Abstract: A face detection device includes processing circuitry configured to acquire, from an imaging device mounted on a vehicle and imaging an at least one occupant inside the vehicle, a captured image in which the at least one occupant is imaged; set a search area in which a face of the at least one occupant is searched in the acquired captured image; search the search area and detect the face of the at least one occupant; determine whether or not the detection of the face of the at least one occupant has succeeded; and expand the search area when determining that the detection of the face of the at least one occupant has failed.

Abstract: An area extraction unit (2) extracts an eye area in an image of a driver. An eyelid detection unit (3) detects an eyelid in the eye area. A reliability calculation unit (4) calculates an eyelid reliability by using luminance information of the eye area and positional information of the eyelid detected by the eyelid detection unit (3). A determination unit (5) determines that the eyelid has not been properly detected, when the eyelid reliability is less than a first threshold value.

Abstract: An area extraction unit (2) extracts an eye area in an image of a driver. An eyelid detection unit (3) detects an eyelid in the eye area. A reliability calculation unit (4) calculates an eyelid reliability by using luminance information of the eye area and positional information of the eyelid detected by the eyelid detection unit (3). A determination unit (5) determines that the eyelid has not been properly detected, when the eyelid reliability is less than a first threshold value.

Abstract: A mode selecting unit (6) selects a processing mode of occupant monitoring by using vehicle information, and an area selecting unit (7) selects an area corresponding to the selected processing mode. An exposure control unit (8) then determines an exposure setting that has to be set on an image shooting unit (2) by using luminance information about the selected area.

Abstract: A mode selecting unit (6) selects a processing mode of occupant monitoring by using vehicle information, and an area selecting unit (7) selects an area corresponding to the selected processing mode. An exposure control unit (8) then determines an exposure setting that has to be set on an image shooting unit (2) by using luminance information about the selected area.