Granted Patent B2

US 12,265,988 · App. 17/469,699 · Granted Apr 1, 2025

Inventors: Matthew McLaughlin (Severna Park, MD); Roy Kalman Rosenfeld (Jerusalem, IL); Lahat Michael Abu (New York, NY); Lior Simon (New York, NY)

Assignee: DoubleVerify, Inc.

G06Q30/0248G06F16/951G06Q30/0201G06Q30/0251G06Q30/0277H04L63/1408H04L9/40

Abstract

Claims (48)

1. A method performed by a content analysis server for identifying hidden content, the method comprising:

obtaining rendered code for a digital display page using any of a web crawler implemented on the content analysis server to download page content and cause a browser to render the digital display page, a piece of hypertext markup language (html) code obtained via the web crawler to access a document object module (DOM) for the digital display page, or a plugin implemented on any of the content analysis server or an end user device to access the digital display page rendered on the browser;

analyzing the rendered code for the digital display page to identify one or more objects embedded in the digital display page by inspecting data types in the DOM for known types that correspond with objects and inspecting signatures and/or content sizes in the DOM that correspond with objects;

extracting graphic images displayed in the objects included in the digital display page and identified from the DOM;

creating a repository of the extracted graphic images at a memory of the content analysis server, wherein the extracted graphic images are arranged in the repository by at least one characteristic, and wherein the repository is configured to store extracted graphic images that are visible or invisible on the digital display page;

rendering the rendered code for the digital display page at the content analysis server using the DOM to cause display of a rendered version of digital display page and saving the rendered version of the digital display page as a rendered page image file at the memory of the content analysis server;

performing a geometric analysis of the digital display page by:

calculating, for each extracted graphic image, a location of the extracted graphic image on the digital display page based on inspecting the DOM;

identifying a viewable portion of the digital display page; and

determining, for each extracted graphic image, whether the extracted graphic image is outside the viewable portion of the digital display page, wherein all extracted graphic images outside the viewable portion of the digital display page are marked as invisible images in the repository;

determining, for each extracted graphic image, whether a threshold amount of CPU resources are used for optimizing each extracted graphic image, wherein each extracted graphic image that has the threshold amount of CPU resources used to optimize the extracted graphic image is marked as visible images in the repository;

comparing the extracted graphic images stored in the repository to the rendered page image file to identify all extracted graphic images from the repository that are included within the rendered page image file by using an image analysis routine to determine whether a threshold number of pixels of each extracted graphic image are represented in the rendered page image file, wherein extracted graphic images that are identified in the rendered page image are marked as visible images in the repository, and extracted graphic images that are not identified in the rendered page image are marked as invisible images in the repository; and

when one or more extracted graphical images are marked as invisible images, classifying the digital display page as containing hidden content.

2. The method of claim 1 , wherein the digital display page is classified as containing an invisible object if any extracted graphic image of the digital display page is marked as an invisible image.

3. The method of claim 1 , wherein, when there is identified an animated object in a digital display page, each frame of the animation is extracted as a separate graphic image of the animated object, and the animated object is marked as a visible object if any of the separate graphic images of the animated object is identified in the rendered page image as a visible image.

4. A system for identifying hidden content comprising a processor, memory accessible by the processor, and program instructions and data stored in the memory, the program instructions executable by the processor to perform a method of comprising:

obtaining rendered code for a digital display page using any of a web crawler implemented on the system to download page content and cause a browser to render the digital display page, a piece of hypertext markup language (html) code obtained via the web crawler to access a document object module (DOM) for the digital display page, or a plugin implemented on any of the system or an end user device to access the digital display page rendered on the browser;

analyzing the rendered code for the digital display page to identify one or more objects embedded in the digital display page by inspecting data types in the DOM for known types that correspond with objects and inspecting signatures and/or content sizes in the DOM that correspond with objects;

extracting graphic images displayed in the objects included in the digital display page and identified from the DOM;

creating a repository of the extracted graphic images at the memory, wherein the extracted graphic images are arranged in the repository by at least one characteristic, and wherein the repository is configured to store extracted graphic images that are visible or invisible on the digital display page;

rendering the rendered code for the digital display page at the system using the DOM to cause display of a rendered version of digital display page and saving the rendered version of the digital display page as a rendered page image file at the memory;

performing a geometric analysis of the digital display page by:

calculating, for each extracted graphic image, a location of the extracted graphic image on the digital display page based on inspecting the DOM;

identifying a viewable portion of the digital display page; and

determining, for each extracted graphic image, whether the extracted graphic image is outside the viewable portion of the digital display page, wherein all extracted graphic images outside the viewable portion of the digital display page are marked as invisible images in the repository;

