A roadside perspective over-the-horizon global fusion perception system based on deep learning

Abstract

The invention relates to a roadside viewing angle over-the-horizon global fusion perception system based on deep learning, which collects and fuses heterogeneous data of cameras, laser radars and millimeter-wave radars arranged on the roadside, and uses deep learning to process the fusion respectively. The final image and lidar point cloud data are finally used to achieve fusion perception through decision layer fusion. By arranging the perception system on the roadside, and combining the respective advantages of image, lidar point cloud, and millimeter wave radar point cloud data, the perception distance can be improved, and the scene understanding of the monitoring area can be carried out at multiple levels, and finally the network is automatically connected. Driving a vehicle provides sufficient and reliable sensory information.

Description

technical field [0001] The invention relates to an intelligent traffic infrastructure system, in particular to a roadside viewing angle over-the-horizon global fusion perception system based on deep learning. Background technique [0002] The development of intelligent transportation infrastructure is the key to realizing the function of connected autopilot. As the source of roadside sensing information in the vehicle-road collaborative sensing system, the stability and reliability of the roadside perception system will directly affect the performance of the connected autopilot function. implementation and security. Today, with the rapid development of intelligent transportation technology and autonomous driving technology, the stability and reliability of the roadside perception system are still the key points that need to be solved urgently. [0003] An autonomous driving system usually includes three modules: perception, decision-making, and control. Perception is the in...

AI Technical Summary

Problems solved by technology

Multi-sensor data fusion has expanded the perception distance and optimized the perception viewing angle to a certain extent, but the limitations of the perception method based on vehicle sensors in terms of field of view, viewing angle, price, etc. cannot be completely overcome due to the location of the sensor, and road congestion conditions Occurs often

Images