Device and method for controlling a passenger airbag of a vehicle

Abstract

Device for controlling a passenger airbag in a vehicle, with a sensor unit integrated into a passenger seat (202) which is configured to determine at least one measured value of the weight of an object located on the passenger seat (202), with at least one processing unit (136) which is at least configured to process the measured value transmitted from the sensor unit to the processing unit (136), to perform a classification of the object based on the measured value and at least one threshold value stored in the processing unit (136) and to determine a first classification result, with at least one camera (134) of the vehicle (102) which is designed to capture at least one image (200, 300) with an image of an area of ​​the vehicle interior encompassing at least part of the passenger seat (202) and to generate image data corresponding to the image (200, 300), wherein the processing unit (136) is further configured to process the image data, wherein the processing unit (136) is further configured to perform a classification of an object located on the passenger seat (202) based on the image data and to determine a second classification result, wherein the processing unit (136) is further configured to adjust the threshold in the event of a deviation between the first classification result and the second classification result, such that the first classification result matches the second classification result, and wherein the processing unit (136) is further configured to determine, starting from an adapted first classification result determined after adaptation during classification, a control parameter of an airbag control unit (142) for triggering the passenger airbag (146).

Description

[0001] The invention relates to devices and methods for controlling a passenger airbag of a vehicle.

[0002] Document DE 10 2017 000 734 A1 discloses a system for detecting objects on the passenger seat of a vehicle and controlling the passenger airbag depending on the detected objects. The airbag is activated when a forward-facing child seat or a regular occupant is detected. The airbag is deactivated when a rear-facing child seat is detected.

[0003] Document DE 694 31 667 T2 discloses a method for deactivating the passenger airbag upon detection of a child seat without a child and in a rear-facing position. Detection is achieved using distance sensors.

[0004] Document US 6,608,910 B1 discloses the classification of objects on the front passenger seat as "adult person," "forward-facing child seat," and "rear-facing child seat." Document US 2006 / 0030988 A1 discloses a vehicle camera for detecting children inside the vehicle.

[0005] There is a general requirement to activate and deactivate passenger airbags for children in order to minimize the risks associated with deployment in so-called out-of-position scenarios. For example, the US Federal Automotive Standard FMVSS 208 provides the option to deactivate the passenger airbag upon detection of a child in the front passenger seat.

[0006] In known solutions, the classification of children in the front passenger seat is primarily achieved by determining their weight using capacitive or pressure-sensitive sensor mats in the passenger seat. Camera-based seat occupancy detection allows for differentiation between "empty seat," "child in child seat," and "adult" (greater than or equal to the 5th percentile for women).

[0007] A purely weight-based distinction between an empty seat, an empty child seat, a child in a child seat, a child in an infant carrier, and an adult is becoming increasingly inaccurate due to the ever-increasing heaviness and complexity of child seat systems. Increasingly flexible child seats (forward-facing / rear-facing, Isofix / vehicle seat belt) and approved for a very wide weight range further exacerbate the problem. The invention is based on the understanding that the risk of injury in severe crashes can be reduced for larger, forward-facing children by reactivating the airbags. Unfortunately, the orientation of the child seat and the precise size differentiation of children are very difficult to determine reliably using seat sensors alone. Due to this uncertainty, potential airbag reactivation in severe crashes is not possible.

[0008] Document DE 10 2019 220 119 A1 discloses a method for resuscitating a vehicle occupant. A vehicle occupant is detected by a sensor in the vehicle's interior. Sensor data generated by the sensor is evaluated to classify the occupant. Depending on the occupant's classification, one or more parameters of a resuscitation signal are selected. A vehicle component is then controlled according to the resuscitation signal to influence a vital function of the vehicle occupant.

[0009] Document DE 10 2020 215 653 A1 discloses a method for providing protection to a vehicle occupant, in which at least one real-time image of the vehicle interior is captured and an occupant is detected in this image. The detected occupant is classified based on the at least one real-time image. A vehicle operator is notified of the detected classification. At least one deployment characteristic of an airbag assigned to the detected occupant is determined based on the classification.

[0010] Document DE 10 2019 110 429 A1 discloses a control system for the airbag activation status of a motor vehicle. It detects when a passenger in a motor vehicle is not in a safe position, at which point an airbag can be deactivated.

[0011] Document DE 10 2021 102 814 A1 discloses a system for regulating the deployment of a restraint device. Based on the output of a seat occupancy sensor, it is determined whether a minor occupant is occupying a vehicle seat. Based on this, the deployment of an airbag module associated with the vehicle seat can be suppressed.

