Machine-learning for enhanced machine reading of non-ideal capture conditions
View Patent ↗Implementations of the present disclosure include receiving a training image, providing a hash pattern that is representative of the training image, applying a plurality of filters to the training image to provide a respective plurality of filtered training images, identifying a filter to be associated with the hash pattern based on the plurality of filtered training images, and storing a mapping of the filter to the hash pattern within a set of mapping in a data store.
1. A computer-implemented method executed by at least one processor, the method comprising:
maintaining a mapping of filters to hash patterns in a data store;
receiving a transaction image captured by a computing device, the transaction image representing a document to be used in executing a transaction;
processing the transaction image to generate a transaction hash pattern for the entire transaction image, comprising:
scaling the transaction image that represents the document by using at least one of the following scaling techniques: sinc resampling, box sampling, or mipmap to generate a micro-image of the transaction image,
applying a convolution edge filter to the micro-image generated by scaling the transaction image to generate a filtered micro-image, wherein the filtered micro-image includes one or more edges of one or more objects within the micro-image, and wherein the one or more objects include text,
applying bi-level encoding to the filtered micro-image to generate an encoded micro-image, and
generating the transaction hash pattern from the entire encoded micro-image;
identifying a first filter to be applied to the transaction image based on the mapping, wherein identifying the first filter to be applied to the transaction image based on the mapping comprises determining that the transaction hash pattern sufficiently matches a first hash pattern mapped to the first filter;
providing a filtered transaction image using the first filter;
determining text data from the filtered transaction image; and
executing the transaction with the determined text data from the filtered transaction image.
2. The method of claim 1 , wherein the transaction hash pattern comprises a binary string.
3. The method of claim 1 , wherein determining that the transaction hash pattern sufficiently matches the first hash pattern mapped to the first filter comprises calculating a Hamming distance between the transaction hash pattern and the first hash pattern, and comparing the Hamming distance to a threshold Hamming distance.
4. The method of claim 1 , wherein determining text data is performed by optical character recognition (OCR) of the filtered transaction image.
5. The method of claim 1 , wherein the document comprises a negotiable instrument.
6. The method of claim 5 , wherein the transaction is to deposit the negotiable instrument at a financial institution.
7. A system comprising:
a data store for storing data; and
at least one processor configured to interact with the data store, the at least one processor being further configured to execute computer-readable instructions to perform operations comprising:
maintaining a mapping of filters to hash patterns in a data store;
receiving a transaction image captured by a computing device, the transaction image representing a document to be used in executing a transaction;
processing the transaction image to generate a transaction hash pattern for the entire transaction image, comprising:
scaling the transaction image that represents the document by using at least one of the following scaling techniques: sinc resampling, box sampling, or mipmap to generate a micro-image of the transaction image,
applying a convolution edge filter to the micro-image generated by scaling the transaction image to generate a filtered micro-image, wherein the filtered micro-image includes one or more edges of one or more objects within the micro-image, and wherein the one or more objects include text,
applying bi-level encoding to the filtered micro-image to generate an encoded micro-image, and
generating the transaction hash pattern from the entire encoded micro-image;
identifying a first filter to be applied to the transaction image based on the mapping, wherein identifying the first filter to be applied to the transaction image based on the mapping comprises determining that the transaction hash pattern sufficiently matches a first hash pattern mapped to the first filter;
providing a filtered transaction image using the first filter;
determining text data from the filtered transaction image; and
executing the transaction with the determined text data from the filtered transaction image.
8. The system of claim 7 , wherein the transaction hash pattern comprises a binary string.
9. The system of claim 7 , wherein the operations for determining that the transaction hash pattern sufficiently matches the first hash pattern mapped to the first filter comprise calculating a Hamming distance between the transaction hash pattern and the first hash pattern, and comparing the Hamming distance to a threshold Hamming distance.
10. The system of claim 7 , wherein determining text data is performed by optical character recognition (OCR) of the filtered transaction image.
11. The system of claim 7 , wherein the document comprises a negotiable instrument.
12. The system of claim 11 , wherein the transaction is to deposit the negotiable instrument at a financial institution.
13. A non-transitory computer-readable storage medium coupled to one or more processors and having instructions stored thereon which, when executed by the one or more processors, cause the one or more processors to perform operations comprising:
maintaining a mapping of filters to hash patterns in a data store;
receiving a transaction image captured by a computing device, the transaction image representing a document to be used in executing a transaction;
processing the transaction image to generate a transaction hash pattern for the entire transaction image, comprising:
scaling the transaction image that represents the document by using at least one of the following scaling techniques: sinc resampling, box sampling, or mipmap to generate a micro-image of the transaction image,
applying a convolution edge filter to the micro-image generated by scaling the transaction image to generate a filtered micro-image, wherein the filtered micro-image includes one or more edges of one or more objects within the micro-image, and wherein the one or more objects include text,
applying bi-level encoding to the filtered micro-image to generate an encoded micro-image, and
generating the transaction hash pattern from the entire encoded micro-image;
identifying a first filter to be applied to the transaction image based on the mapping, wherein identifying the first filter to be applied to the transaction image based on the mapping comprises determining that the transaction hash pattern sufficiently matches a first hash pattern mapped to the first filter;
providing a filtered transaction image using the first filter;
determining text data from the filtered transaction image; and
executing the transaction with the determined text data from the filtered transaction image.
14. The non-transitory computer-readable storage medium of claim 13 , wherein the transaction hash pattern comprises a binary string.
15. The non-transitory computer-readable storage medium of claim 13 , wherein the operations for determining that the transaction hash pattern sufficiently matches the first hash pattern mapped to the first filter comprise calculating a Hamming distance between the transaction hash pattern and the first hash pattern, and comparing the Hamming distance to a threshold Hamming distance.
16. The non-transitory computer-readable storage medium of claim 13 , wherein determining text data is performed by optical character recognition (OCR) of the filtered transaction image.
17. The non-transitory computer-readable storage medium of claim 13 , wherein the transaction is to deposit a negotiable instrument at a financial institution.