Method for image product recommendation

Patent metadata

Field	Value
Publication number	US-11741522-B2
Application number	US-201615395837-A
Country	US
Kind code	B2
Filing date	Dec 30, 2016
Priority date	Dec 31, 2015
Publication date	Aug 29, 2023
Grant date	Aug 29, 2023

Abstract

Embodiments of the present disclosure a method for determining product relevancy including extracting metadata from an image file of a digital image collection, the metadata being indicative of at least one feature of the image file. The method includes creating an input profile corresponding to the metadata extracted from the image files of the digital image collection. The method includes comparing the input profile to a product profile, the product profile having one or more rules corresponding to a consumer product, wherein the rules are indicative of the requirements of the product. The method includes determining a match score, the match score indicative of a relevancy of the product profile to the input profile such that a high relevancy correlates to a consumer product that is suited to the input profile and a low relevancy correlates to the consumer product that is not suited to the input profile.

First claim

The invention claimed is:

1. A method for creating a consumer photo product recommendation, comprising:
receiving, at an image processing unit, a digital image collection comprising a plurality of image files;
extracting recorded metadata from the image files, wherein the recorded metadata comprises location information and capture time information;
extracting content-based metadata indicative of one or more characterizing content profiles from the image files independent of other types of metadata, wherein the content-based metadata is derived from an analysis of image pixels of a background or scenery of each of the image files;
extracting face-based metadata from the image files, wherein the face-based metadata is derived from an analysis of image pixels in each of the image files;
extracting event-based metadata including a temporal duration of an event, number of images in the event, and event user tags from the image files, wherein the event-based metadata is derived from an analysis of image pixels in each of the image files;
creating an input profile for the digital image collection defined by the extracted recorded metadata, extracted content-based metadata, extracted face-based metadata, and extracted event-based metadata;
comparing the input profile to a plurality of product profiles, wherein each product profile corresponds to a different consumer photo product and comprises one or more product profile rules corresponding to each respective consumer photo product, wherein the product profile rules define features of each respective consumer photo product, and wherein each product profile rule is configured to provide a reward if it is satisfied, or a penalty or no value if it is not satisfied;
determining a match score for each product profile compared to the input profile, wherein the match score is indicative of a relevancy of the product profile to the input profile and a higher match score correlates to a higher relevancy, and wherein the match score is determined in part by evaluating whether the input profile satisfies the product profile rules; and
displaying, on an interface of the image processing unit, a rendering of the consumer photo products corresponding to the product profiles having a match score above a pre-established threshold.

2. The method of claim 1 , wherein determining the match score for each product profile compared to the input profile comprises increasing the match score when a product profile rule is met by the input profile.

3. The method of claim 1 , wherein determining the match score for each product profile compared to the input profile comprises decreasing the match score when a product profile rule is not met by the input profile.

4. The method of claim 1 , further comprising receiving the product profiles from a pre-loaded catalog on a server.

5. The method of claim 1 , further comprising displaying, on the interface of the image processing unit, the consumer product associated with the product profile having the highest match score.

6. The method of claim 1 , wherein displaying a rendering of the consumer photo products comprises listing the consumer products from highest match score to lowest match score.

7. The method of claim 1 , wherein determining the match score for each product profile compared to the input profile comprises neither increasing nor decreasing the match score when a product profile rule is not met by the input profile.

Assignees

• Kodak Alaris Inc

Inventors

• Loui Alexander
• Das Madirakshi

Classifications

• G06V2201/10

  Recognition assisted with metadata · CPC title

• G06V20/35

  Categorising the entire scene, e.g. birthday party or wedding scene · CPC title

• G06Q30/0631 Primary

  Recommending goods or services · CPC title

• G06F16/583

  using metadata automatically derived from the content · CPC title

• G06Q30/0603

  Catalogue creation or management · CPC title