[0012] Devices and methods for controlling a passenger airbag in a vehicle are known from the prior art. These devices and methods include at least one camera of the vehicle configured to capture at least one image of an area of ​​the vehicle interior encompassing at least part of the passenger seat and to generate corresponding image data. A processing unit is configured to process the image data and, based on this data, to classify an object or person located in the passenger seat and determine a classification result. Based on the classification result obtained during the classification process, the processing unit determines a control parameter for an airbag control unit to trigger the passenger airbag.

[0013] Based on the known state of the art, the object of the invention is therefore to provide a device and a method for controlling a passenger airbag of a vehicle, which increases the accuracy of the detection of objects on the passenger seat.

[0014] This problem is solved by a device having the features of claim 1 and claim 2, and by a method having the features of the independent method claims. Advantageous embodiments are specified in the dependent claims.

[0015] The proposed device for controlling a passenger airbag of a vehicle according to claim 1 utilizes a two-stage and, according to claim 2, a dual detection principle with the aid of a sensor unit in the passenger seat and image processing of a camera image. This improves the precise and correct classification, particularly of children in child seats, so that, for example, in severe crashes and for children older than 6 or older than 10 years in forward-facing child seats, the passenger airbag can be reactivated even if the passenger airbag is deactivated when a child in a child seat is detected in the passenger seat. This reduces the risk of serious injury for children in the passenger seat in severe crashes. Based on the correct classification of an object in the passenger seat, a control parameter is generated, which is preferably transmitted to an airbag control unit.The airbag control unit can then reactivate the passenger airbag depending on the control parameter and preferably on other parameters, particularly based on measurements from collision sensors, even if a child is detected in the passenger seat. This allows for the reliable classification of children in the passenger seat even when a clear classification by the sensor unit in the passenger seat for determining weight is not possible under conventional use of this sensor unit.

[0016] The sensor unit integrated into the passenger seat is in particular a capacitive or pressure-sensitive sensor mat, which is preferably integrated into the seat cushion of the passenger seat.

[0017] In a beneficial further development, the processing unit is configured to transmit the control parameter to the airbag control unit. Furthermore, the airbag control unit is configured to activate, deactivate, or reactivate the passenger airbag depending on other parameters. These other parameters may include, in particular, an expected collision intensity, a collision intensity measured by sensors such as accelerometers, and / or a collision speed between the vehicle and a collision object.

[0018] Furthermore, it is advantageous if the following are classified as classification results, first classification results, second classification results, further classification results, and / or overall classification results during the classification, further classification, or overall classification: "empty front passenger seat," "empty child seat in the front passenger seat," "empty infant carrier in the front passenger seat," "child in a forward-facing child seat in the front passenger seat," "child in a rear-facing infant carrier in the front passenger seat," "child in a forward-facing infant carrier in the front passenger seat," and "adult in the front passenger seat" (at least 5th percentile female and taller). Preferably, the following are classified as subclassification results: "child up to 1 year old," "child 1 to 3 years old," "child 3 to 6 years old," and / or "child 6 to 10 years old."The processing unit determines the control parameter preferably depending on the classification result and the subclassification result.

[0019] It is particularly advantageous if the processing unit performs the classification and / or secondary classification and / or subclassification of the image data using a human body pose estimation method, a pattern recognition method, and / or a machine learning method. Using a human body pose estimation method, in particular, the dimensions of limbs can be determined from a two-dimensional image, which can then be used for object recognition, especially for determining the age of a child.

[0020] Furthermore, it is advantageous if the airbag control unit is configured to deactivate the airbag depending on the control parameter, which depends on the classification result and / or the subclassification result, and if the airbag control unit is configured to reactivate the airbag depending on the control parameter and the sensor signals from at least one airbag trigger sensor unit. This makes reactivation of the passenger airbag particularly easy in severe crashes.

[0021] It is particularly advantageous if the airbag control unit reactivates the airbag upon detecting a collision at a collision speed with a collision object greater than or equal to a preset stored speed threshold, or if the airbag control unit determines the medium or high severity of an expected or actual collision based on sensor signals and a control parameter such as the classification of a child in a forward-facing child seat and / or the subclassification of a child aged 6 to 10 years. A high severity level, for example, is present in a collision with an impact speed of 56 km / h.

