4D millimeter wave radar point cloud target detection method based on Darknet53 network

• Main Authors: CAI YUHENG, ZHOU GUAN, ZHOU RUCHEN
• Format: Patent
• Language: Chinese; English
• Published: 19.01.2024
• Subjects: ANALOGOUS ARRANGEMENTS USING OTHER WAVES; CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; IMAGE DATA PROCESSING OR GENERATION, IN GENERAL; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION ORRERADIATION OF RADIO WAVES; MEASURING; PHYSICS; RADIO DIRECTION-FINDING; RADIO NAVIGATION; TESTING

The invention discloses a Darknet53 network-based 4D millimeter wave radar point cloud target detection method, which belongs to the field of computer deep learning and automotive electronics, and mainly provides a new 4D millimeter wave radar target detection method, which comprises the following steps of: in a pre-processing part, projecting and dispersing a 3D radar point cloud into a bird's-eye view 2D grid map; a height feature, a speed feature, a reflection intensity feature and a grid density feature form a four-channel feature pseudo graph to be input into a detection network, and the calculation amount of the network can be greatly reduced. A Darknet53 network is used as a main body target detection network framework, and a pooling pyramid (SPP) module is added to realize fusion of local features and global features of the whole network, so that the expression ability of a feature map can be effectively enriched. According to the method, target features can be efficiently and accurately extracted fro