Produce and Non-produce Verification Using Hybrid Scanner

• US 20150054959A1
• Filed: 08/26/2013
• Published: 02/26/2015
• Est. Priority Date: 08/26/2013
• Status: Active Grant

• Current Assignee: NCR Voyix Corporation
• Original Assignee: NCR Corporation (NCR Voyix Corporation)
• Inventors: He, Chao, Yu, Meng, Quipanes, Richard C., Parac, Em, Amora, Kristoffer Dominic
• Granted Patent: US 10,650,232 B2
• Time in Patent Office: Days
• US Class Current: 348/150
• CPC Class Codes: G06F 18/00 Pattern recognition
• CPC Class Codes: G06F 2218/08 Feature extraction
• CPC Class Codes: G06V 10/40 Extraction of image or video features
• CPC Class Codes: G06V 10/70 using pattern recognition or machine learning optical pattern recognition or electronic computations therefor G06V10/88
• CPC Class Codes: G06V 20/68 Food, e.g. fruit or vegetables

Abstract

Approaches to produce verification are addressed which may be suitably employed in conjunction with a low resolution image scanner. Features, such as lines, corners, text, texture, edges and the like are extracted and compared with features for an identified item or evaluated to determine if the features collectively are indicative of a produce item or a non-produce item.

Claims

1. Apparatus for verifying whether an item is produce or not comprising:

a scale utilized to weigh produce items;

a camera for generating an image of the item when the item is placed within a field of view of the camera on the scale;

an illumination mechanism for selectively illuminating the field of view of the camera;

a produce verification model database storing data for produce items to be verified; and

a processor for controlling the camera and the illumination mechanism, and retrieving a model for a particular item of produce to be weighed from the produce verification model database, wherein the processor controls the camera to capture image data for the item;

extracts features from the captured image data;

fits the extracted features to a retrieved model for the particular item of produce; and

verifies the item as produce or not based upon the fit of the extracted features to the retrieved model.

2. The apparatus of claim 1 wherein the processor further operates to verify the item is the particular item or not.

3. The apparatus of claim 1 further comprising:

a touchscreen, and wherein the particular item is identified by a selection by a user from a menu displayed on the touchscreen.

4. The apparatus of claim 1 wherein if the item is not identified as produce, the processor communicates a potential error message to an attendant display station.

5. The apparatus of claim 1 wherein the illumination mechanism comprises a plurality of light emitting diodes (LEDs);

and the processor controls the camera and LEDs to generate illuminated and non-illuminated background images, and illuminated and non-illuminated item images.

6. The apparatus of claim 5 wherein the processor computes a first difference of the two background images, bksub, and a second difference of the two item images, imgsub.

7. The apparatus of claim 6 wherein the processor subtracts bksub from imgsub to obtain, img, a difference of the first and second differences.

8. The apparatus of claim 7 wherein the camera comprises a low resolution gray scale camera also utilized to perform reading of two dimensional bar codes.

9. The apparatus of claim 8 wherein the processor generates a binary item segmentation map by thresholding img.

10. The apparatus of claim 1 wherein the processor extracts line features, corner features, edge features and texture features and applies feature detectors to select a group of most discriminating features.

11. Apparatus for categorizing an item as produce or non-produce comprising:

a camera for generating an image of the item when the item is placed within a field of view of the camera on the scale;

a categorization database storing data for produce and non-produce items to be categorized; and

a processor for controlling the camera, and retrieving product categorization criteria from the categorization database, wherein the processor controls the camera to capture image data for the item;

extracts features from the captured image data;

applies a retrieved categorization criteria; and

categorizes the item as produce or non-produce based upon the application of the categorization criteria to the extracted features from the captured image data.

12. The apparatus of claim 11 further comprising:

an illumination mechanism for selectively illuminating the field of view of the camera, and wherein the processor operates to control the illumination mechanism.

13. The apparatus of claim 12 wherein the illumination mechanism comprises a plurality of light emitting diodes (LEDs);

and the processor controls the camera and LEDs to generate illuminated and non-illuminated background images, and illuminated and non-illuminated item images.

14. The apparatus of claim 13 wherein the processor computes a first difference of the two background images, bksub, and a second difference of the two item images, imgsub.

15. The apparatus of claim 14 wherein the processor subtracts bksub from imgsub to obtain, img, a difference of the first and second differences.

16. The apparatus of claim 15 wherein the camera comprises a low resolution gray scale camera also utilized to perform reading of two dimensional bar codes.

17. The apparatus of claim 16 wherein the processor generates a binary item segmentation map by thresholding img.

18. The apparatus of claim 17 wherein the processor extracts item pixels using the segmentation map.

19. The apparatus of claim 18 wherein the processor extracts line features, corner features, edge features and texture features and applies feature detectors to select a group of most discriminating features.

20. The apparatus of claim 19 wherein the processor applies a probabilistic approach to classification and further classifies the item into a sub-category.