TECHNIQUES FOR GENERATING AN ELECTRONIC SHOPPING LIST
Techniques are disclosed for generating an electronic shopping list based on a digital photograph of a handwritten shopping list. The techniques include obtaining a digital photograph of a handwritten shopping list and identifying a handwritten text portion in the digital photograph. The techniques further include determining a set of candidate items corresponding to the handwritten text portion, each candidate item of the set of candidate items representing a different item that may have written in the handwritten text portion of the handwritten shopping list and being indicative of a product sold by a retailer. The techniques further include selecting one of the set of candidate items for inclusion in the electronic shopping list. The techniques can be performed by a mobile computing device or by a shopping list server.
1 . A computer-implemented method for generating an electronic shopping list comprising:
obtaining, at a processing device, a digital photograph of a handwritten shopping list;
identifying, at the processing device, a handwritten text portion in the digital photograph;
determining, at the processing device, a set of candidate items corresponding to the handwritten text portion, each candidate item of the set of candidate items representing a different item that may have written in the handwritten text portion of the handwritten shopping list and being indicative of a product sold by a retailer; and
selecting, at the processing device, one of the set of candidate items for inclusion in the electronic shopping list.
2 . The method of claim 1 , further comprising:
inserting, at the processing device, the selected candidate item in the electronic shopping list; and
storing, by the processing device, the electronic shopping list in a memory device.
3 . The method of claim 1 , further comprising providing, at the processing device, the selected candidate item to a computing device, the mobile computing device having provided the digital photograph.
4 . The method of claim 1 , wherein determining the set of candidate items includes determining, at the processing device, an alphanumeric string based on the handwritten text portion, wherein the set of candidate items are determined based on the alphanumeric string and each candidate item is an approximate match to the alphanumeric string.
5 . The method of claim 4 , wherein determining the alphanumeric string includes performing optical character recognition on the handwritten text portion, the output of the optical character recognition being the alphanumeric string.
6 . The method of claim 4 , wherein the set of candidate items are determined from a product language model, the product language model being trained with words corresponding to products sold by the retailer.
7 . The method of claim 6 , wherein the words used to train the product language model includes generic products and specific products, the generic products indicating different types of products and specific products indicating different brands of products.
8 . The method of claim 6 , wherein the language model receives the alphanumeric string and determines one or more candidate items that are approximate matches to the alphanumeric string and a score corresponding to each candidate item, the score of each candidate item being indicative of a degree of likelihood that the alphanumeric string is the product indicated by the candidate item.
9 . The method of claim 8 , wherein the selected candidate item is selected based on its score, the selected candidate item having a higher score than the other candidate items in the set of candidate items.
10 . A computer-implemented method for generating an electronic shopping list, the method comprising:
capturing, by a camera of a mobile computing device, a digital photograph of a handwritten shopping list;
identifying, at a processing device of the mobile computing device, a handwritten text portion in the digital photograph;
determining, at the processing device, a set of candidate items corresponding to the handwritten text portion, each candidate item of the set of candidate items representing a different item that may have written in the handwritten text portion of the handwritten shopping list and being indicative of a product sold by a retailer; and
selecting, at the processing device, one of the set of candidate items for inclusion in the electronic shopping list;
generating, at the processing device, the electronic shopping list based on the selected candidate item; and
displaying, at the processing device, the electronic shopping list in a user interface of the mobile computing device.
11 . The method of claim 10 , further comprising:
inserting, at the processing device, the selected candidate item in the electronic shopping list; and
storing, by the processing device, the electronic shopping list in a memory device.
12 . The method of claim 11 , wherein determining the set of candidate items includes determining, at the processing device, an alphanumeric string based on the handwritten text portion, wherein the set of candidate items are determined based on the alphanumeric string and each candidate item is an approximate match to the alphanumeric string.
13 . The method of claim 12 , wherein determining the alphanumeric string includes performing optical character recognition on the handwritten text portion, the output of the optical character recognition being the alphanumeric string.
14 . The method of claim 12 , wherein the set of candidate items are determined from a product language model, the product language model being trained with words corresponding to products sold by the retailer.
15 . The method of claim 14 , wherein the words used to train the product language model includes generic products and specific products, the generic products indicating different types of products and specific products indicating different brands of products.
16 . The method of claim 14 , wherein the language model receives the alphanumeric string and determines one or more candidate items that are approximate matches to the alphanumeric string and a score corresponding to each candidate item, the score of each candidate item being indicative of a degree of likelihood that the alphanumeric string is the product indicated by the candidate item.
17 . The method of claim 10 , wherein selecting one of the one or more candidate items includes:
displaying, at the processing device, the one or more candidate items in the user interface; and
receiving, at the processing device, a user selection indicating the selected candidate item.
18 . A computer-implemented method for generating an electronic shopping list comprising:
receiving, at a processing device of a shopping list server, a digital photograph of a handwritten shopping list from a mobile computing device;
identifying, at the processing device, a handwritten text portion in the digital photograph;
determining, at the processing device, a set of candidate items corresponding to the handwritten text portion, each candidate item of the set of candidate items representing a different item that may have written in the handwritten text portion of the handwritten shopping list and being indicative of a product sold by a retailer; and
selecting, at the processing device, one of the set of candidate items for inclusion in the electronic shopping list;
generating, at the processing device, the electronic shopping list based on the selected candidate item; and
providing, the processing device, the electronic shopping list to the mobile computing device.
19 . The method of claim 18 , wherein determining the set of candidate items includes determining, at the processing device, an alphanumeric string based on the handwritten text portion, wherein the set of candidate items are determined based on the alphanumeric string and each candidate item is an approximate match to the alphanumeric string.
20 . The method of claim 19 , wherein the set of candidate items are determined from a product language model, the product language model being trained with words corresponding to products sold by the retailer, wherein the words used to train the product language model includes generic products and specific products, the generic products indicating different types of products and specific products indicating different brands of products.
21 . The method of claim 20 , wherein the language model receives the alphanumeric string and determines one or more candidate items that are approximate matches to the alphanumeric string and a score corresponding to each candidate item, the score of each candidate item being indicative of a degree of likelihood that the alphanumeric string is the product indicated by the candidate item.
22 . The method of claim 21 , wherein the selected candidate item is selected based on its score, the selected candidate item having a higher score than the other candidate items in the set of candidate items.