determining, for each extracted graphic image, whether a threshold amount of CPU resources are used for optimizing each extracted graphic image, wherein each extracted graphic image that has the threshold amount of CPU resources used to optimize the extracted graphic image is marked as visible images in the repository;

comparing the extracted graphic images stored in the repository to the rendered page image file to identify all extracted graphic images from the repository that are included within the rendered page image file by using an image analysis routine to determine whether a threshold number of pixels of each extracted graphic image are represented in the rendered page image file, wherein extracted graphic images that are identified in the rendered page image are marked as visible images in the repository, and extracted graphic images that are not identified in the rendered page image are marked as invisible images in the repository; and

when one or more extracted graphical images are marked as invisible images, classifying the digital display page as containing hidden content.

5. The system of claim 4 , wherein the system is configured to compare the extracted graphic images in the repository to the rendered page image using image analysis to compare each extracted graphic image in the repository with graphic images contained within the rendered page image, and to identify as visible those extracted graphic images in the repository that match with a graphic image within the rendered page image.

6. The system of claim 4 , wherein the system is configured to classify a digital display page as containing an invisible object if any extracted graphic image of the digital display page is marked as an invisible image.

7. The system of claim 4 , wherein the system is configured such that, when there is identified an animated object in a digital display page, each frame of the animation is extracted as a separate graphic image of the animated object, and the animated object is marked as a visible object if any of the separate graphic images of the animated object is identified in the rendered page image as a visible image.

8. A computer program product for identifying hidden content comprising a non-transitory computer readable medium storing program instructions that when executed by a processor perform a method comprising:

obtaining rendered code for a digital display page using any of a web crawler implemented on the computer program product to download page content and cause a browser to render the digital display page, a piece of hypertext markup language (html) code obtained via the web crawler to access a document object module (DOM) for the digital display page, or a plugin implemented on any of the computer program product or an end user device to access the digital display page rendered on the browser;

analyzing the rendered code for the digital display page to identify one or more objects embedded in the digital display page by inspecting data types in the DOM for known types that correspond with objects and inspecting signatures and/or content sizes in the DOM that correspond with objects;

extracting graphic images displayed in the objects included in the digital display page and identified from the DOM;

creating a repository of the extracted graphic images at the memory, wherein the extracted graphic images are arranged in the repository by at least one characteristic, and wherein the repository is configured to store extracted graphic images that are visible or invisible on the digital display page;

rendering the rendered code for the digital display page at the computer program product using the DOM to cause display of a rendered version of digital display page and saving the rendered version of the digital display page as a rendered page image file at the memory;

performing a geometric analysis of the digital display page by:

calculating, for each extracted graphic image, a location of the extracted graphic image on the digital display page based on inspecting the DOM;

identifying a viewable portion of the digital display page; and

determining, for each extracted graphic image, whether the extracted graphic image is outside the viewable portion of the digital display page, wherein all extracted graphic images outside the viewable portion of the digital display page are marked as invisible images in the repository;

determining, for each extracted graphic image, whether a threshold amount of CPU resources are used for optimizing each extracted graphic image, wherein each extracted graphic image that has the threshold amount of CPU resources used to optimize the extracted graphic image is marked as visible images in the repository;

comparing the extracted graphic images stored in the repository to the rendered page image file to identify all extracted graphic images from the repository that are included within the rendered page image file by using an image analysis routine to determine whether a threshold number of pixels of each extracted graphic image are represented in the rendered page image file, where in extracted graphic images that are identified in the rendered page image are marked as visible images in the repository, and extracted graphic images that are not identified in the rendered page image are marked as invisible images in the repository; and

when one or more extracted graphical images are marked as invisible images, classifying the digital display page as containing hidden content.

9. The computer program product of claim 8 , wherein the program instructions are executable to capture the rendered code for the digital display page using a crawler or a client side browser plugin.

10. The computer program product of claim 8 , wherein the program instructions are executable to compare the extracted graphic images in the repository to the rendered page image using image analysis to compare each extracted graphic image in the repository with graphic images contained within the rendered page image, and to identify as visible those extracted graphic images in the repository that match with a graphic image within the rendered page image.

11. The computer program product of claim 8 , wherein the program instructions are executable to classify a digital display page as containing an invisible object if any extracted graphic image of the digital display page is marked as an invisible image.

12. The computer program product of claim 8 , wherein the program instructions are executable such that, when there is identified an animated object in a digital display page, each frame of the animation is extracted as a separate graphic image of the animated object, and the animated object is marked as a visible object if any of the separate graphic images of the animated object is identified in the rendered page image as a visible image.

Assignments (1)

Continuity (3)

References Cited (71)