Detecting parametric objects in large scenes by Monte Carlo sampling
- 831 Downloads
Point processes constitute a natural extension of Markov random fields (MRF), designed to handle parametric objects. They have shown efficiency and competitiveness for tackling object extraction problems in vision. Simulating these stochastic models is however a difficult task. The performances of the existing samplers are limited in terms of computation time and convergence stability, especially on large scenes. We propose a new sampling procedure based on a Monte Carlo formalism. Our algorithm exploits the Markovian property of point processes to perform the sampling in parallel. This procedure is embedded into a data-driven mechanism so that the points are distributed in the scene in function of spatial information extracted from the input data. The performances of the sampler are analyzed through a set of experiments on various object detection problems from large scenes, including comparisons to the existing algorithms. The sampler is also tested as optimization algorithm for MRF-based labeling problems.
KeywordsStochastic modeling Monte Carlo sampling Object detection Large scenes Energy minimization Point processes Markov random fields
This work was partially funded by the European Research Council (ERC Starting Grant “Robust Geometry Processing”, Grant agreement 257474). The authors thank A. Lehmussola, V. Lempitsky, H. Bischof, R. Ehrich, the French Mapping Agency (IGN), the Tour du Valat, and the BRGM for providing the datasets, as well as the reviewers for their valuable comments.
- Benchmark, (2013). Datasets, results and evaluation tools. http://www-sop.inria.fr/members/Florent.Lafarge/benchmark/evaluation.html.
- Byrd, J., Jarvis, S., & Bhalerao, A. (2010). On the parallelisation of mcmc-based image processing. IEEE International Symposium on Parallel and Distributed Processing. Atlanta, US.Google Scholar
- Chai, D., Forstner, W., & Lafarge, F. (2013). Recovering line-networks in images by junction-point processes. Computer Vision and Pattern Recognition, Portland.Google Scholar
- Chai, D., Forstner, W., & Yang, M. Y. (2012). Combine Markov random fields and marked point processes to extract building from remotely sensed images. International Society for Photogrammetry and Remote Sensing Congress. Melbourne, Australia.Google Scholar
- Descombes, X. (2011). Stochastic geometry for image analysis. Oxford: Wiley.Google Scholar
- Ge, W., & Collins, R. (2009). Marked point processes for crowd counting. Computer Vision and Pattern Recognition. Miami. Google Scholar
- Gonzalez, J., Low, Y., Gretton, A., & Guestrin, C. (2011). Parallel Gibbs sampling: From colored fields to thin junction trees. Journal of Machine Learning Research, 15, 324–332.Google Scholar
- Harkness, M., & Green, P. (2000). Parallel chains, delayed rejection and reversible jump mcmc for object recognition. British Machine Vision Conference. Bristol, United Kingdom.Google Scholar
- Lehmussola, A., Ruusuvuori, P., Selinummi, J., Huttunen, H., & Yli-Harja, O. (2007). Computational framework for simulating fluorescence microscope images with cell populations. IEEE Transactions on Medical Imaging, 26(7), 1010–1016.Google Scholar
- Lempitsky, V., & Zisserman, A. (2010). Learning to count objects in images. Conference on Neural Information Processing Systems. Vancouver, Canada.Google Scholar
- Nguyen, H.-G., Fablet, R., & Bouchet, J. (2010). Spatial statistics of visual keypoints for texture recognition. European Conference on Computer Vision. Heraklion, Greece.Google Scholar
- Salamon, P., Sibani, P., & Frost, R. (2002). Facts, Conjectures, and Improvements for Simulated Annealing. Philadelphia: SIAM Monographs on Mathematical Modeling and Computation.Google Scholar
- Stoica, R. S., Martinez, V., & Saar, E. (2007). A three dimensional object point process for detection of cosmic filaments. Journal of the Royal Statistical Society, 56(4), 459.Google Scholar
- Sun, K., Sang, N., & Zhang, T. (2007). Marked point process for vasculartree extraction on angiogram. Energy Minimization Methods in Computer Vision and Pattern Recognition. Ezhou, China.Google Scholar
- Utasi, A., & Benedek, C. (2011). A 3-D marked point process model for multi-view people detection. Conference on Computer Vision and Pattern Recognition. Colorado Springs, US. Google Scholar
- Verdie, Y., & Lafarge, F. (2012). Efficient Monte Carlo sampler for detecting parametric objects in large scenes. European Conference on Computer Vision. Firenze, Italy.Google Scholar
- Zhu, S., Guo, C., Wang, Y., & Xu, Z. (2005). What are textons? International Journal of Computer Vision, 62(1–2), 121–143.Google Scholar