# US 12,223,818 B2 · Image-based evaluation of baby bodily position with alert > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,223,818 B2 - Document ID: US12223818B2 - Status: Granted - Kind: B2 - Application number: 18/346,750 - Filed: 2023-07-03 - Published: 2024-01-11 - Granted: 2025-02-11 - Earliest claimed: 2022-07-06 - Applicants: Gabriel Johnson - Inventors: Gabriel Johnson - Web page: https://patentgrep.com/patent/US12223818B2 ## CPC - G08B 21/0208: {Combination with audio or video communication, e.g. combination with "baby phone" function} - G06T 7/70: Determining position or orientation of objects or cameras (camera calibration G06T7/80) - G06V 10/454: {Integrating the filters into a hierarchical structure, e.g. convolutional neural networks [CNN]} - G06V 10/82: using neural networks - G06V 20/52: Surveillance or monitoring of activities, e.g. for recognising suspicious objects (recognising microscopic objects G06V20/69) - G06V 40/10: Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands - G06V 40/23: {Recognition of whole body movements, e.g. for sport training} - G08B 7/06: using electric transmission {, e.g. involving audible and visible signalling through the use of sound and light sources} - H04N 7/183: {for receiving images from a single remote source} - G06T 2207/20081: Training; Learning - G06T 2207/20084: Artificial neural networks [ANN] - G06T 2207/30196: Human being; Person - G06T 2207/30232: Surveillance - G08B: SIGNALLING SYSTEMS, e.g. PERSONAL CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS - G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL - G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING - H04N: PICTORIAL COMMUNICATION, e.g. TELEVISION ## Abstract A method to monitor a baby by evaluating live images through use of machine learning techniques. The bodily position of the baby is determined and a decision is made whether the position constitutes a potentially unsafe situation, such as a prone position correlated with Sudden Infant Death Syndrome events. In the event of an unsafe situation, an alarm is generated to alert the caretaker and allow a chance to intervene. This system is highly adaptable to user preferences by nature of allowing the user to select from a variety of image evaluation schemes. ## Claims ### Claim 1 A monitoring system for a sleeping area of a human baby, comprising: at least one processor; at least one communication interface coupled to said at least one processor which enables said at least one processor to communicate with other elements selected from the group consisting of at least one external device and at least one component of said monitoring system; at least one image sensor coupled to said at least one communication interface, wherein said at least one image sensor is controllable by said at least one processor, with said at least one image sensor being positioned such that said sleeping area is within said at least one image sensor's sensing range; at least one sound transducer coupled to said at least one communication interface, wherein output of said at least one sound transducer is controllable by said at least one processor, whereby an audible sound may alert a user; at least one memory coupled to said at least one processor, wherein said at least one memory has in its storage: at least one evaluation scheme, which has been preconfigured by at least one separate system and transferred to said at least one memory of said monitoring system; computer-executable instructions that when executed, cause said at least one processor to: communicate with at least one user interface by means of said at least one communication interface; obtain a plurality of images of said sleeping area from said at least one image sensor; store said plurality of images; identify at least one selected evaluation scheme, wherein said at least one selected evaluation scheme is selected from the group consisting of default selections and selections made by said user through said at least one user interface; process said plurality of images utilizing said at least one selected evaluation scheme, thereby providing an analysis of said plurality of images; determine a conclusion about said human baby's body position, based on said analysis; enable an alarm when said conclusion does not indicate a safe body position of said human baby. ### Claim 2 (depends on 1) The system as claimed in claim 1, wherein said at least one separate system is a remotely located computerized system which is not physically incorporated as part of said monitoring system itself, yet generates said at least one evaluation scheme which is transferable by way of said at least one communication interface and stored within said at least one memory. ### Claim 3 (depends on 1) The system as claimed in claim 1, further comprising at least one auxiliary light emitter coupled to said at least one communication interface, wherein output of said at least one auxiliary light emitter is controllable by said at least one processor. ### Claim 4 (depends on 1) The system as claimed in claim 1, wherein a software application is executed on said at least one external device which communicates with said monitoring system by way of said at least one communication interface to provide a graphical user interface, whereby said user may designate said at least one selected evaluation scheme which is used to process said plurality of images. ### Claim 5 (depends on 1) The system as claimed in claim 1, wherein said at least one evaluation scheme includes at least one deep learning model which has been trained on a dataset of images. ### Claim 6 (depends on 5) The system as claimed in claim 5, wherein said at least one deep learning model is employed to identify at least one general orientation of said human baby's head, which yields at least one confidence level of said at least one general orientation for use to infer a generalized body position of said human baby. ### Claim 7 (depends on 5) The system as claimed in claim 5, wherein a plurality of clothing articles worn by said human baby incorporate at least one distinct visible design, wherein each of the aid at least one distinct visible designs is uniquely positioned on each of the said plurality of clothing articles in a region selected from the group consisting of front side and back side and left side and right side. ### Claim 8 (depends on 7) The system as claimed in claim 7, wherein said at least one deep learning model is employed to identify at least one general orientation of said plurality of clothing articles, which yields at least one confidence level of said at least one general orientation of said plurality of clothing articles for use to infer a generalized body position of said human baby. ### Claim 9 A computerized method for monitoring a sleeping area of a human baby, comprising: Communicating, by at least one communication interface between at least one processor and other elements selected from the group consisting of at least one external device and at least one internal