System and Method for Human Motion Detection and Tracking

A system for human motion detection and tracking, the system comprising:

• a housing securing an optical sensing instrument, a processor, and memory therewith;
• a busing architecture communicatively interconnecting the optical sensing instrument, the processor, and the memory;
• and
• the memory accessible to the processor, the memory including processor-executable instructions that, when executed, cause the processor to:
• convert a depth frame into a designated depth frame format, the designated depth frame format including at each image element second coordinate values relative to the depth frame,
• acquire a plurality of probability distribution models for the designated depth frame format to linearly identify a respective plurality of body parts, the plurality of probability distribution models being single source generated by a neural network trained with markerless, colorless synthetic depth frames,
• calculate a position of each of the respective plurality of body parts in the designated depth frame format, and
• calculate the position of each of the respective plurality of body parts in the depth frame.

A system for human motion detection and tracking, the system comprising:

• an integrated goniometer comprising:
• a housing securing an optical sensing instrument, a processor, and memory therewith;
• a busing architecture communicatively interconnecting the optical sensing instrument, the processor, and the memory;
• the optical sensing instrument monitoring a stage, the stage being a virtual volumetric area that is compatible with human exercise positions and movement, the stage including a volume in space at a monitoring distance from the optical sensing instrument;
• a server comprising:
• a server housing securing a server processor, and a server memory therewith, and
• a server busing architecture communicatively interconnecting the server processor and the server memory;
• a data link providing communication between the integrated goniometer and the server;
• and
• the memory accessible to the processor and the server memory accessible to the server processor, processor-executable instructions distributed between the memory and the server memory that, when executed, cause the processor and the server processor, in a distributed manner via the data link, to:
• capture a depth frame from the optical sensing instrument, the depth frame including at each image element first coordinate values including a point related to a distance from the optical sensing instrument,
• convert the depth frame into a designated depth frame format, the designated depth frame format including at each image element second coordinate values relative to the depth frame,
• acquire a plurality of probability distribution models for the designated depth frame format to linearly identify a respective plurality of body parts, the plurality of probability distribution models being single source generated by a neural network trained with markerless, colorless synthetic depth frames,
• calculate a position of each of the respective plurality of body parts in the designated depth frame, and
• calculate the position of each of the respective plurality of body parts in the depth frame.