[0022] Furthermore, it is advantageous if the camera is an interior camera of the vehicle, wherein the camera's field of view has in particular a detection angle in the range of 100° to 150°, wherein the camera is preferably a monocular camera and / or an RGB camera and / or an RGB-IR camera.

[0023] In general, the images captured by the camera, in combination with a sensor unit integrated into the passenger seat, particularly the sensor mat, can be used to classify an object on the passenger seat. The trigger logic of an airbag control unit for deactivating or reactivating the passenger airbag depends on the severity of the crash.

[0024] The methods with the features of the dependent method claims have the same advantages as the claimed devices. In particular, the methods can be further developed with the features of the dependent claims directed to the devices.

[0025] Exemplary embodiments of the invention are explained in more detail below with reference to the figures. These show: Fig. 1. A perspective schematic representation of a vehicle cockpit; Fig. 2. In schematic representation, a first image of the vehicle's interior taken by the camera; and Fig. Figure 3 shows a schematic representation of a second image of the vehicle interior taken by the camera.

[0026] Fig. Figure 1 shows a perspective schematic representation of the cockpit 100 of a vehicle 102. A driver 104 is seated in a driver's seat 106 of the vehicle 102. A display 108 for showing information when the passenger airbag 146 of the vehicle 102 is deactivated is arranged, by way of example, in the area of ​​a center console 110 of the cockpit 100. The cockpit 100 also includes a central information display (CID) 112, a head-up display 114, and a graphic instrument cluster 116, which are arranged in a dashboard 118 of the vehicle 102. The dashboard 118 also includes air outlet nozzles 120 of an air conditioning system of the vehicle 102. Fig. Figure 1 shows, by way of example, air outlet nozzles 120 in the area of ​​the center console 110 and on one passenger side. The cockpit 100 also includes a steering wheel 122 with controls 124, a gear selector 126, pedals 128, and an input unit 130 with a rotary dial and push-button function and / or a touch input panel. This input unit 130 is also referred to as the Ergo Commander.

[0027] A camera 134 is arranged in the upper area of ​​the windshield 132 of the vehicle 102. The camera 134 is arranged and oriented such that it captures images 200, 300 depicting at least part of the interior of the vehicle 102, including the passenger seat 202. Specifically, the field of view of the camera 134 is directed towards the driver's seat 106 and a passenger seat 202 of the vehicle 102. Exemplary images 200, 300 from the camera 134 are shown in the Fig. 2 and Fig. Figure 3 shows that the camera 134 is specifically designed to capture several successive images, generate image data corresponding to images 200 and 300, and transmit this data to a control unit 136. The camera 134 has a field of view of 120 degrees. In other embodiments, the camera 134 can also have a field of view of 150 degrees or more.

[0028] The control unit 136 has data outputs 138 and data inputs 140, which serve to connect to other units of the vehicle 102, for example, to other sensors, input and output units, and control units of assistance systems. In particular, the control unit 136 is connected to an airbag control unit 142, which is designed to trigger the airbags of the vehicle 102 in accident situations, i.e., in crashes, and in particular to control the deployment of the passenger airbag 146.

[0029] The control unit 136 also includes a communication module 144, which is designed to establish a connection with a telecommunications network, in particular a mobile communications network.

[0030] A sensor mat integrated into the front passenger seat 202 detects the weight of any object located on the front passenger seat 202. Such a sensor mat can be capacitive or pressure-sensitive. Depending on the detected weight, any object on the front passenger seat 202 is classified as "no object present or empty front passenger seat", "empty child seat on the front passenger seat", "empty infant carrier on the front passenger seat", "child in a forward-facing child seat on the front passenger seat", "child in a rear-facing infant carrier on the front passenger seat", "child in a forward-facing infant carrier on the front passenger seat", and "adult on the front passenger seat".Furthermore, based on the determined weight, in addition to the classifications "child in a forward-facing child seat on the front passenger seat," "child in a rear-facing infant carrier on the front passenger seat," and "child in a forward-facing infant carrier on the front passenger seat," a subclassification result of "up to 1 year old," "1 to 3 year old," "3 to 6 year old," and / or "6 to 10 year old" can preferably be generated. Depending on the classification result and the subclassification result, the control unit 136 determines a control parameter, which is transmitted to the airbag control unit 142. It is advantageous if the sensor mat integrated in the front passenger seat 202 determines the weight of the child 302 including the child seat 310. Previously, for example, with the vehicle 102 closed, the weight of the empty child seat 310 was determined using the sensor mat and stored in the control unit 136.Thus, the weight of child 302 can be determined relatively accurately and used to determine the age of child 302.

