various target surfaces and compared with conventional light models to demonstrate feasibility and accuracy improvement by the proposed calibration method.
In this paper,a new calibration method for a specially designed photometric stereo system was presented.Different from conventional photometric stereo systems which adopt parallel or distant point light sources,the quasi-point light sources are used in the proposed system,which permits the system with compact size and short working distance.Subject to the non-isotropic radiance property of the quasi-pointlight source,the lighting field should be precisely calibrated so as to determine the intensity and direction for each incident light ray.The proposed calibration method was performed in two steps.The first step was to calculate light source position,where a multi-sphere-based method was introduced.By extracting the highlight image points formed from the light source on the specular sphere,and utilizing the sphere contour and radius,it shows that the light source coordinate under camera reference frame can be precisely estimated.In second step,a reference plane-based method was proposed to calculate the principle optical axis of light source.It proves that,the principle optical axis should cross the brightness point in the image of reference plane,and the brightness point can be estimated by fitting the iso-luminance curve in the image.With the calibration parameters,and considering the radiance model of light source,lighting condition for each scene point can be determined.The experiments were conducted with various target surfaces and compared with conventional light models to demonstrate feasibility and accuracy improvement by the proposed calibration method.
引文
[1]Woodham RJ.Photometric method for determining surface orientation from multiple images[M]//Shape from Shading.MIT Press,1989:513-531.
[2]Abrams A,Hawley C,Pless R.Heliometric stereo:shape from sun position[C]//Proceedings of the12th European Conference on Computer Vision,2012:357-370.
[3]Schnieders D,Wong KYK,Dai ZW.Polygonal light source estimation[C]//The 9th Asian Conference on Computer Vision,2010:96-107.
[4]Yu LF,Yeung SK,Tai YW,et al.Outdoor photometric stereo[C]//IEEE International Conference on Computational Photography,2013:1-8.
[5]Han TQ,Shen HL.Photometric stereo for general BRDFs via reflection sparsity modeling[J].IEEE Transaction on Image Processing,2015,24(12):4888-4903.
[6]Zhou W,Kambhamettu C.A unified framework for scene illuminant estimation[J].Image and Vision Computing,2008,26(3):415-429.
[7]Zhou SK,Aggarwal G,Chellappa R,et al.Appearance characterization of linear lambertian objects,generalized photometric stereo,and illumination-invariant face recognition[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2007,29(2):230-245.
[8]Xie WY,Song Z,Chung RC.Real-time threedimensional fingerprint acquisition via a new photometric stereo means[J].Optical Engineering,2013,52(10):103103.
[9]Powell MW,Sarkar S,Goldgof D.A simple strategy for calibrating the geometry of light sources[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2001,23(9):1022-1027.
[10]Rykowski R,Kostal H.Novel approach for LED luminous intensity measurement[C]//Proceedings of SPIE-The International Society for Optical Engineering,2008:69100.
[11]Wu TP,Tang CK.Dense photometric stereo using a mirror sphere and graph cut[C]//Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition,2005:140-147.
[12]Wong KYK,Schnieders D,Li S.Recovering light directions and camera poses from a single sphere[C]//Computer Vision-ECCV 2008:the 10th European Conference on Computer Vision,2008:631-642.
[13]Kato K,Sakaue F,Sato J.Extended multiple view geometry for lights and cameras from photometric and geometric constraints[C]//The 20th International Conference on Pattern Recognition,2010:2110-2113.
[14]Schnieders D,Wong KYK.Camera and light calibration from reflections on a sphere[J].Computer Vision and Image Understanding,2013,117(10):1536-1547.
[15]Alldrin NG,Kriegman DJ.A planar light probe[C]//Proceedings of the 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition,2006:2324-2330.
[16]Weber M,Cipolla R.A practical method for estimation of point light-sources[C]//British Machine Vision Conference,2001:1-10.
[17]Ackermann J,Fuhrmann S,Goesele M.Geometric point light source calibration[M]//Vision,Modeling&Visualization,2013:161-168.
