Distance-adaptive sensor fusion method, vehicle, and computer-readable storage medium storing instructions for performing distance-adaptive sensor fusion method

A vehicle comprising:

• at least one sensor configured to detect a target object present within a preset threshold distance from the vehicle during autonomous driving of the vehicle;
• a processor;
• and
• a memory storing at least one instruction that, when executed by the processor, is configured to cause the vehicle to:
• obtain radar feature information using a radar signal input from a radar sensor of the at least one sensor,
• determine radar bird's-eye-view (BEV) feature information based on the radar feature information,
• obtain ultrasonic feature information using an ultrasonic signal input from an ultrasonic sensor of the at least one sensor,
• determine ultrasonic BEV feature information based on the ultrasonic feature information,
• obtain image feature information using image data captured from an image sensor of the at least one sensor,
• update the image feature information based on first depth information associated with a first detection area of the radar sensor and second depth information associated with a second detection area of the ultrasonic sensor,
• perform view transformation on the updated image feature information,
• determine image BEV feature information based on the view-transformed image feature information,
• output a signal indicating detected objects based on fusion of the radar BEV feature information, the ultrasonic BEV feature information, and the image BEV feature information, and
• control, based on the signal, an operation of the vehicle.

A vehicle comprising:

• a plurality of sensors configured to capture sensor data associated with a surrounding environment of the vehicle, wherein the plurality of sensors comprises:
• a radar sensor configured to detect objects within a first detection area,
• an ultrasonic sensor configured to detect objects within a second detection area, and
• a camera configured to capture image data within a third detection area;
• a driving control circuit configured to control autonomous driving of the vehicle;
• and
• a processor circuit configured to cause the vehicle to:
• obtain radar feature information based on a radar signal of the radar sensor, obtain ultrasonic feature information based on an ultrasonic signal of the ultrasonic sensor, and obtain camera-image feature information based on the image data of the camera, and
• generate radar bird's-eye-view (BEV) feature information based on the radar feature information, generate ultrasonic BEV feature information based on the ultrasonic feature information, and generate camera-image BEV feature information based on the camera-image feature information, and
• determine a radar weight map, an ultrasonic weight map, and a camera-image weight map respectively corresponding to the radar sensor, the ultrasonic sensor, and the camera, wherein the radar weight map assigns a first weight to a region corresponding to the first detection area, wherein the ultrasonic weight map assigns a second weight to a region corresponding to the second detection area, and wherein the camera-image weight map assigns a uniform weight to a region corresponding to the third detection area,
• generate weighted BEV feature information based on applying the radar weight map, the ultrasonic weight map, and the camera-image weight map, respectively, to the radar BEV feature information, the ultrasonic BEV feature information, and the camera-image BEV feature information,
• output, based on the weighted BEV feature information, a signal indicating detected objects in the surrounding environment of the vehicle, and
• control, via the driving control circuit and based on the signal, autonomous driving of the vehicle.