[0031] Depending on the control parameter, the airbag control unit 142 permanently deactivates the passenger airbag 146 if the passenger seat 202 is empty or if an empty child seat 310 or an empty infant carrier 204 is located on the passenger seat 202. For a small child 206 up to 6 years of age, the airbag control unit 142 can also permanently deactivate the passenger airbag 146. For children 302 from 6 to 14 years of age, the airbag control unit 142 can reactivate the passenger airbag 146 only in moderate and severe crashes and leave it deactivated in minor crashes. For children over 14 years of age and adults, the airbag control unit 142 permanently activates the passenger airbag 146.

[0032] However, a purely weight-based classification of child restraint systems and children 206, 302 as well as of adults is becoming increasingly inaccurate due to ever heavier and more diverse child seat systems 204, 310. Increasingly flexible (forward-facing / rear-facing, isofix / vehicle belt) child seats 204, 310, approved for a very wide weight range, further exacerbate the problem.

[0033] According to a first embodiment, the first classification of the object on the passenger seat 202, performed using the determined weight, is verified by a second classification based on at least one image captured by the camera 134. For this purpose, the camera 134 captures an image depicting at least part of the passenger seat 202 within the vehicle interior and generates corresponding image data, which the control unit 136 processes. Based on this image data, the control unit 136 performs a second classification of the object located on the passenger seat 202 and determines a second classification result.If the first weight-based classification result differs from the second image-based classification result, the control unit 136 adjusts at least one threshold used for the first classification so that the first classification result matches the second classification result after the adjustment. This permanently improves the first classification, so that the first classification is performed during or before every trip, and the second classification only needs to be performed at longer intervals. However, the second classification can also be performed during or before every trip to verify the first classification result.

[0034] The control unit 136 determines the control parameter for controlling the airbag control unit 142 based on an adapted first classification result determined after adjustment during classification and transmits this to the airbag control unit 142.

[0035] According to a second embodiment for controlling a passenger airbag 146 of a vehicle 102, at least one camera 134 of the vehicle 102 is provided, which captures at least one image depicting an area of ​​the vehicle interior encompassing at least part of the passenger seat 202 and generates corresponding image data, which the control unit 136 then processes. Based on the image data, the control unit 136 performs a classification of an object located on the passenger seat 202 and determines a classification result. Based on the classification result determined during the classification process, the control unit 136 determines a control parameter for an airbag control unit 142 to trigger the passenger airbag 146. This enables precise classification and also correct subclassification of the object simply and reliably. Misclassifications are avoided.

[0036] Additionally, a sensor mat integrated into a passenger seat 202 can determine at least one measurement of the weight of an object located on the passenger seat 202. The control unit 136 processes the measurement transmitted from the sensor mat to the control unit 136 and, based on the measurement and at least one threshold value stored in the control unit 136, performs a classification of the object to determine a further classification result. Based on the determined classification results of the first classification and the second classification, the control unit 136 calculates an overall classification result. Based on the overall classification result, the control unit 136 determines the control parameter for the airbag control unit 142. This enables a particularly reliable classification and preferably also a correct subclassification of an object on the passenger seat 202.

[0037] The following are specifically provided as classification results, first classification results, second classification results, further classification results and / or overall classification results: - “empty passenger seat”, - “empty child seat on the passenger seat”, - "empty baby car seat on the passenger seat", - “Child in a forward-facing child seat in the front passenger seat”, - “Child in a rear-facing infant car seat in the front passenger seat”, - “Child in a forward-facing infant car seat in the front passenger seat” and / or - “adult person in passenger seat”.

[0038] The following is specifically intended as a subclassification result: - “a person up to 1 year old”, - “a person aged 1 to 3 years”, - “a person aged 3 to 6” and / or - "a person aged 6 to 10".

[0039] To determine the subclassification result, the control unit 136 performs a human body pose estimation procedure to determine the measurement of at least one limb of a child and, based on this, the child's age. Alternatively or additionally, the control unit 136 can also use a pattern recognition procedure and / or a machine learning procedure to determine the child's age.

[0040] The control unit 136 determines a different control parameter for each classification result. Likewise, the control unit 136 determines further, different control parameters for each combination of a classification result and a subclassification result.