component; Controlling, by said at least one processor through said at least one communication interface, at least one image sensor, which is positioned such that said sleeping area is within said at least one image sensor's sensing range; Controlling, by said at least one processor through said at least one communication interface, the output of at least one sound transducer, whereby an audible sound may alert a user; Storing, in at least one memory coupled to said at least one processor, at least one evaluation scheme, which has been preconfigured by at least one separate system and transferred to said at least one memory; Communicating, by said at least one processor through said at least one communication interface, with at least one user interface; Obtaining, by said at least one processor through said at least one communication interface, a plurality of images of said sleeping area from said at least one image sensor; Storing, in said at least one memory, said plurality of images; Identifying, by said at least one processor, at least one selected evaluation scheme, wherein said at least one selected evaluation scheme is selected from the group consisting of default selections and selections made by said user through said at least one user interface; Processing, by said at least one processor, said plurality of images utilizing said at least one selected evaluation scheme, thereby providing an analysis of said plurality of images; Determining, by said at least one processor, a conclusion about said human baby's body position based on said analysis; Enabling, by said at least one processor, an alarm when said conclusion does not indicate a safe body position of said human baby. ### Claim 10 (depends on 9) The method as claimed in claim 9, wherein storing, in at least one memory, said at least one evaluation schemes may be accomplished by transferring said at least one evaluation scheme, by way of said at least one communication interface, to said at least one memory, which evaluation scheme originates in said at least one separate system, which said at least one separate system is a remotely located computerized system which generates said at least one evaluation scheme. ### Claim 11 (depends on 9) The method as claimed in claim 9, wherein controlling at least one auxiliary light emitter, by said at least one processor through said at least one communication interface, enables variable amounts of illumination to be produced by said at least one auxiliary light emitter. ### Claim 12 (depends on 9) The method as claimed in claim 9, wherein communicating, by said at least one communication interface between said at least one processor and said at least one external device, provides a means for said user to access a graphical user interface, whereby said user may designate said at least one selected evaluation scheme which is used to process said plurality of images. ### Claim 13 (depends on 9) The method as claimed in claim 9, wherein processing, by said at least one processor, includes use of at least one deep learning model which has been trained on a dataset of images. ### Claim 14 (depends on 13) The method as claimed in claim 13, wherein processing, by said at least one processor, using said at least one deep learning model identifies at least one general orientation of said human baby's head, which yields at least one confidence level of said at least one general orientation for use to infer a generalized body position of said human baby. ### Claim 15 (depends on 13) The method as claimed in claim 13, wherein clothing said human baby makes use of a plurality of clothing articles worn by said human baby, wherein each said clothing article incorporates at least one distinct visible design, wherein each of the said at least one distinct visible designs is uniquely positioned on each of the said plurality of clothing articles in a region selected from the group consisting of front side and back side and left side and right side. ### Claim 16 (depends on 13) The method as claimed in claim 13, wherein processing, by said at least one processor, using said at least one deep learning model identifies at least one general orientation of said plurality of clothing articles, which yields at least one confidence level of said at least one general orientation of said plurality of clothing articles for use to infer a generalized body position of said human baby. ### Claim 17 A non-transitory computer-readable storage medium, having computer executable instructions stored thereon that, when executed by at least one processor, cause the at least one processor to: Communicate between said at least one processor and other elements selected from the group consisting of at least one external device and at least one internal component of the monitoring system; Control at least one image sensor, which is positioned such that said sleeping area is within said at least one image sensor's sensing range; Control the output of at least one sound transducer, whereby an audible sound may alert a user; Store at least one evaluation scheme, which has been preconfigured by at least one separate system and transferred to said at least one memory; Communicate with at least one user interface; Obtain a plurality of images of said sleeping area from said at least one image sensor; Store said plurality of images; Identify at least one selected evaluation scheme, wherein said at least one selected evaluation scheme is selected from the group consisting of default selections and selections made by said user through said at least one user interface; Process said plurality of images utilizing said at least one selected evaluation scheme, thereby providing an analysis of said plurality of images; Determine a conclusion about said human baby's body position based on said analysis; Enable an alarm when said conclusion does not indicate a safe body position of said human baby. ### Claim 18 (depends on 17) The non-transitory computer-readable storage medium of claim 17 wherein storing said at least one evaluation schemes may be accomplished by transferring said at least one evaluation scheme, by way of said at least one communication interface, to said at least one memory, which evaluation scheme originates in said at least one separate system, which said at least one separate system is a remotely located computerized system which generates said at least one evaluation scheme. ### Claim 19 (depends on 17) The non-transitory computer-readable storage medium of claim 17, wherein communicating by said at least one communication interface between said at least one processor and said at least one external device, provides a means for said user to access a graphical user interface, whereby said user may designate said at least one selected evaluation scheme which is used to process said plurality of images. ### Claim 20 (depends on 17) The non-transitory computer-readable storage medium of claim 17, wherein processing includes use of at least one deep learning model which has been trained on a dataset of images.