Hyperspectral scanning to determine skin health

A non-transitory computer readable storage medium storing instructions that, when executed by at least one processor, cause the at least one processor to perform operations comprising:

• obtaining, from an imaging system, first skin data associated with a user, wherein the first skin data comprises an ultraviolet (UV) image of the user's skin;
• performing a correction on the first skin data, based on second skin data associated with the user, the second skin data having been collected at a time prior to obtaining the first skin data, wherein the second skin data is stored in a repository;
• quantifying a plurality of skin parameters of the user's skin, wherein the at least one of the plurality of skin parameters is a bacterial load, the quantifying comprising:
• applying a brightness filter that isolates portions of at least one UV image that contain fluorescence,
• applying a dust filter that removes portions of the at least one UV image that contain fluorescence caused by dust particles, wherein the dust filter comprises an expectation maximization algorithm and a two-component Gaussian mixture model,
• applying a loopy belief propagation algorithm that isolates portions of the at least one UV image that contain fluorescence due to bacteria, and
• determining a bacterial load quantity associated with the user's skin based on an output of the loopy belief propagation algorithm;
• determining, based on providing the plurality of quantified skin parameters as input features to a machine learning model, an output associated with a normal skin state of the user and a current skin state of the user;
• and
• providing, for display on a user computing device, data indicating the output.

A computer-implemented skin health assessment method executed by one or more processors and comprising:

• obtaining, from an imaging system, first skin data associated with a user, wherein the first skin data comprises an ultraviolet (UV) image of the user's skin;
• performing a correction on the first skin data, based on second skin data associated with the user, the second skin data having been collected at a time prior to obtaining the first skin data, wherein the second skin data is stored in a repository;
• quantifying a plurality of skin parameters of the user's skin, wherein the at least one of the plurality of skin parameters is a bacterial load, the quantifying comprising:
• applying a brightness filter that isolates portions of at least one UV image that contain fluorescence,
• applying a dust filter that removes portions of the at least one UV image that contain fluorescence caused by dust particles, wherein the dust filter comprises an expectation maximization algorithm and a two-component Gaussian mixture model,
• applying a loopy belief propagation algorithm that isolates portions of the at least one UV image that contain fluorescence due to bacteria, and
• determining a bacterial load quantity associated with the user's skin based on an output of the loopy belief propagation algorithm;
• determining, based on providing the plurality of quantified skin parameters as input features to a machine learning model, an output associated with a normal skin state of the user and a current skin state of the user;
• and
• providing, for display on a user computing device, data indicating the output.