[0041] Depending on the control parameter, the airbag control unit 142 permanently deactivates the passenger airbag 146 if the passenger seat 202 is empty or if an empty child seat 310 or an empty infant carrier 204 is located on the passenger seat 202. For a small child 206 up to 6 years of age, the airbag control unit 142 can also permanently deactivate the passenger airbag 146. For children 302 from 6 to 14 years of age, the airbag control unit 142 can reactivate the passenger airbag 146 only in moderate and severe crashes and leave it deactivated in minor crashes. For children over 14 years of age and adults, the airbag control unit 142 permanently activates the passenger airbag 146.

[0042] It is particularly advantageous if camera 134 is an RGB-IR interior camera. The recorded images are used in combination with the seat sensors, especially the sensor mat, to classify an object on the passenger seat 202. The trigger logic of an airbag control unit 142 for deactivating or reactivating the passenger airbag 146 depends on the severity of a crash.

[0043] Fig. Figure 2 shows a schematic representation of an image 200 of the interior of vehicle 102, taken by camera 134. A baby carrier 204 is arranged on the front passenger seat 202 of vehicle 102, in which a toddler 206 is sitting. To determine the age of the toddler 206 from the image data, the control unit 136 performs a human body pose estimation procedure and, in particular, determines the length of a limb of the toddler 206.

[0044] Fig. Figure 3 shows a schematic representation of another image 300 of the vehicle interior taken by camera 134. In the Fig. In the situation shown in Figure 3, a taller child 302, approximately 10 years old, is sitting in the front passenger seat 202 of the vehicle 102. To determine the age of the child 302, the control unit 136 performs a human body pose estimation procedure and determines the length of at least one limb of the child 302. In other embodiments, the age of the child 302 can also be determined using a different method.

[0045] In the based on the Fig. In the embodiments described in 1 to 3, the camera 134 and the control unit 136 form a device for controlling a passenger airbag 146 of the vehicle 102. Further embodiments are described in the Fig.The elements and features shown in Figures 1 to 3 and mentioned in the preceding description can be part of this device. Instead of the sensor mat, a sensor unit can generally be used to determine at least one measurement of the weight of an object located on the passenger seat 202. The control unit 136 is also referred to as the processing unit. Reference symbol list 100 Cockpit 102 vehicles 104 drivers 106 Driver's seat 108 ads 110 Center console 112, 114, 116 Display element 118 Dashboard 120 air outlet nozzles 122 Steering wheel 124 Control element 126 gear selector 128 Pedals 130 Input unit 132 Windscreen 134 Camera 136 Control unit 138 Data output 140 Data input 142 Airbag control unit 144 Communication module 146 Passenger airbag 200, 300 image 202 Passenger seat 204 Baby car seat 206 Toddler 302 children 310 child seat

