Vehicle cabin-based target detection and tracking

A method for tracking a target in a vehicle interior of a motor vehicle, the method comprising:

• receiving from a target sensor, via an electronic control unit (ECU) of the motor vehicle, a sensor data stream of a seat surface in the vehicle interior;
• segmenting the sensor data stream into an output file via a fully convolutional neural network (FCN) of the ECU;
• detecting the target on the seat surface using the output file from the FCN, including
• calculating a segmentation mask (M) from the output file, extracting a largest fully-connected region in the segmentation mask (M) using a Depth First Search (DFS) to identify pixel clusters or regions, and extracting the largest fully-connected region from amongst the pixel regions or clusters;
• classifying the target into a target class via the ECU;
• and
• executing a control action, via the ECU, in response to the target class.

A motor vehicle comprising:

• a vehicle body defining a vehicle interior;
• and
• a target tracking system for use in the vehicle interior, the target tracking system comprising:
• at least one target sensor connected to the vehicle body and situated in the vehicle interior, wherein the at least one target sensor is configured to output a sensor data stream of a seat surface in the vehicle interior, and wherein the target sensor includes a red, green, blue (RGB) camera;
• and
• an electronic control unit (ECU) in communication with the at least one target sensor, the ECU being configured to:
• receive the sensor data stream of the seat surface from the at least one target sensor;
• segment the sensor data stream into an output file using a fully convolutional neural network (FCN);
• detect a target on the seat surface using the output file, including calculating a segmentation mask (M) from the output file, extracting a largest fully-connected region in the segmentation mask (M) using a Depth First Search (DFS) to identify pixel clusters or regions, and extracting the largest fully-connected region from amongst the pixel regions or clusters;
• classify the target into a target class;
• and
• execute a control action in response to the target class.