Resolution-Independent Meshes of Superpixels
The over-segmentation into superpixels is an important pre-processing step to smartly compress the input size and speed up higher level tasks. A superpixel was traditionally considered as a small cluster of square-based pixels that have similar color intensities and are closely located to each other. In this discrete model the boundaries of superpixels often have irregular zigzags consisting of horizontal or vertical edges from a given pixel grid. However digital images represent a continuous world, hence the following continuous model in the resolution-independent formulation can be more suitable for the reconstruction problem.
Instead of uniting squares in a grid, a resolution-independent superpixel is defined as a polygon that has straight edges with any possible slope at subpixel resolution. The harder continuous version of the over-segmentation problem is to split an image into polygons and find a best (say, constant) color of each polygon so that the resulting colored mesh well approximates the given image. Such a mesh of polygons can be rendered at any higher resolution with all edges kept straight.
We propose a fast conversion of any traditional superpixels into polygons and guarantees that their straight edges do not intersect. The meshes based on the superpixels SEEDS (Superpixels Extracted via Energy-Driven Sampling) and SLIC (Simple Linear Iterative Clustering) are compared with past meshes based on the Line Segment Detector. The experiments on the Berkeley Segmentation Database confirm that the new superpixels have more compact shapes than pixel-based superpixels.
The work has been supported by the EPSRC grant “Application-driven Topological Data Analysis” (2018-2023), EP/R018472/1.
- 4.Duan, L., Lafarge, F.: Image partitioning into convex polygons. In: Proceedings of CVPR (Computer Vision and Pattern Recognition), pp. 3119–3127 (2015)Google Scholar
- 7.Forsythe, J., Kurlin, V., Fitzgibbon, A.: Resolution-independent superpixels based on convex constrained meshes. In: Proceedings of ISVC (2016)Google Scholar
- 10.Kurlin, V., Muszynski, G.: A persistence-based approach to automatic detection of line segments in images. In: Proceedings of CTIC, pp. 137–150 (2019)Google Scholar
- 12.Li, Z., Chen, J.: Superpixel segmentation using linear spectral clustering. In: Proceedings of CVPR, pp. 1356–1363 (2015)Google Scholar
- 13.Liu, M.Y., Tuzel, O., Ramalingam, S., Chellappa, R.: Entropy rate superpixel segmentation. In: Proceedings of CVPR, pp. 2097–2104 (2011)Google Scholar
- 14.Luengo, I., Basham, M., French, A.: Smurfs: Superpixels from multi-scale refinement of super-regions. In: Proceedings of BMVC (2016)Google Scholar
- 15.Moore, A., Prince, S., Warrell, J.: Lattice cut - constructing superpixels using layer constraints. In: Proceedings of CVPR, pp. 2117–2124 (2010)Google Scholar
- 17.Veksler, O., Boykov, Y., Mehrani, P.: Superpixels and supervoxels in an energy optimization framework. In: Proceedings of ECCV, pp. 211–224 (2010)Google Scholar
- 18.Viola, F., Fitzgibbon, A., Cipolla, R.: A unifying resolution-independent formulation for early vision. In: Proceedings of CVPR, pp. 494–501 (2012)Google Scholar
- 19.Yao, J., Boben, M., Fidler, S., Urtasun, R.: Real-time coarse-to-fine topologically preserving segmentation. In: Proceedings of CVPR, pp. 216–225 (2015)Google Scholar
- 20.Zhang, Y., Hartley, R., Mashford, J., Burn, S.: Superpixels via pseudo-boolean optimization. In: Proceedings of ICCV, pp. 211–224 (2011)Google Scholar