Claims

[1] Device for controlling a passenger airbag of a vehicle, with a sensor unit integrated into a passenger seat (202) which is configured to determine at least one measured value of the weight of an object located on the passenger seat (202), with at least one processing unit (136) which is at least configured to process the measured value transmitted from the sensor unit to the processing unit (136), to perform a classification of the object based on the measured value and at least one threshold value stored in the processing unit (136) and to determine a first classification result, with at least one camera (134) of the vehicle (102) which is designed to capture at least one image (200, 300) with an image of an area of ​​the vehicle interior encompassing at least part of the passenger seat (202) and to generate image data corresponding to the image (200, 300), wherein the processing unit (136) is further configured to process the image data, wherein the processing unit (136) is further configured to perform a classification of an object located on the passenger seat (202) based on the image data and to determine a second classification result, wherein the processing unit (136) is further configured to adjust the threshold in the event of a deviation between the first classification result and the second classification result, such that the first classification result matches the second classification result, and wherein the processing unit (136) is further configured to determine, starting from an adapted first classification result determined after adaptation during classification, a control parameter of an airbag control unit (142) for triggering the passenger airbag (146). [2] Device for controlling a passenger airbag of a vehicle, with at least one camera (134) of the vehicle (102) which is designed to capture at least one image (200, 300) with an image of an area of ​​the vehicle interior encompassing at least part of the passenger seat (202) and to generate image data corresponding to the image (200, 300), with at least one processing unit (136) which is at least equipped to process the image data, wherein the processing unit (136) is further configured to perform a classification of an object located on the passenger seat (202) based on the image data and to determine a classification result, wherein the processing unit (136) is further configured to determine, based on the classification result obtained during the classification, a control parameter for an airbag control unit (142) to trigger the passenger airbag (146), with a sensor unit integrated into a passenger seat (202) which is configured to determine at least one measured value of the weight of an object located on the passenger seat (202), wherein the processing unit (136) is further configured to process the measured value transmitted from the sensor unit to the processing unit (136), to perform a classification of the object based on the measured value and at least one threshold value stored in the processing unit (136) and to determine a further classification result, and wherein the processing unit (136) is further configured to determine an overall classification result based on the determined classification results of the classification and the further classification, and to determine the control parameter for the airbag control unit (142) based on the overall classification result. [3] Device according to any one of the preceding claims, characterized by , that the processing unit (136) is configured to transmit the control parameter to the airbag control unit (142), and that the airbag control unit (142) is configured to activate, deactivate or reactivate the passenger airbag (146) depending on further parameters. [4] Device according to any one of the preceding claims, characterized by, that in the classification, the further classification or the overall classification as classification result, firstly classification result, as second classification result, as further classification result and / or overall classification result “empty front passenger seat”, “empty child seat on the front passenger seat”, “empty infant carrier on the front passenger seat”, “child in a forward-facing child seat on the front passenger seat”, “child in a rear-facing infant carrier on the front passenger seat”, “child in a forward-facing infant carrier on the front passenger seat” and “adult person on front passenger seat”. [5] Device according to any one of the preceding claims, characterized by, that the processing unit (136) performs the classification and / or the second classification and / or the subclassification based on the image data using a human body pose estimation method, a pattern recognition method and / or a machine learning method. [6] Device according to any one of the preceding claims, characterized by , that the airbag control unit (142) is configured to deactivate the passenger airbag (146) depending on the control parameter which depends on the classification result and / or the subclassification result, and that the airbag control unit (142) is configured to reactivate the passenger airbag (146) depending on the control parameter and depending on the sensor signals from at least one airbag trigger sensor unit. [7] Device according to claim 6, characterized by, that the airbag control unit (142) reactivates the passenger airbag (146) upon detection of a collision at an impact speed on an obstacle greater than or equal to a preset stored speed threshold or at a medium or high expected severity of an expected or actual collision determined by the airbag control unit (142) based on sensor signals and a control parameter based on a classification of a child (206, 302) in a forward-facing child seat (310) and / or a subclassification of a 6- to 10-year-old child (302). [8] Device according to any one of the preceding claims, characterized by , that the field of view of the camera (134) has a detection angle in the range of 100° to 150°. [9] Method for controlling a passenger airbag of a vehicle, in which at least one measurement of the weight of an object located on the passenger seat (202) is determined with the aid of a sensor unit integrated into a passenger seat (202), in which the measured value is processed with the aid of at least one processing unit (136), a classification of the object is carried out based on the measured value and at least one threshold value stored in the processing unit (136), and a first classification result is determined. in which at least one image (200, 300) of the vehicle interior, encompassing at least part of the passenger seat (202), is captured using at least one camera (134) of the vehicle (102) and corresponding image data is generated from the image (200, 300), in which the image data are processed with the aid of the processing unit (136), whereby a classification of an object located on the passenger seat (202) is carried out based on the image data and a second classification result is determined, in which, with the aid of the processing unit (136), if the first classification result differs from the second classification result, the threshold is adjusted such that the first classification result matches the second classification result, and in which, with the help of the processing unit (136), after adaptation based on an adapted first classification result determined during classification, a control parameter of an airbag control unit (142) for triggering the passenger airbag (146) is determined. [10] Method for controlling a passenger airbag of a vehicle, in which at least one image (200, 300) of the vehicle interior, encompassing at least part of the passenger seat (202), is captured using at least one camera (134) of the vehicle (102) and corresponding image data is generated from the image (200, 300), in which the image data are processed with the aid of at least one processing unit (136), wherein a classification of an object located on the passenger seat (202) is carried out based on the image data and a classification result is determined, where, based on the classification result obtained during the classification, a control parameter for an airbag control unit (142) for triggering the passenger airbag (146) is determined, in which at least one measured value of the weight of an object located on the passenger seat (202) is determined with the aid of a sensor unit integrated into a passenger seat (202), in which a classification of the object is carried out based on the measured value and at least one threshold value and a further classification result is determined, and where, starting from the determined classification results of the classification and the further classification, an overall classification result is determined and, starting from the overall classification result, the control parameter for the airbag control unit (142) is determined.

Images