System and method for centralized, automatic extraction of data from remotely transmitted forms
View Patent ↗One or more parameters such as time is reported to a center using a form. The center then automatically extracts data from the form and converts it for storage and subsequent processing or review. The form is preferably a physical form, which the sender fills in and then faxes to the center. Image capture, registration, and feature recognition routines are included to enable the center to extract the reported data with no need for data reentry or human intervention. The invention is particularly suitable for employees who need to submit time sheets, expense reports, and the like to an employer, or for automatic analysis of submitted inventory reports, product orders, etc.
1. A method for collecting reports of at least one parameter comprising the following steps:
all in a central computer system:
automatically receiving from any of a plurality of arbitrary senders, via a publicly accessible transmission channel, an electronic representation of an image of a physical form, the form having a plurality of data fields, each corresponding to an indicator, which may be alphanumeric, of at least a partial value of at least one of the parameters;
automatically and uniquely identifying the physical form from the electronic representation of its received image;
automatically identifying the location of the data fields in the received representation of the image of the form by comparing the received electronic representation of the image of the physical form with at least one pre-stored electronic representation of at least one template;
automatically extracting from the identified data fields the at least partial values of the corresponding parameters; and
automatically storing the extracted values in a predetermined format in a memory for subsequent processing as well as the representation of the received physical form as it was received; and
transferring the stored extracted values to an external recipient via a network, all processing of the physical form after transmission by the sender up to and including transfer to the external recipient via the network thereby taking place automatically;
in which:
the electronic representation of the image of the physical form is generated using a standard, conventional facsimile machine, whereby the transmission channel is a standard telephone line and the central computer system is separate from the facsimile machine other than though its connection via the transmission channel;
each data field indicates a quantifiable or itemizable value of a corresponding one of the parameters, further including the additional step of storing the received electronic representation of the image of the physical form in the memory, whereby non-quantifiable and non-itemizable entries by the user onto the physical form are made available for subsequent review;
the step of automatically identifying the location of the data fields comprises the following sub-steps:
storing an electronic representation of a template of each of a plurality of physical forms;
automatically identifying each received form by performing a best-fit comparison of each received electronic representation of the image of the corresponding physical form with the stored electronic representations of the templates;
automatically registering the received electronic representation of the received physical form image with the best-fit electronic template representation; and
matching the data fields in the received electronic representation of the received physical form image with corresponding data fields in the best-fit electronic template representation.