Systems and methods for classifying payment documents during mobile image processing
Systems and methods are provided for processing an image of a financial payment document captured using a mobile device and classifying the type of payment document in order to extract the content therein. These methods may be implemented on a mobile device or a central server, and can be used to identify content on the payment document and determine whether the payment document is ready to be processed by a business or financial institution. The system can identify the type of payment document by identifying features on the payment document and performing a series of steps to determine probabilities that the payment document belongs to a specific document type. The identification steps are arranged starting with the fastest step in order to attempt to quickly determine the payment document type without requiring lengthy, extensive analysis.
1. A method comprising using at least one hardware processor to:
receive an image to be classified as one of a plurality of types of payment document; and
extract one or more features from the image, wherein the one or more features comprise at least one geometric characteristic of the image, and wherein the at least one geometric characteristic of the image comprises an aspect ratio; and
classify the image using a classification algorithm that comprises a plurality of classifiers, wherein the plurality of classifiers include at least a first classifier and a second classifier that requires a more time-consuming computation than the first classifier, wherein at least one of the plurality of classifiers eliminates one or more of the plurality of types of payment document based on the aspect ratio, and wherein classifying the image comprises,
prior to executing the second classifier, executing the first classifier,
when the first classifier classifies the image as one of the plurality of types of payment document with a probability that exceeds a threshold, classifying the image as the certain one type of payment document without executing the second classifier, and,
when the first classifier is unable to classify the image as one of the plurality of types of payment document with a probability that exceeds the threshold, executing the second classifier.
2. The method of claim 1 , further comprising using the at least one hardware processor to, before classifying the image, crop, correct, and binarize the image.
3. The method of claim 1 , wherein the one or more features comprise a magnetic ink character recognition (MICR) line.
4. The method of claim 1 , wherein the one or more features comprise an optical character recognition type A (OCR-A) codeline.
5. The method of claim 1 , wherein the one or more features comprise a barcode.
6. The method of claim 1 , wherein the one or more features comprise a check number.
7. The method of claim 1 , wherein the one or more features comprise a lock icon.
8. The method of claim 1 , wherein at least one of the plurality of classifiers:
determines a position of a component in the image; and
compares the determined position of the component to known positions of the component in one or more of the plurality of types of payment document to identify at least one of the plurality of types of payment document for which the known position matches the determined position.
9. The method of claim 8 , wherein the component is a lock icon.
10. The method of claim 1 , wherein at least one of the plurality of classifiers:
identifies a magnetic ink character recognition (MICR) line in the image; and
determines a probability that the image is of one or more of the plurality of types of payment document based on one or more characteristics of the MICR line.
11. The method of claim 1 , wherein at least one of the plurality of classifiers identifies a specific linear pattern, associated with one of the plurality of types of payment document, in the image.
12. The method of claim 1 , wherein the second classifier:
performs optical character recognition on the image to recognize one or more character strings; and
analyzes the one or more character strings to classify the image.
13. The method of claim 1 , wherein the plurality of types of payment document comprises a personal check, a business check, and one or more categories of irregular check.
14. The method of claim 1 , wherein the first classifier does not utilize keywords obtained by optical character recognition (OCR) performed on the image, and wherein the second classifier does utilize keywords obtained by OCR performed on the image.
15. A system comprising:
at least one hardware processor; and
one or more software modules that are configured to, when executed by the at least one hardware processor,
receive an image to be classified as one of a plurality of types of payment document,
extract one or more features from the image, wherein the one or more features comprise at least one geometric characteristic of the image, and wherein the at least one geometric characteristic of the image comprises an aspect ratio, and
classify the image using a classification algorithm that comprises a plurality of classifiers, wherein the plurality of classifiers include at least a first classifier and a second classifier that requires a more time-consuming computation than the first classifier, wherein at least one of the plurality of classifiers eliminates one or more of the plurality of types of payment document based on the aspect ratio, and wherein classifying the image comprises,
prior to executing the second classifier, executing the first classifier,
when the first classifier classifies the image as one of the plurality of types of payment document with a probability that exceeds a threshold, classifying the image as the certain one type of payment document without executing the second classifier, and,
when the first classifier is unable to classify the image as one of the plurality of types of payment document with a probability that exceeds the threshold, executing the second classifier.
16. A non-transitory computer-readable medium having instructions stored thereon, wherein the instructions, when executed by a processor, cause the processor to:
receive an image to be classified as one of a plurality of types of payment document;
extract one or more features from the image, wherein the one or more features comprise at least one geometric characteristic of the image, and wherein the at least one geometric characteristic of the image comprises an aspect ratio; and
classify the image using a classification algorithm that comprises a plurality of classifiers, wherein the plurality of classifiers include at least a first classifier and a second classifier that requires a more time-consuming computation than the first classifier, wherein at least one of the plurality of classifiers eliminates one or more of the plurality of types of payment document based on the aspect ratio, and wherein classifying the image comprises,
prior to executing the second classifier, executing the first classifier,
when the first classifier classifies the image as one of the plurality of types of payment document with a probability that exceeds a threshold, classifying the image as the certain one type of payment document without executing the second classifier, and,
when the first classifier is unable to classify the image as one of the plurality of types of payment document with a probability that exceeds the threshold, executing the second classifier.