Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.
Moving beyond the box:Moving beyond the box:
automating the digitisation ofautomating the digitisation of
insect collectio...
Blagoderov et al (2012) No specimen left behind: industrial scale digitization of natural history collections.
ZooKeys 209...
Drawer level imaging is (mostly) a solved problem
1. Place drawer 2. Scan 3. Stitch
Result: High resolution composite image
Drawer level imaging is (mostly) a solved problem
• Fast (5 mins per drawer)
• Hig...
But, two key problems remain…
Synchronisation Label data
Keeping the physical & digital
copies in sync
Capturing data from...
• Don’t worry about it, re-image as required
• Lock down the drawers
• Crop-out each specimen image
– Automate the croppin...
Annotation software, NHM working with SmartDrive:
Initial supporting software
• No automated cropping
• Manually link imag...
Automating specimen segmentation
Starting image
Auto-segment
Mark errors
Correct
Work with Pieter Holtzhausen and Stéfan v...
Original Primary segmentation
(contrast based)
Segmentation methods
Secondary
(seed growing)
Inselect
http://naturalhistorymuseum.github.io/inselect/
• Currently pre-release (alpha)
• Automatically detects specimens...
Whole Drawer image
Inselect: segmentation of specimen images
Auto-segmented
images in sidebar
Whole Drawer image
Easy to spot &
correct errors
Secondary re-segmentation (seed growing)
Whole Drawer image
Whole Drawer image
Easy to spot &
correct errors
Secondary re-segmentation (seed growing)
Works on slides & pinned insects
Whole Drawer image Also testing mineral & fossil samples
Planned UX / UI Enhancements
Whole Drawer image
Unit tray
recognition
Multi-
specimen
annotation
Plug in controlled
vocabu...
1D & 2D barcode recognition
Whole Drawer image
• Recognition and reading of 1D &
2D matrix barcodes from images
• Differen...
The Holy Grail: label imaging & text recognition
Whole Drawer image
Chauliodes pectinicornis
DorsalCaudalFrontal
Reconstru...
The Holy Grail: label imaging & text recognition
Whole Drawer image
Agulla astuta
DorsalCaudalLateral
Reconstructed labels
Approaches to label imaging pinned specimens
Whole Drawer image
Could be incorporated as part of the barcode dispensing pr...
Acknowledgements
Whole Drawer image
Segmentation algorithm & app. development
Stefan van der Walt and Pieter Holtzhausen
A...
Upcoming SlideShare
Loading in …5
×

Moving beyond the box: automating the digitisation of insect collections

1,188 views

Published on

Presented at the 2014 TDWG annual conference, Elmia Congress Centre in Jönköping, Sweden, October 27-31, 2014.

Published in: Science
  • Be the first to comment

  • Be the first to like this

Moving beyond the box: automating the digitisation of insect collections

  1. 1. Moving beyond the box:Moving beyond the box: automating the digitisation ofautomating the digitisation of insect collectionsinsect collections
  2. 2. Blagoderov et al (2012) No specimen left behind: industrial scale digitization of natural history collections. ZooKeys 209: 131–146, doi: 10.3897/zookeys.209.3178
  3. 3. Drawer level imaging is (mostly) a solved problem 1. Place drawer 2. Scan 3. Stitch
  4. 4. Result: High resolution composite image Drawer level imaging is (mostly) a solved problem • Fast (5 mins per drawer) • High resolution (circa 500MB per image) • No specimen handling
  5. 5. But, two key problems remain… Synchronisation Label data Keeping the physical & digital copies in sync Capturing data from multiple pinned labels
  6. 6. • Don’t worry about it, re-image as required • Lock down the drawers • Crop-out each specimen image – Automate the cropping process – Link each specimen to its digital image – Make it easy to collect label data Approaches to the synchronisation problem The only practical solution, but a new rate limiting step
  7. 7. Annotation software, NHM working with SmartDrive: Initial supporting software • No automated cropping • Manually link images & specimens • Poor UX/UI • Closed source and proprietary • Not cross-platform (Windows only) A good first step to understanding the problems
  8. 8. Automating specimen segmentation Starting image Auto-segment Mark errors Correct Work with Pieter Holtzhausen and Stéfan van der Walt (Stellenbosch University) Software: Inselect, written in Python
  9. 9. Original Primary segmentation (contrast based) Segmentation methods Secondary (seed growing)
  10. 10. Inselect http://naturalhistorymuseum.github.io/inselect/ • Currently pre-release (alpha) • Automatically detects specimens • Creates bounding boxes for cropping and exporting images • Rapid annotation interface • Persistent settings & keyboard shortcuts • Data export in JSON format • Open source & modular • Python based (OpenCV, scikit- image libraries) • Windows, OSX & Linux Automated recognition, cropping and annotation of specimens
  11. 11. Whole Drawer image Inselect: segmentation of specimen images Auto-segmented images in sidebar Whole Drawer image
  12. 12. Easy to spot & correct errors Secondary re-segmentation (seed growing) Whole Drawer image
  13. 13. Whole Drawer image Easy to spot & correct errors Secondary re-segmentation (seed growing)
  14. 14. Works on slides & pinned insects Whole Drawer image Also testing mineral & fossil samples
  15. 15. Planned UX / UI Enhancements Whole Drawer image Unit tray recognition Multi- specimen annotation Plug in controlled vocabulary services
  16. 16. 1D & 2D barcode recognition Whole Drawer image • Recognition and reading of 1D & 2D matrix barcodes from images • Different physical requirements (smallest 6x6mm matrix, readable via handheld scanners) • Testing open source & commercial libraries at different scan resolutions Initial results • Commercial solutions outperform open source • Max. read success 94% • Idiosyncratic results (different results on different OS) • Testing continues…
  17. 17. The Holy Grail: label imaging & text recognition Whole Drawer image Chauliodes pectinicornis DorsalCaudalFrontal Reconstructed labels
  18. 18. The Holy Grail: label imaging & text recognition Whole Drawer image Agulla astuta DorsalCaudalLateral Reconstructed labels
  19. 19. Approaches to label imaging pinned specimens Whole Drawer image Could be incorporated as part of the barcode dispensing process 1. Barcode dispenser & scanner (two sides barcode labels) 2. Freshly pinned barcode label 3. Other collection labels 4. Multiple label imaging cameras 5. Assemble labels from composite images
  20. 20. Acknowledgements Whole Drawer image Segmentation algorithm & app. development Stefan van der Walt and Pieter Holtzhausen Application development Alice Heaton Barcode recognition & testing Lawrence Hudson Analysis & testing Laurence Livermore, Vladimir Blagoderov and Ben Price Initial specification and funding Vince Smith

×