A microwave radar device for identifying car models

Abstract

Disclosed is a microwave radar device for identifying vehicle models, comprising: a digital signal processor; a transmitting antenna array whose normal direction is at an angle with the road direction and includes at least one array element and a transmitting antenna array including at least two array elements In the first mode, all transmitting array elements transmit signals in turn, and all receiving array elements receive echo signals. In the second mode, one transmitting array element transmits signals, and one receiving array element receives echoes. signal; the digital signal processor is configured to: perform two-dimensional Fourier transform, incoherent accumulation, target detection, angle measurement and target tracking operations for the first digital signal in the first mode, thereby obtaining target information; for the first digital signal; For the second digital signal in the 2 mode, one-dimensional Fourier transform is performed, and then according to the target information, the backward projection algorithm is performed to obtain the image information of each point in the ground coordinate system; the vehicle recognition algorithm is performed for the image information to obtain each target. model information.

Description

technical field [0001] The invention relates to the field of microwave distance measurement and speed measurement, in particular to a microwave radar system for traffic control, and more particularly to a microwave radar device for identifying vehicle types. Background technique [0002] The recognition technology for vehicle models has a broad application space in intelligent traffic scenarios such as traffic light timing and ETC inspection. At present, the technologies for identifying vehicle models mainly include the following three types: [0003] 1. Obtain image data through visual sensors. The image data is rich in target feature information, and the brand, model, color and other information of the vehicle can be extracted by using deep learning technology. However, in extreme weather such as heavy rain, heavy snow, and heavy fog, or in low-light conditions such as night, the camera cannot obtain effective image data, and the vehicle type recognition function fails. ...

AI Technical Summary

Problems solved by technology

However, the angle resolution performance of microwave radar is far inferior to that of cameras and lidar; although some target features can be extracted through micro-Doppler technology and spectrum analysis technology, these features are not robust enough, resulting in poor vehicle recognition effect. good

For example, the existing patent with the title of invention "Method for Vehicle Type Recognition Based on Doppler Traffic Radar" and publication number CN102360528B performs fast Fourier transform on the time-domain echo signal of the radar, and then extracts Doppler Spectrum features to infer vehicle length and other contour features, but its recognition accuracy and resolution are not high, and the detection distance is short

Images