METHOD AND APPARATUS FOR DETECTING PEDESTRIAN AND VEHICLE BASED ON CONVOLUTIONAL NEURAL NETWORK USING STEREO CAMERA

摘要

According to the present invention, provided is an object detection apparatus which comprises: an object candidate detection unit for acquiring a disparity image through stereo matching of a stereo image captured from a stereo camera including two or more lenses; and an object detection unit for detecting either one or both of a vehicle and a pedestrian among object candidates detected by the object candidate detection unit.