Produce identification and pricing system for checkouts

• US 6,363,366 B1
• Filed: 08/31/1998
• Issued: 03/26/2002
• Est. Priority Date: 08/31/1998
• Status: Expired due to Term

• Current Assignee: ITAB SCANFLOW AB
• Original Assignee: David L. Henty
• Inventors: Henty, David L.
• Primary Examiner(s): Trammell, James P.
• Assistant Examiner(s): Dixon, Thomas A.
• Application Number:
  US09/143,903
• Time in Patent Office: 1,303 Days
• Field of Search: 705/400, 705/16, 705/414, 177/25.15, 382/100, 382/110, 382/203, 382/141, 382/142, 382/143, 382/170, 382/199, 382/209
• US Class Current: 705/400
• CPC Class Codes: G01G 19/4144 for controlling weight of goods in commercial establishments, e.g. supermarket, P.O.S. systems
• CPC Class Codes: G06F 18/00 Pattern recognition
• CPC Class Codes: G06Q 30/0283 Price estimation or determination
• CPC Class Codes: G06V 20/68 Food, e.g. fruit or vegetables
• CPC Class Codes: G07G 1/0054 with control of supplementary check-parameters, e.g. weight or number of articles

Abstract

A produce identification and pricing system for a supermarket checkout or other retail outlet where produce is handled includes one or more imagers which provide image data from produce to be identified, which image data includes spectral information, preferably including information from the nearer infrared as well as RGB information. The spectral information is transformed into a signature vector which is compared with target signature vectors for the variety of types of produce handled by the supermarket or other retail outlet. The best match to the target signature vectors is employed to obtain price information from a produce price database. If the produce price is based on weight, information from a scale is retrieved and the total price is calculated. If the price is based on the number of items, a count operation is enabled and the count value determined. The price is then determined based on the count value of the number of items. The total price is output to a register for computation of the total checkout price. The produce identification information is also forwarded to a database for inventory monitoring of the produce in the store. The produce identification system may be combined with a more conventional bar code scanning system at single imaging station to provide automated handling of identification and pricing of all the items at the retail outlet without requiring any additional obtrusive hardware.

Claims

1. A method for identification and pricing of produce comprising:

receiving image data from the produce to be identified;

transforming the image data to a signature vector characterizing the produce to be identified, the signature vector including at least spectral information from the produce to be identified;

comparing the signature vector to stored signature vectors representing a target set of produce types and identifying the produce based on the best match with the stored signature vectors;

accessing a price information database and retrieving price information for the identified produce type;

determining if the identified produce price is based on weight and retrieving the weight of the produce from a scale if the price is based on weight;

computing the total price of the produce based on the price information for the identified produce and the total weight of the produce, if the price is based on weight;

determining if the identified produce price is based on the number of items; and

automatically counting the number of items of produce if the price is based on the number of items.

2. A method for identification and pricing of produce, comprising:

receiving image data from the produce to be identified;

transforming the image data to a signature vector characterizing the produce to be identified, the signature vector including at least spectral information from the produce to be identified;

comparing the signature vector to stored signature vectors representing a target set of produce types and identifying the produce based on the best match with the stored signature vectors;

accessing a price information database and retrieving price information for the identified produce type;

determining if the identified produce price is based on weight and retrieving the weight of the produce from a scale if the price is based on weight; and

computing the total price of the produce based on the price information for the identified produce and the total weight of the produce;

wherein transforming the image data to a signature vector further comprises transforming two-dimensional shape information from the image data to a signature vector, wherein said two-dimensional shape information comprises the estimated area of the imaged object.

3. A method for identification and pricing of produce as set out in claim 2, wherein transforming the image data to a signature vector further comprises transforming three-dimensional shape information from the image data to a signature vector.

4. A method for identification and pricing of produce as set out in claim 2, wherein said two-dimensional shape information comprises a histogram of deviations of the imaged object'"'"'s boundaries from a best fit ellipse.

5. A method for identification and pricing of produce, comprising:

receiving image data from the produce to be identified;

transforming the image data to a signature vector characterizing the produce to be identified, the signature vector including at least spectral information from the produce to be identified;

comparing the signature vector to stored signature vectors representing a target set of produce types and identifying the produce based on the best match with the stored signature vectors;

determining a confidence value associated with the identification of the produce type;

accessing a price information database and retrieving price information for the identified produce type;

determining if the identified produce price is based on weight and retrieving the weight of the produce from a scale if the price is based on weight; and

computing the total price of the produce based on the price information for the identified produce and the total weight of the produce.

6. A method for identification and pricing of produce as set out in claim 5, further comprising:

comparing the confidence value with a predetermined threshold confidence value to determine if the identification is potentially inaccurate; and

if the confidence value is less than the predetermined threshold, displaying a plurality of possible produce types.

7. A method for identification and pricing of produce as set out in claim 6, further comprising:

receiving an input from an operator selecting a correct produce type from the displayed plurality of produce types; and

correcting the stored signature vectors based on the selected produce type information provided from the operator.

8. A method for identification and pricing of produce as set out in claim 7, further comprising providing the corrected stored signature vectors to additional pricing systems configured in a network.

9. A produce identification and pricing system, comprising:

at least one imager for providing image data from produce to be identified, the image data including spectral information;

preprocessing means for receiving the image data from said at least one imager and providing a signature corresponding to spectral information contained in the image data;

processing means for comparing the signature to stored target signatures corresponding to target produce types and obtaining a best match identification of the produce type of the imaged produce;

means for storing price information for the target produce types;

processing means for counting the number of items of produce based on three-dimensional image data; and

means for calculating the total price of the produce based on the price information of the identified produce type and an additional input parameter including either weight or count information for the imaged produce.

10. A produce identification and pricing system as set out in claim 9, wherein said at least one imager comprises a first imager which provides spectral information in the near infrared region and a second imager.

11. A produce identification and pricing system as set out in claim 9, wherein said at least one imager comprises a CMOS imager.