Methods and apparatus adapted to identify a specimen container from multiple lateral views
A model-based method of determining characteristics of a specimen container. The method includes providing a specimen container, capturing images of the specimen container at different exposures times and at different spectra having different nominal wavelengths, selecting optimally-exposed pixels from the images at different exposure times at each spectra to generate optimally-exposed image data for each spectra, and classifying the optimally-exposed pixels as at least being one of tube, label or cap, and identifying a width, height, or width and height of the specimen container based upon the optimally-exposed image data for each spectra. Quality check modules and specimen testing apparatus adapted to carry out the method are described, as are other aspects.
1. A method of determining characteristics of a specimen container, comprising:
providing a specimen container;
capturing images of the specimen container at different exposures times and at different spectra having different nominal wavelengths;
selecting optimally-exposed pixels from the images at different exposure times at each spectra to generate optimally-exposed image data for each spectra;
classifying the optimally-exposed pixels as at least being one of tube, label or cap; and
identifying a width, height, or width and height of the specimen container based upon the optimally-exposed image data for each spectra.
2. The method of claim 1 , comprising identifying a cap type of the specimen container.
3. The method of claim 1 , comprising identifying a label of the specimen container.
4. The method of claim 1 , comprising identifying any region which is a holder.
5. The method of claim 1 , wherein the capturing images of the specimen container comprises capturing multiple images from a number of different viewpoints.
6. The method of claim 5 , wherein the number of different viewpoints comprises 3 or more.
7. The method of claim 1 , wherein the classifying the optimally-exposed pixels is based upon a multi-class classifier generated from multiple training sets.
8. The method of claim 7 , wherein the multi-class classifier comprises a support vector machine.
9. The method of claim 1 , comprising determining an inner width of the specimen container based upon the width and height of the specimen container.
10. A quality check module configured to determine characteristics of a specimen container, comprising:
a plurality of cameras arranged at multiple viewpoints around an imaging location configured to receive the specimen container, each of the plurality of cameras configured to capture multiple images of at least a portion of the specimen container at different exposures times and at different spectra having different nominal wavelengths from the multiple viewpoints; and
a computer coupled to the plurality of cameras, the computer configured to:
select optimally-exposed pixels from the images at the different exposure times at each of the different spectra to generate optimally-exposed image data for each spectra and viewpoint,
classify the optimally-exposed image data as at least being one of tube, cap, or label, and
identify a width, height, or width and height of the specimen container based upon the optimally-exposed image data for each spectra.
11. The quality check module of claim 10 , comprising a housing surrounding the imaging location.
12. The quality check module of claim 10 , comprising a back lighting surrounding the imaging location.
13. The quality check module of claim 10 , comprising front lighting surrounding the imaging location.
14. The quality check module of claim 10 , wherein the imaging location is on a track, the specimen container configured to be received in a receptacle of a carrier moveable on the track.
15. The quality check module of claim 10 , configured to identify a cap type of the specimen container.
16. The quality check module of claim 10 , configured to identify a label of the specimen container.
17. The quality check module of claim 10 , wherein the classifying the optimally-exposed image data is based upon a multi-class classifier generated from multiple training sets.
18. The quality check module of claim 17 , wherein the multi-class classifier comprises a support vector machine.
19. The quality check module of claim 10 , configured to determine an inner width of the specimen container based upon the width and height of the specimen container.
20. A specimen testing apparatus, comprising:
a track;
specimen carriers moveable on the track, the specimen carriers configured to carry specimen containers; and
a quality check module arranged on the track and configured to determine characteristics of a specimen container, the quality check module comprising:
a plurality of cameras arranged at multiple viewpoints around an imaging location configured to receive the specimen container, each of the plurality of cameras configured to capture multiple images of at least a portion of the specimen container at different exposures times and at different spectra having different nominal wavelengths from the multiple viewpoints; and
a computer coupled to the plurality of cameras, the computer configured to:
select optimally-exposed pixels from the images at the different exposure times at each of the spectra to generate optimally-exposed image data for each spectra and viewpoint,
classify the optimally-exposed image data as at least being one of tube, cap, or label, and
identify a width, height, or width and height of the specimen container based upon the optimally-exposed image data for each spectra.