[18]Takai T,Maki A,Niinuma K,et al.Difference sphere:an approach to near light source estimation[J].Computer Vision and Image Understanding,2009,113(9):966-978.
[19]Park J,Sinha SN,Matsushita Y,et al.Calibrating a non-isotropic near point light source using a plane[C]//2014 IEEE Conference on Computer Vision and Pattern Recognition,2014:2267-2274.
[20]Canny J.A computational approach to edge detection[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1986,8(6):679-698.
[21]Zhang ZY.A flexible new technique for camera calibration[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2000,22(11):1330-1334.
[22]Koks D.Explorations in Mathematical Physics:The Concepts Behind an Elegant Language[M].Springer Science&Business Media,2006.
[2]Abrams A,Hawley C,Pless R.Heliometric stereo:shape from sun position[C]//Proceedings of the12th European Conference on Computer Vision,2012:357-370.
[3]Schnieders D,Wong KYK,Dai ZW.Polygonal light source estimation[C]//The 9th Asian Conference on Computer Vision,2010:96-107.
[4]Yu LF,Yeung SK,Tai YW,et al.Outdoor photometric stereo[C]//IEEE International Conference on Computational Photography,2013:1-8.
[5]Han TQ,Shen HL.Photometric stereo for general BRDFs via reflection sparsity modeling[J].IEEE Transaction on Image Processing,2015,24(12):4888-4903.
[6]Zhou W,Kambhamettu C.A unified framework for scene illuminant estimation[J].Image and Vision Computing,2008,26(3):415-429.
[7]Zhou SK,Aggarwal G,Chellappa R,et al.Appearance characterization of linear lambertian objects,generalized photometric stereo,and illumination-invariant face recognition[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2007,29(2):230-245.
[8]Xie WY,Song Z,Chung RC.Real-time threedimensional fingerprint acquisition via a new photometric stereo means[J].Optical Engineering,2013,52(10):103103.
[9]Powell MW,Sarkar S,Goldgof D.A simple strategy for calibrating the geometry of light sources[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2001,23(9):1022-1027.
[10]Rykowski R,Kostal H.Novel approach for LED luminous intensity measurement[C]//Proceedings of SPIE-The International Society for Optical Engineering,2008:69100.
[11]Wu TP,Tang CK.Dense photometric stereo using a mirror sphere and graph cut[C]//Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition,2005:140-147.
[12]Wong KYK,Schnieders D,Li S.Recovering light directions and camera poses from a single sphere[C]//Computer Vision-ECCV 2008:the 10th European Conference on Computer Vision,2008:631-642.
[13]Kato K,Sakaue F,Sato J.Extended multiple view geometry for lights and cameras from photometric and geometric constraints[C]//The 20th International Conference on Pattern Recognition,2010:2110-2113.
[14]Schnieders D,Wong KYK.Camera and light calibration from reflections on a sphere[J].Computer Vision and Image Understanding,2013,117(10):1536-1547.
[15]Alldrin NG,Kriegman DJ.A planar light probe[C]//Proceedings of the 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition,2006:2324-2330.
[16]Weber M,Cipolla R.A practical method for estimation of point light-sources[C]//British Machine Vision Conference,2001:1-10.
[17]Ackermann J,Fuhrmann S,Goesele M.Geometric point light source calibration[M]//Vision,Modeling&Visualization,2013:161-168.
[18]Takai T,Maki A,Niinuma K,et al.Difference sphere:an approach to near light source estimation[J].Computer Vision and Image Understanding,2009,113(9):966-978.
[19]Park J,Sinha SN,Matsushita Y,et al.Calibrating a non-isotropic near point light source using a plane[C]//2014 IEEE Conference on Computer Vision and Pattern Recognition,2014:2267-2274.
[20]Canny J.A computational approach to edge detection[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1986,8(6):679-698.
[21]Zhang ZY.A flexible new technique for camera calibration[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2000,22(11):1330-1334.
[22]Koks D.Explorations in Mathematical Physics:The Concepts Behind an Elegant Language[M].Springer Science&Business Media,2006.