Cost effective, mass producible system for rapid detection of fever conditions based on thermal imaging

Patent metadata

Field	Value
Publication number	US-11625828-B2
Application number	US-202117202976-A
Country	US
Kind code	B2
Filing date	Mar 16, 2021
Priority date	Mar 17, 2020
Publication date	Apr 11, 2023
Grant date	Apr 11, 2023

Abstract

Systems and methods based on thermal imaging for rapid detection of fever conditions in humans that provide for extremely inexpensive, mass producible, field deployable devices accurate in specific, relatively low temperature ranges, and in particular temperatures near nominal human body temperature. The system may include a thermal imager tailored for the application and a corresponding mass producible controlled temperature calibration source configured to provide real time calibration near the human body temperature of interest. The imager and source are deployed in a way such that target people and the calibration source will be within the imager FOV for fever detection. The combination of real time near measurement temperature calibration, with suitable thermography approaches, yield fast, accurate measurements in the fever range using low cost, easy-to-produce components. In combination with a visible imager and pattern/facial recognition techniques, detection of a human target's facial regions of interest suitable for fever detection can be accurately accomplished.

First claim

What is claimed is:

1. A system for measuring the temperature of a region of interest of a target, the system comprising:
at least one system controller;
at least one visible camera interfaced to the controller;
at least one thermal camera interfaced to the controller, wherein at least a portion of image pixel locations from the visible and thermal cameras are mapped to each other, and wherein the thermal camera has a thermography function based on a previous pixel-by-pixel calibration of at least some pixels of the thermal camera; and
at least one temperature controlled calibration source of a known shape;
wherein, when the target and the calibration source are within a field of view (FOV) of both cameras, the system controller is configured to:
obtain a visible image using the at least one visible camera and a thermal image using the at least one thermal camera;
locate the calibration source in the thermal image by performing pattern recognition on at least one of the visible image and the thermal image;
perform a temperature calibration of a subset of pixels in the thermal image, the subset of pixels corresponding to the calibration source;
update a thermography function of the thermal camera based on a known temperature of the calibration source, the known temperature of the calibration source being at least one of within 15 degrees, within 10 degrees, or within 5 degrees of a nominal human body temperature;
detect, using pattern recognition, visible region of interest pixels in the visible image and thermal region of interest pixels in the thermal image;
refine the thermal region of interest pixels based on the pattern recognition;
perform an updated thermography readout of the refined thermal region of interest pixels to produce a real-time calibrated temperature measurement based on the thermal region of interest pixels, the temperature measurement being within a range corresponding to within at least one of 15 degrees, 10 degrees, or 5 degrees above the nominal human body temperature; and
periodically update detection and mapping of the visible region of interest pixels, update location of the calibration source, and update the temperature calibration after a number of frames, wherein the number of frames is one and the periodically updating is performed on a frame-by-frame basis.

2. The system of claim 1 , wherein the target comprises an upper body region of a human, and wherein the region of interest comprises a portion of a facial region of the target.

3. The system of claim 1 , wherein the temperature measurement is a threshold measurement indicative of the target having a fever.

4. The system of claim 1 , wherein the threshold temperature measurement is based on a highest temperature pixel of the thermal region of interest pixels, statistically derived over the number of frames.

5. The system of claim 4 , wherein the number of frames is 8 and the statistical derivation is a boxcar average of the highest temperature pixel of the thermal region of interest pixels in each of the 8 frames.

6. The system of claim 5 , wherein the threshold temperature measurement is based on a statistical derivation of a number of the highest temperature pixels in the region and is derived over one or more frames.

7. The system of claim 1 , further comprising a user interface interfaced to the system controller, the user interface comprising at least one of audio or visible indicators indicating at least one of measurement validity based on verified detection of the target and the calibration source or a target temperature above a threshold.

8. The system of claim 1 , wherein locating the calibration source in the thermal image comprises at least one of identifying pixels of the known shape of the calibration source in the thermal image, or recognizing at least one of the source shape or color in the visible image and mapping to the corresponding thermal pixels.

9. The system of claim 1 , wherein pixels in a mapped facial region within a predetermined temperature range of the calibration source temperature are identified, and facial region coordinates in the thermal image are updated based on the identified pixels.

10. The system of claim 1 , wherein an actual size of at least one of the calibration source and a facial region are compared to the known size at a nominal distance from the cameras, and wherein the thermography function is adjusted for distance based on an estimated distance between the at least one thermal camera and the at least one of the calibration source and the facial region, the estimated distance being determined based on the comparison.

11. The system of claim 1 , wherein the system controller is further configured to:
use a nominal target temperature for a first target;
acquire and store actual measured skin temperatures for a number of successive targets;
keep a running statistical value, including at least one of an average, median, or other statistical value, of actual skin temperatures of the successive targets;
obtain a number of data points including at least one of a predetermined threshold number of data points or a number of data points such that deviations from the statistical value fall within a predetermined range; and
substitute the running statistical value for the nominal target temperature for at least some subsequent temperature measurements.

12. The system of claim 11 , wherein outliers are eliminated, and a statistical value of remaining center temperatures are used as the nominal target temperature.

13. The system of claim 11 , wherein the nominal target temperature is periodically updated after at least one of every successive target after the initial number or after a predetermined number of targets.

14. A method for measuring the temperature of a region of interest of a target with a system comprising at least one visible camera and at least one thermal camera interfaced to a system controller, wherein at least a portion of image pixel locations from the visible and thermal cameras are mapped to each other, and wherein the thermal camera has a thermography function based on a previous pixel-by-pixel calibration of at least some pixels of the thermal camera, the system further comprising at least one temperature controlled calibration source of a known shape, the method comprising:
causing the target and the calibration source to be within a field of view (FOV) of both cameras;
obtaining a visible image using the at least one visible camera and a thermal image using the at least one thermal camera;
locating the calibration source in the thermal image by performing pattern recognition on at least one of the visible image and the thermal image;
performing a temperature calibration of a subset of pixels in the thermal image, the subset of pixels corresponding to the calibration source;
updating a thermography function of the thermal camera based on a known temperature of the calibration source, the known temperature of the calibration source being at least one of within 15 degrees, within 10 degrees, or within 5 degrees of a nominal human body temperature;
detecting, using pattern recognition, visible region of interest pixels in the visible image and thermal region of interest pixels in the thermal image;
refining the thermal region of interest pixels based on the pattern recognition;
performing an updated thermography readout of the refined thermal region of interest pixels to produce a real-time calibrated temperature measurement based on the thermal region of interest pixels, the temperature measurement being within a range corresponding to within at least one of 15 degrees, 10

Assignees

• Seek Thermal Inc

Inventors

• Parrish William J
• Collins Jacob
• Williams Ross E

Classifications

• G06T2207/30088

  Skin; Dermal · CPC title

• G06T7/0014 Primary

  using an image reference approach · CPC title

• G06T7/80

  Analysis of captured images to determine intrinsic or extrinsic camera parameters, i.e. camera calibration · CPC title

• A61B5/0077 Primary

  Devices for viewing the surface of the body, e.g. camera, magnifying lens · CPC title

• A61B5/7405

  using sound · CPC title