Computational Visual Media

, Volume 4, Issue 3, pp 277–285 | Cite as

GPU based techniques for deep image merging

  • Jesse ArcherEmail author
  • Geoff Leach
  • Ron van Schyndel
Open Access
Research Article


Deep images store multiple fragments perpixel, each of which includes colour and depth, unlike traditional 2D flat images which store only a single colour value and possibly a depth value. Recently, deep images have found use in an increasing number of applications, including ones using transparency and compositing. A step in compositing deep images requires merging per-pixel fragment lists in depth order; little work has so far been presented on fast approaches.

This paper explores GPU based merging of deep images using different memory layouts for fragment lists: linked lists, linearised arrays, and interleaved arrays. We also report performance improvements using techniques which leverage GPU memory hierarchy by processing blocks of fragment data using fast registers, following similar techniques used to improve performance of transparency rendering. We report results for compositing from two deep images or saving the resulting deep image before compositing, as well as for an iterated pairwise merge of multiple deep images. Our results show a 2 to 6 fold improvement by combining efficient memory layout with fast register based merging.


deep image composite GPU performance 



The authors would like to thank Pyar Knowles for his original deep image software on which this work is based. It is available at


  1. [1]
    Heckenberg, D.; Saam, J.; Doncaster, C.; Cooper, C. Deep compositing. 2010. Available at Google Scholar
  2. [2]
    Duff, T. Deep compositing using lie algebras. ACM Transactions on Graphics Vol. 36, No. 3, Article No. 26, 2017.Google Scholar
  3. [3]
    Maule, M.; Comba, J. L. D.; Torchelsen, R.; Bastos, R. Memory-efficient order-independent transparency with dynamic fragment buffer. In: Proceedings of the 25th SIBGRAPI Conference on Graphics, Patterns and Images, 134–141, 2012.Google Scholar
  4. [4]
    Knowles, P.; Leach, G.; Zambetta, F. Efficient layered fragment buffer techniques. In: OpenGL Insights. Cozzi, P.; Riccio, C. Eds. CRC Press, 279–292, 2012.CrossRefGoogle Scholar
  5. [5]
    Knowles, P.; Leach, G.; Zambetta, F. Fast sorting for exact OIT of complex scenes. The Visual Computer Vol. 30, Nos. 6–8, 603–613, 2014.CrossRefGoogle Scholar
  6. [6]
    Porter, T.; Duff, T. Compositing digital images. In: Proceedings of the 11th Annual Conference on Computer Graphics and Interactive Techniques, 253–259, 1984.Google Scholar
  7. [7]
    Egstad, J.; Davis, M.; Lacewell, D. Improved deep image compositing using subpixel masks. In: Proceedings of the 2015 Symposium on Digital Production, 21–27, 2015.CrossRefGoogle Scholar
  8. [8]
    Hillman, P. The theory of OpenEXR deep samples. Technical Report. Weta Digital Ltd., 2013.Google Scholar
  9. [9]
    Knowles, P.; Leach, G.; Zambetta, F. Backwards memory allocation and improved OIT. In: Proceedings of the Pacific Graphics, 59–64, 2013.Google Scholar
  10. [10]
    McGuire, M. Computer graphics archive. 2017. Available at Scholar

Copyright information

© The Author(s) 2018

Authors and Affiliations

  1. 1.School of ScienceRMIT UniversityMelbourneAustralia

Personalised recommendations