Occupant position detection system
AssigneesToyota Jidosha Kabushiki Kaisha, Aisin Seiki Kabushiki Kaisha
InventorsUdara Manawadu, Yoshihiro Saeki
Life of this patent
Estimating the term…
Abstract
An occupant position detection system includes an imaging device configured to capture an image of an occupant seated on a seat provided in an interior of a vehicle, a sensor provided on the seat, a first calculating unit configured to calculate a first position of the occupant from the image captured by the imaging device, a second calculating unit configured to calculate a second position of the occupant from a detection result of the sensor, an occupant position calculating unit configured to calculate an occupant position by fusing the first position and the second position, and an output unit configured to output information on the occupant position calculated by the occupant position calculating unit, to a safety device.
Claims5
An occupant position detection system, comprising:
- an imaging device configured to capture an image of an occupant seated on a seat provided in an interior of a vehicle;
- a sensor provided on the seat;
- and
- a central processing unit configured to:
- calculate a first position which includes a first head position of the occupant from the image captured by the imaging device;
- calculate a second position which includes a second head position of the occupant from a detection result of the sensor;
- calculate an occupant position by fusing the first position and the second position;
- and
- output information on the occupant position to a safety device,
- wherein the central processing unit includes:
- a neural network that receives the image captured by the imaging device, and outputs information about facial landmarks and a body skeleton of the occupant and a first seat position in two dimensions;
- a triangulation model that receives information about the facial landmarks and the body skeleton of the occupant and the first seat position from the neural network, and generates the first head position in three dimensions, the body position in three dimensions and the first seat position in three dimensions, and
- a fusion unit including a fusion algorithm that fuses the first head position, the body position, and the first seat position generated by the triangulation model, so as to output the first position including the body position of the occupant in three dimensions which includes the first head position in three dimensions.
Dependencies are inferred from the claim text.