Scanning recognition decisions
Images of scanned documents are processed to identify an amount for each document. At least one document has two possible amounts with each amount having a different confidence level associated therewith. A provided total amount is balanced to a determined total amount by resolving a permutation of the available amounts from the images that matches the provided total amount.
1. A method, comprising:
obtaining, by executable instructions that execute on a processor as a balance manager from a non-transitory computer-readable storage medium, a balance total for a plurality of documents;
receiving, by the balance manager, images, each image representing a particular document;
identifying, by the balance manager, at least one amount for each image, wherein at least one image includes two or more selectable amounts; and
resolving, by the balance manager, a permutation of the amounts for the images that represents a total amount and that equals the balance total by serially presenting each possible combination that equals the balance total to a user in an interface for selection of one of the possible combinations as the permutation.
2. The method of claim 1 , wherein receiving further includes obtaining the images from a scanner interfaced to the method.
3. The method of claim 2 , wherein receiving further includes providing each image to an Optical Character Recognition (OCR) engine.
4. The method of claim 3 , wherein identifying further includes obtaining the amounts from the OCR engine.
5. The method of claim 4 , wherein obtaining further includes obtaining a unique confidence value for each of the two or more selectable amounts.
6. The method of claim 5 , wherein resolving further includes initially selecting one of the two or more selectable amounts based on the selected one having a higher confidence value than remaining ones of the two or selectable amounts.
7. The method of claim 6 , wherein initially selecting further includes selecting a different one of the two or more selectable amounts based on a difference between the balance total and a running total for the total amount.
8. The method of claim 1 , wherein resolving further includes maintaining differences between a running total for the total amount and specific selections of the two or more selectable amounts.
9. The method of claim 8 further comprising, selecting, by the balance manager, a particular one of the two or more selectable amounts based on a specific difference that makes the running total equal the balance total.
10. The method of claim 1 , wherein resolving further includes iterating with different selections of the two or more selectable amounts until the total amount is equal to the balance total.
11. A method, comprising:
maintaining, by executable instructions that execute on a processor as a balancer from a non-transitory computer-readable storage medium, a combination of selectable amounts for each document image of a plurality of document images;
assigning, by the balancer, a unique confidence level to each selectable amount;
deriving, by the balancer, a permutation of selectable amounts with just one selectable amount selected for each document image that when totaled equals an inputted balance total by processing the unique confidence levels, wherein deriving further includes identifying possible combinations for the selectable amounts, each possible combination equaling the inputted balance total and presenting the possible combinations in an interface for selection by a user to derive the permutation.
12. The method of claim 11 further comprising processing the balancer as a balance operation for a plurality of deposited physical checks, wherein each physical check is scanned to provide a particular document image.
13. The method of claim 11 , wherein maintaining further includes receiving the combinations of amounts for the document images from an OCR engine.
14. The method of claim 11 , wherein deriving further includes iterating the deriving to find the possible permutations.
15. The method of claim 14 , wherein iterating further includes selecting during a first iteration a particular amount from a particular combination of selectable amounts based on that particular amount having a higher corresponding unique confidence level over other selectable amounts available from the particular combination.
16. The method of claim 15 , wherein selecting further includes selecting during a subsequent iteration a different amount from the particular combination based on a difference between the different amount and the particular amount balancing with the inputted balance total.
17. The method of claim 11 , wherein deriving further includes displaying a specific amount within a context of a particular document image for user correction based on the specific amount having a lowest confidence level from remaining confidence levels associated with other amounts when no permutation exists that totals to the inputted balanced total.
18. A terminal, comprising:
a hardware processor configured to execute executable instructions representing a decision maker;
a non-transitory computer-readable storage medium having the executable instructions that are executed by the processor;
the decision maker;
wherein the decision maker is configured to: maintain selectable amounts for each of a plurality of document images, and select a permutation of the selectable amounts with just one selectable amount per document image where a total amount for the permutation balances with an inputted total for a total amount of the document images by identifying possible combinations for the selectable amounts that balances with the inputted total and presenting the possible combinations in an interface to a user for selecting one of the possible combinations as the permutation.
19. The SST of claim 18 , wherein the terminal is one of: a Self-Service Terminal (SST), an Automated Teller Machine (ATM), a kiosk, a server, a cloud, a laptop, a mobile phone, a wearable processing device, a desktop, and a tablet.
20. The SST of claim 19 , wherein the terminal is one of: interfaced over a network connection to a scanner to receive the document images and interfaced over an internal connection to an integrated scanner that is integrated into the terminal for receiving the document images.