Abstract
Past research into multi-modality sensor data fusion has given rise to approaches that are generally heuristic and ad hoc. In this paper we utilize the calculus of variations as the underlying framework for fusing registered images of different modalities when models relating these modalities are available. The result is a mathematically rigorous method for improving the accuracy with which parameters can be estimated. Using both dense and sparse simulated range and intensity data, the proposed approach is demonstrated on the problem of estimating the surface representing the three dimensional structure of a scene. The results indicate that a four to five-fold increase in surface estimation accuracy with respect to the original input data can be realized. Furthermore, an 8%–250% increase in accuracy over surface estimation from each sensing modality alone (i.e., via shape from shading or surface reconstruction) can be realized.
Similar content being viewed by others
References
J.J. Clark and A.L. Yuille,Data Fusion for Sensory Information Processing Systems, Kluwer, Boston, MA, 1990.
N. Nandhakumar and J. Aggarwal, “Multisensory computer vision,” inAdvances in Computers, Vol. 34, M. Yovits (ed.), Academic Press, San Diego, CA, 1992.
A. Blake, A. Zisserman, and G. Knowles, “Surface descriptions from stereo and shading,” inImage and Vision Computing, 3(4), pp. 183–191, 1985.
M.J. Magee, B.A. Boyter, C.H. Chien, and J.K. Aggarwal, “Experiments in intensity guided range sensing recognition of three-dimensional objects,” inIEEE Trans. Pattern Analysis and Machine Intelligence, 7(6), pp. 629–637, 1985.
H. Pien, “Achieving safe autonomous landings on Mars using vision-based approaches,” inProceedings of the SPIE Conference on Cooperative Intelligent Robotics in Space, 1991.
K. Ikeuchi and K. Sato, “Determining reflective properties of an object using range and brightness images,” inIEEE Trans. Pattern Analysis and Machine Intelligence, 13(11), pp. 1139–1153, Nov. 1991.
D. Nitzan, A. Brain, and R. Duda, “The Measurement and Use of Registered Reflectance and Range Data in Scene Analysis,” inProc. of the IEEE, 65(2), February 1977.
R. Duda, D. Nitzan, and P. Barrett, “Use of range and reflectance data to find planar surface regions,” inIEEE Trans. Pattern Analysis and Machine Intelligence, PAMI-1(3), July 1979.
B. Gil, A. Mitiche, and J.K. Aggarwal, “Experiments in combining intensity and range edge maps,” inCVGIP, 21(3), pp. 395–441, 3/1983.
I.S. Kweon, M. Hebert, and T. Kanade, “Sensor fusion of range and reflectance data for outdoor scene analysis,” in2nd Annual Workshop on Space Operations, Automation, and Robotics, NASA, Washington, DC, 1988.
C.J. Delcroix and M.A. Abidi, “Fusion of range and intensity edge maps,” inSPIE Proc. on Sensor Fusion: Spatial Reasoning and Scene Interpretation, Vol. 1003, SPIE, pp. 145–152, 1988.
T.A. Mancini and L.B. Wolff, “3D shape and source location from depth and reflectance,” inSPIE Proc. Optics, Illumination, and Image Sensing for Machine Vision VI, Vol. 1614, 1991.
Y.F. Wang and D.I. Cheng, “Three dimensional shape construction and recognition by fusing intensity and structured lighting,” inPattern Recognition, 25(12), pp. 1411–1425, 1992.
J.E. Cryer, P.S. Tsai, and M. Shah, “Combining shape from shading and stereo using human vision model,” Department of Computer Science Technical Report CS-TR-92-25, University of Central Florida, Orlando, FL, 1992.
E.B. Gamble, D. Geiger, T. Poggio, and D. Weinshall, “Integration of vision modules and labeling of surface discontinuities,” inIEEE Trans. Systems, Man, Cybernetics, 19(6), pp. 1156–1161, November/December 1989.
T. Poggio, J. Little, W. Gillett, D. Geiger, D. Wienshall, M. Villalba, N. Larson, T. Cass, H. Bulhoff, M. Drumheller, P. Oppenheimer, W. Yang, and A. Hurlbert, “The MIT Vision Machine,” inProc. DARPA Image Understanding Workshop, Morgan Kaufmann, San Mateo, CA, April 1988.
A. Blake, “Comparison of the efficiency of deterministic and stochastic algorithms for visual reconstruction,” inIEEE Trans. on Pattern Analysis and Machine Intelligence, 11(11), pp. 2–12, 1989.
A.A. Amini, T.E. Weymouth, and R.C. Jain, “Using dynamic programming for solving variational problems in vision,” inIEEE Trans. on Pattern Analysis and Machine Intelligence, 12(9), pp. 855–867, 1990.
M. Kass, A. Witkins, and D. Terzopoulos, “Snakes: active contour models,” inProc. First Int'l Conf. on Computer Vision, London, 1987, pp. 259–268, 1987.
L.D. Cohen, “On active contour models and balloons,” inComputer Vision, Graphics, and Image Understanding: Image Understanding, Vol. 53, pp. 211–218, March 1991.
