Classification of barcode tag conditions from top view sample tube images for laboratory automation

Patent metadata

Field	Value
Publication number	US-10325182-B2
Application number	US-201615551566-A
Country	US
Kind code	B2
Filing date	Feb 16, 2016
Priority date	Feb 17, 2015
Publication date	Jun 18, 2019
Grant date	Jun 18, 2019

Abstract

Embodiments are directed to classifying barcode tag conditions on sample tubes from top view images to streamline sample tube handling in advanced clinical laboratory automation systems. The classification of barcode tag conditions leads to the automatic detection of problematic barcode tags, allowing for a user to take necessary steps to fix the problematic barcode tags. A vision system is utilized to perform the automatic classification of barcode tag conditions on sample tubes from top view images. The classification of barcode tag conditions on sample tubes from top view images is based on the following factors: (1) a region-of-interest (ROI) extraction and rectification method based on sample tube detection; (2) a barcode tag condition classification method based on holistic features uniformly sampled from the rectified ROI; and (3) a problematic barcode tag area localization method based on pixel-based feature extraction.

First claim

We claim:

1. A method of classifying barcode tag conditions on sample tubes held in a tube tray, the method comprising:
acquiring, by an image capture system comprised of at least one camera, top view image sequences of the tube tray; and
analyzing, by one or more processors in communication with the image capture system, the top view image sequences, the analyzing comprising, for each sample tube:
rectifying a region of interest (ROI) from each input image of the top view image sequences;
extracting features from the rectified ROI; and
inputting the extracted features from the rectified ROI into a classifier to determine the barcode tag condition, the barcode tag condition based upon a barcode tag condition category stored in the classifier;

wherein the classifier comprises a pixel-based classifier trained to localize and segment the ROI with visible deformation, and
the localization and segmentation of the ROI is performed on each pixel in the ROI to determine a likelihood that a particular pixel belongs to a problematic area.

2. The method of claim 1 , wherein the analyzing by the one or more processors further comprises:
if the determined barcode tag condition comprises a problematic identification, localizing a problematic area and issuing a warning via an output device in communication with the one or more processors.

3. The method of claim 1 , wherein the tube tray is configured to fit within a portion of a drawer movable between an open and a closed position.

4. The method of claim 3 , wherein the top view image sequences of the tube tray comprises images of the tube tray at predetermined positions in the drawer.

5. The method of claim 1 , wherein rectifying a region of interest (ROI) comprises rectifying the ROI to a canonical geometric orientation.

6. The method of claim 1 , wherein the ROI for a particular sample tube comprises a region including the particular sample tube from a top portion of the particular sample tube to a surface area of the tube tray to a given region extending outward from the particular sample tube, wherein the given region comprises the barcode tag for the particular sample tube.

7. The method of claim 1 , wherein the barcode tag conditions are grouped into a predetermined number of main categories, each of the main categories comprising a plurality of subcategories.

8. A vision system for use in an in vitro diagnostics environment for classifying barcode tag conditions on sample tubes held in a tube tray, the vision system comprising:
a surface configured to receive the tube tray, wherein the tube tray comprises a plurality of slots, each configured to receive a sample tube;
at least one camera configured to capture top view image sequences of the tube tray positioned on the surface; and
a processor in communication with the at least one camera, the processor configured to perform the following steps for each sample tube:
rectify a region of interest (ROI) from each input image of the top view image sequences;
extract features from the rectified ROI; and
input the extracted features from the rectified ROI into a classifier to determine the barcode tag condition, the barcode tag condition based upon a barcode tag condition category stored in the classifier,

wherein the classifier comprises a pixel-based classifier trained to localize and segment the ROI with visible deformation, and
the localization and segmentation of the ROI is performed on each pixel in the ROI to determine a likelihood that a particular pixel belongs to a problematic area.

9. The system of claim 8 , wherein the processor is further configured to:
if the determined barcode tag condition comprises a problematic identification, localize a problematic area and issue a warning via an output device in communication with the processor.

10. The system of claim 8 , wherein the surface comprises a drawer movable between an open and a closed position.

11. The system of claim 10 , wherein the top view image sequences of the tube tray comprises images of the tray at predetermined positions in the drawer.

12. The system of claim 8 , wherein rectifying a region of interest (ROI) comprises rectifying the ROI to a canonical geometric orientation.

13. The system of claim 8 , wherein the ROI for a particular sample tube comprises a region including the particular sample tube from a top portion of the particular sample tube to a surface area of the tube tray to a given region extending outward from the particular sample tube, wherein the given region comprises the barcode tag for the particular sample tube.

14. The system of claim 8 , wherein the barcode tag conditions are grouped into a predetermined number of main categories, each of the main categories comprising a plurality of subcategories.

Assignees

• Siemens Healthcare Diagnostics Inc

Inventors

• Soomro Khurram
• Chang Yao-Jen
• Kluckner Stefan
• Wu Wen
• Pollack Benjamin
• Chen Terrence

Classifications

• G06V10/774 Primary

  Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting · CPC title

• G06F18/2411

  based on the proximity to a decision surface, e.g. support vector machines · CPC title

• G06F18/2415

  based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate · CPC title

• G06F18/214

  Generating training patterns; Bootstrap methods, e.g. bagging or boosting · CPC title

• G06V10/25

  Determination of region of interest [ROI] or a volume of interest [VOI] · CPC title