-
公开(公告)号：US11282367B1
公开(公告)日：2022-03-22
申请号：US17402583
申请日：2021-08-15
摘要： A system includes video cameras arranged to monitor a vulnerable person, and a processor system that receives video frames from the video cameras, the processor system comprising a processor and a non-transitory, computer-readable storage medium having machine instructions executed by the processor. The processor detects and identifies objects in a current received video frame, classifies an identified object as the person by applying a facial recognition algorithm that identifies the person, determines a posture of the person by identifying joints, limbs, and body parts, and their respective orientations to each other and to a plane, and immediately discards the current video frame. The processor then determines a change in motion, of the person, between the current received video frame and one or more prior received video frames, and, based on the determined posture and the change in motion, determines that the person has experienced a defined event.
-
公开(公告)号：US11688265B1
公开(公告)日：2023-06-27
申请号：US17692396
申请日：2022-03-11
CPC分类号： G08B21/0476 , G06V20/41 , G06V40/166 , G06V40/23 , G08B21/043 , G06V20/44
摘要： A system includes video cameras arranged to monitor a vulnerable person, and a processor system that receives video frames from the video cameras, the processor system comprising a processor and a non-transitory, computer-readable storage medium having machine instructions executed by the processor. The processor detects and identifies objects in a current received video frame, classifies an identified object as the person by applying a facial recognition algorithm that identifies the person, determines a posture of the person by identifying joints, limbs, and body parts, and their respective orientations to each other and to a plane, and immediately discards the current video frame. The processor then determines a change in motion, of the person, between the current received video frame and one or more prior received video frames, and, based on the determined posture and the change in motion, determines that the person has experienced a defined event.
An example method for monitoring health and well-being of persons includes a processor integral to a video camera capturing video frames containing raw image data of a scene containing person objects and non-person objects. The processor, for a first video frame, executes video frame operations, including detecting the person objects in the first video frame, detecting and identifying the non-person objects in the first video frame, identifying and classifying a person object as a person for metric analysis, and determining a posture of the person. The method further includes the processor repeating the video frame operations for second and subsequent video frames; identifying changes in posture of the person from the second and subsequent video frames; using the changes in posture, determining motion of the person; providing person privacy, comprising automatically and immediately following each video frame operation on a video frame, deleting all raw image data of the video frame; and, based on the posture changes and the motion, generating a private, personalized metric analysis for the person. In an aspect, the private, personalized metric analysis includes outcome measures of interest selected from a group consisting of steps taken/distance walked/changes in gait by elderly persons, positional changes for bedridden persons, and developmental milestones in child persons; and the processor conducts the metric analysis over time to identify trends and corresponding changes in a health condition of the person, comprising progress in rehabilitation, deterioration from a degenerative disease, and physical and mental development.
-
搜索结果
2 条记录 (耗时 0.011 秒)