J.M. Gauch and M. Seaidoun, “Comparison of implementation strategies for deformable surfaces in computer vision,” inProc. SPIE Conf. Mathematical Methods in Medical Imaging, Vol. 1768, San Diego, CA, July 1992.
F. Leymarie and M.D. Levine, “Tracking deformable objects in the plane using an active contour model,” inIEEE. Trans. Pattern Analysis and Machine Intelligence, 15(6), pp. 617–634, June 1993.
W.E.L. Grimson,From Images to Surfaces, MIT Press, Cambridge, MA, 1981.
D. Terzopoulos, “Regularization of inverse visual problems involving discontinuities,” inIEEE Trans. on Pattern Analysis and Machine Intelligence, 8(4), July 1986.
R.M. Bolle and B.C. Vemuri, “On three dimensional surface reconstruction methods,” inIEEE Trans. on Pattern Analysis and Machine Intelligence, 13(1), pp. 2–13, 1991.
A. Blake and A. Zisserman,Visual Reconstruction, MIT Press, Cambridge, MA, 1987.
S. Geman and D. Geman, “Stochastic Relaxation, Gibbs Distributions, and the Bayesian Restoration of Images,” inIEEE Trans. Pattern Analysis and Machine Intelligence, PAMI-6(6), Nov. 1984.
D. Mumford and J. Shah, “Boundary Detection by Minimizing Functionals, I,” inProc. of the IEEE Conference on Computer Vision and Pattern Recognition, IEEE, 1985.
B.K.P. Horn and M.J. Brooks (ed.),Shape from Shading, MIT Press, Cambridge, MA, 1989.
N.E. Hurt, “Mathematical methods in shape-from-shading: a review of recent results,” inActa Applicandae Mathematicae, 23, pp. 163–188, 1991.
B.K.P. Horn, “Shape from shading: a method for obtaining the shape of a smooth opaque object from one view,” Ph.D. Thesis, Dept. of Electrical Engineering, MIT-AI-TR-232, 1970.
B.K.P. Horn, “Obtaining shape from shading information,” inThe Psychology of Computer Vision, P.H. Winston (ed.), McGraw-Hill, NY, pp. 115–155, 1975.
K. Ikeuchi and B.K.P. Horn, “Numerical shape from shading and occluding boundaries,” inArtificial Intelligence, 17(1–3), pp. 141–184, 1981.
T. Strat, “A numerical method for shape from shading for a single image,” S.M. Thesis, MIT EECS, MIT, Cambridge, MA, 1979.
B.K.P. Horn and M.J. Brooks, “The variational approach to shape from shading,” inComputer Vision, Graphics, and Image Processing, 33(2), pp. 174–208, 1986.
C.H. Lee and A. Rosenfeld, “Improved methods of estimating shape from shading using the light source coordinate system,” in Horn and Brooks (ed.),Shape from Shading, MIT Press, Cambridge, MA, revised version ofArtificial Intelligence, 1985, 26(2), 1989.
B.K.P. Horn, “Height and gradient from shading,” inInt'l Journal of Computer Vision, 5(1), pp. 37–75, 1990.
A.P. Pentland, “Local shading analysis,” inIEEE Trans. on Pattern Analysis and Machine Intelligence, 6(2), pp. 170–187, 1984. Also appeared in Horn and Brooks [1989] in a revised form.
A.P. Pentland, “Shape information from shading: a theory about human perception,” inProc. 2nd Int'l Conf. on Computer Vision, IEEE, Los Alamitos, CA, 1988.
M. Bichsel and A.P. Pentland, “A simple algorithm for shape from shading,” inProc. Computer Vision and Pattern Recognition 1992 Conference, IEEE, Los Alamitos, CA, 1992.
P. Dupuis and J. Oliensis, “Direct method for reconstructing shape from shading,” inProc. Computer Vision and Pattern Recognition 1992 Conference, IEEE, Los Alamitos, CA, 1992.
B.V.H. Saxberg, “A modern differential geometric approach to shape from shading,” Ph.D. Thesis, MIT EECS, 1989.
T. Poggio, V. Torre, and C. Koch, “Computational vision and regularization theory,” inNature, 317, 26 September 1985.
M. Bertero, T. Poggio, and V. Torre, “Ill-Posed Problems in Early Vision,” inProceedings of the IEEE, 76(8), 1988.
J. Oliensis, “Uniqueness in shape from shading” inInt'l J. of Computer Vision, 6(2), pp. 75–104, 1991a.
R.T. Frankot and R. Chellappa, “A method for enforcing integrability in shape from shading algorithms,” inIEEE Trans. on Pattern Analysis and Machine Intelligence, 10(4), pp. 439–451, (1988).
D. Forsyth and A. Zisserman, “Reflections on shading,” inIEEE Trans. on Pattern Analysis and Machine Intelligence, 13(7), pp. 671–679, 1991.
Q. Zheng and R. Chellappa, “Estimation of illuminant direction, albedo, and shape from shading,” inIEEE PAMI, 13(7), pp. 671–679, 1991.
A.P. Pentland, “Finding the illuminant direction,” inOptical Society of America, 7294, pp. 448–455, 1982.
