US 12,436,253 B1
Coarse to fine movement detection in vehicles
Tarek El Dokor, Phoenix, AZ (US); and James Holmes, Mesa, AZ (US)
Assigned to Edge 3 Technologies, Phoenix, AZ (US)
Filed by Edge 3 Technologies, Inc., Phoenix, AZ (US)
Filed on Jul. 3, 2022, as Appl. No. 17/857,025.
Application 17/857,025 is a continuation of application No. 16/541,096, filed on Aug. 14, 2019, granted, now 11,378,671.
Int. Cl. G01S 13/04 (2006.01); B60N 2/00 (2006.01); G01S 7/35 (2006.01); G01S 13/50 (2006.01); G01S 13/88 (2006.01)
CPC G01S 13/04 (2013.01) [B60N 2/0022 (2023.08); B60N 2/0029 (2023.08); B60N 2/0033 (2023.08); G01S 7/354 (2013.01); G01S 13/50 (2013.01); G01S 13/88 (2013.01)] 5 Claims
OG exemplary drawing
 
1. A non-transitory recording medium having a computer program stored thereon which when run causes a computer to determine the presence of an occupant and to train an artificial intelligence neural network, by performing the steps comprising:
receiving a first signal from a first sensor positioned within a seating area of a seat and having a corresponding sensitive zone passing through the seating area vertically, the first sensor being adapted to sense coarse movements adjacent to and farther from the sensor;
determining whether the first signal includes a value greater than a coarse motion threshold;
receiving a second signal from a second sensor positioned within a back rest area of the seat and having a corresponding sensitive zone passing through a chest area of an occupant horizontally, the second sensor sensing at least one or more fine movements of an occupant, the one or more fine movements being selected from the group of heartbeat and breathing;
determining whether the second signal includes a value greater than a coarse motion threshold;
if it is determined that either the first signal or the second signal is greater than the coarse motion threshold, indicating that an occupant is present;
if it is determined that neither the first signal nor the second signal is greater than the coarse motion threshold, then determining whether the second signal includes a value that is greater than the fine motion threshold, further comprising the steps of:
determining whether the second signal includes a value that is greater than the fine motion threshold further comprises:
performing a spectral analysis on the second signal;
determining whether there is a dominant response between 0.8 and 3 Hz;
if it is determined that there is a dominant response between 0.8 and 3 Hz, thus indicating the presence of a heartbeat, indicate that an occupant is present;
if it is determined that there is no dominant response between 0.8 and 3 Hz, determining whether there is a dominant response between 0.2 to 0.5 Hz;
if it is determined that there is a dominant response between 0.2 and 0.5 Hz, thus indicating the presence of breathing, indicate that an occupant is present; and
if there is no dominant response between 0.2 and 0.5 Hz or between 0.8 and 3 Hz, indicate that no occupant is present;
processing a plurality of sensor samples in accordance with a convolutional neural network; and
outputting one or more class probabilities from a final neural network layer, the classes being used to encapsulate a plurality of different states of the system to be employed by the artificial intelligence neural network.