US 8,510,814 B2Granted

Method and apparatus for network authentication of human interaction and user identity

AssigneeBinary Monkeys Inc.

InventorsWarren Pratte, David Stephenson

Life of this patent

Estimating the term…

Loading patent…

Abstract

A method and apparatus are provided to enable a server to determine if a client connecting to the server is doing so by means of human user interaction, as opposed to an automated process. In order to authorize access to services, the option of determining user identity, such as by means of a graphical shared secret, is also provided. Three aspects are described: (i) image formation from an object model; (ii) presentation of image choices to a user, and (iii) user action. Image formation includes the creation of one or more categorized 2-dimensional images with object regions for each image automatically qualified. These one or more categorized images can be created by means of a Randomizable Image Generation Object for Human Recognition, comprised of (i) a 3-dimensional object model, (ii) a plurality of rendering threshold and constraint parameters, and (iii) categorization and qualitative metadata. The one or more 2-dimensional images are preferably transmitted to the user for authentication without the image metadata, which may be retained on the server. Related inquiry text can be sent when human user interaction is being determined and not user identity authentication.

Claims22

Dependent claims

A computer-implemented method for authenticating a user of a computer comprising:

selecting plurality of 3-dimensional models from a library of 3-dimensional models, each of the 3-dimensional models in the library:
containing geometric shape data for a respective object in three spatial dimensions;
and
having an associated parameterization comprising a set of one or more rendering parameters specific to said model, each parameter having pre-defined thresholds and/or constraints such that said model remains recognizable by a human,
wherein the parameterization of one or more of the 3-dimensional models in the library comprises a parameter determining a deformation of the respective geometric shape data for the respective object;
for each respective selected model, randomly varying the one or more rendering parameters, within the respective pre-defined thresholds and/or constraints for the one or more rendering parameters specific to said model;
generating a 2-dimensional image based on the result of the varying step, such that each respective selected model remains recognizable by a human in said 2-dimensional image;
generating metadata based on each respective selected model, and the one or more rendering parameters, wherein the metadata comprises:
visibility regions identifying where each respective selected model is not occluded in the 2-dimensional image;
one or more descriptors identifying categorical properties of one of the respective selected models;
and
one or more descriptors identifying visual properties of one of the respective selected models,
wherein the categorical and visual properties are recognizable by a human in said 2-dimensional image;
generating a prompt based on at least one of the categorical descriptors and at least one of the visual descriptors;
communicating the 2-dimensional image for presentation at the computer;
communicating the prompt for presentation at the computer;
receiving user input data concerning the 2-dimensional image;
comparing the metadata and the user input data to determine if the user interacting with the computer is a human user, the comparing comprising a comparison of the user input data with the visibility regions;
and
authenticating or not authenticating based on the result of the comparing step.

17 more elements

A system for authenticating a user comprising:

a processor for:
selecting plurality of 3-dimensional models from a library of 3-dimensional models, each of the 3-dimensional models in the library:
containing geometric shape data for a respective object in three spatial dimensions;
and
having an associated parameterization comprising a set of one or more rendering parameters specific to said model, each parameter having pre-defined thresholds and/or constraints such that said model remains recognizable by a human,
wherein the parameterization of one or more of the 3-dimensional models in the library comprises a parameter determining a deformation of the respective geometric shape data for the respective object;
for each respective selected model, randomly varying the one or more rendering parameters, within the respective pre-defined thresholds and/or constraints for the one or more rendering parameters specific to said model;
generating a 2-dimensional image based on the result of the varying step, such that each respective selected model remains recognizable by a human in said 2-dimensional image;
generating metadata based on each respective selected model, and the one or more rendering parameters, wherein the metadata comprises:
visibility regions identifying where each respective selected model is not occluded in the 2-dimensional image;
one or more descriptors identifying categorical properties of one of the respective selected models;
and
one or more descriptors identifying visual properties of one of the respective selected models,
wherein the categorical and visual properties are recognizable by a human in said 2-dimensional image;
generating a prompt based on at least one of the categorical descriptors and at least one of the visual descriptors;
communicating the 2-dimensional image for presentation at a computer;
communicating the prompt for presentation at the computer;
receiving user input data concerning the 2-dimensional image;
comparing the metadata and the user input data to determine if the user interacting with the computer is a human user, the comparing comprising a comparison of the user input data with the visibility regions;
and
authenticating or not authenticating based on the result of the comparing step.

18 more elements

Dependencies are inferred from the claim text.