M.J. Brooks and B.K.P. Horn, “Shape and source from shading,” inProc. 1985 Int'l Joint Conf. Artificial Intelligence, Morgan Kaufmann, San Mateo, CA, pp. 932–936, 1985.
T. Simchony, R. Chellappa, and M. Shao, “Direct analytical methods for solving Poisson equations in computer vision problems,” inIEEE Trans. on Pattern Analysis and Machine Intelligence, 12(5), pp. 435–446, 1990.
E. Ruoy and A. Tourin, “A viscosity solutions approach to shape-from-shading,” inSIAM J. Numerical Analysis, 29(3), pp. 867–884, 1992.
M.A. Penna, “A shape from snading analysis for a single perspective image of a polyhedron,” inIEEE Trans. on Pattern Analysis and Machine Intelligence, 11(6), pp. 545–554, 1989.
J. Malik and D. Maydan, “Recovering three-dimensional shape from a single image of curved objects,” inIEEE Trans. on Pattern Analysis and Machine Intelligence, 11(6), pp. 555–566, 1989.
R. Courant and D. Hilbert,Methods of Mathematical Physics, Interscience, NY, 1953.
D. Weinshall, “The shape of shading,” MIT Artificial Intelligence Laboratory Memo No. 1264, Cambridge, MA, October 1990.
J. Oliensis, “Shape from shading as a partially well-constrained problem,” inCVGIP: Image Processing, 54(2), pp. 163–183, 1991b.
B.K.P. Horn,Robot Vision, MIT Press, Cambridge, MA, 1986.
A.P. Pentland, “Linear shape from shading,” inInt. J. of Computer Vision, 4(2), pp. 153–162, 1990.
G. Wahba, “Practical approximate solutions to linear operator equations when the data are noisy,” inSIAM J. Numerical Analysis, 14(4), pp. 651–667, September 1978.
G. Wahba, “Ill-posed problems: numerical and statistical methods for mildly, moderately, and severely ill-posed problems with noisy data,” University of Wisconsin Department of Statistics Technical Report No. 595, Madison, WI, 1980.
G. Wahba,Spline Models for Observational Data, SIAM, Philadelphia, PA, 1990.
B. Shahraray and D.J. Anderson, “Optimal estimation of contour properties by cross-validated regularization,” inIEEE Trans. Pattern Analysis and Machine Intelligence, 11(6), pp. 600–610, June 1989.
S.J. Reeves and R.M. Mersereau, “Automatic assessment of constraint sets in image restoration,” inIEEE Trans. Image Processing, 1(1), pp. 119–123, January 1992.
A. Rosenfeld and M. Thurston, “Edge and curve detection for visual scene analysis,” inIEEE Trans. Computers, Vol. C-20, 1971.
A.M. Thompson, J.C. Brown, J.W. Kay, and D.M. Titterington, “A study of methods of choosing the smoothing parameter in image restoration by regularization,” inIEEE Trans. on Pattern Analysis and Machine Intelligence, 13(4), pp. 326–339, April 1991.
P. Hall and I. Koch, “On the feasibility of cross-validation in image analysis,” inSIAM J. Applied Math, 52(1), pp. 292–313, February 1992.
H. Pien,A variational framework for multi-sensor data fusion, Northeastern University College of Computer Science Ph.D. Dissertation, September 1993.
L. Ambrosio and V.M. Tortorelli, “Approximation of functionals depending on jumps by elleptic functionals via Γ-convergence,” inComm. Pure and Applied Mathematics, Vol. 43, pp. 999–36, 1990.
L. Uhr, “Layered ‘recognition cone’ networks that preprocess,” inIEEE Trans. Computers, Vol. C-21, 1972.
S. Tanimoto and T. Pavlidis, “A hierarchical data structure for picture processing,” inComputer Graphics and Image Processing, Vol. 4, 1975.
D. Terzopoulos, “Image analysis using multigrid relaxation methods,” inIEEE Trans. on Pattern Analysis and Machine Intelligence, 8(2), March 1986.
M.R. Leuttgen, W.C. Karl, A.S. Willsky, and R.R. Tenney, “Multiscale representations of Markov random fields,” MIT Laboratory for Information and Decision Systems Report No. LIDS-P-2157, December 1992.
M.R. Leuttgen, W.C. Karl, and A.S. Willsky, “Efficient multiscale regularization with applications to the computation of optical flow,” Center for Intelligent Control Systems Technical Report No. CICS-P-370, April 1993.
H. Pien and J. Gauch, “A variational approach to sensor fusion using registered range and intensity data,” inSPIE Conf. on Sensor Fusion and Aerospace Applications, Vol. 1956, April 1993.
Author information
Authors and Affiliations
Additional information
H. Pien is supported by Draper Laboratory under IR&D No. 451; J. Gauch is partially supported by the National Science Foundation under Grant IRI-9109431.
Rights and permissions
About this article
Cite this article
Pien, H.H., Gauch, J.M. A variational approach to multi-sensor fusion of images. Appl Intell 5, 217–235 (1995). https://doi.org/10.1007/BF00872223
Issue Date:
DOI: https://doi.org/10.1007/BF00872223