NANJING UNIVERSITY OF SCIENCE AND TECHNOLOGY

The present disclosure provides a multi-task panoptic driving perception method and system based on improved You Only Look Once version 5 (YOLOv5). The method in the present disclosure includes: performing image preprocessing on an image in a dataset to obtain an input image; extracting a feature of the input image by using a backbone network of improved YOLOv5, to obtain a feature map, where the backbone network is obtained by replacing a C3 module in a backbone network of YOLOv5 with an inverted residual bottleneck module; inputting the feature map into a neck network to obtain a feature map, and fusing the obtained feature map and the feature map obtained by the backbone network; inputting the fused feature map into a detection head to perform traffic target detection; and inputting the feature map of the neck network into a branch network to perform lane line detection and drivable area segmentation.

The present disclosure provides an expressway vehicle speed measuring method and device based on roadside monocular camera calibration. The expressway vehicle speed measuring method based on roadside monocular camera calibration includes: acquiring a traffic monitoring video in a roadside monocular camera; segmenting the traffic monitoring video into a traffic background image set in frames, and transmitting the traffic background image set to a well-trained multi-target detection algorithm to obtain a pixel position of a target vehicle; tracking a target vehicle detection result in each frame through a multi-target tracking algorithm to obtain a movement locus of the target vehicle; and calculating an actual moving distance of the target vehicle according to the movement locus of the target vehicle and a fitted pixel-to-world coordinate conversion curve, thereby calculating a speed of the target vehicle. The present disclosure can achieve a higher accuracy and a lower operational complexity in vehicle speed detection.

The disclosure is a three-dimensional towered checkerboard for multi-sensor calibration, and a LiDAR and camera joint calibration method based on the checkerboard. The joint calibration method includes: establishing a modeling coordinate system taking the three-dimensional towered checkerboard as a basis, and generating a point cloud of the three-dimensional towered checkerboard; denoising a three-dimensional point cloud obtained by LiDAR, and obtaining an actual point cloud of the three-dimensional towered checkerboard under a LiDAR coordinate system; determining a transformation relationship between the LiDAR coordinate system and the modeling coordinate system; generating a corner point set of two-dimensional checkerboards under the modeling coordinate system in sequence according to actual positions of corners of the two-dimensional checkerboards, and transforming into the LiDAR coordinate system; obtaining a corner point set of the two-dimensional checkerboards on a photo; and calculating a transformation relationship between the camera coordinate system and the LiDAR coordinate system.

The disclosure is a three-dimensional towered checkerboard for multi-sensor calibration, and a LiDAR and camera joint calibration method based on the checkerboard. The joint calibration method includes: establishing a modeling coordinate system taking the three-dimensional towered checkerboard as a basis, and generating a point cloud of the three-dimensional towered checkerboard; denoising a three-dimensional point cloud obtained by LiDAR, and obtaining an actual point cloud of the three-dimensional towered checkerboard under a LiDAR coordinate system; determining a transformation relationship between the LiDAR coordinate system and the modeling coordinate system; generating a corner point set of two-dimensional checkerboards under the modeling coordinate system in sequence according to actual positions of corners of the two-dimensional checkerboards, and transforming into the LiDAR coordinate system; obtaining a corner point set of the two-dimensional checkerboards on a photo; and calculating a transformation relationship between the camera coordinate system and the LiDAR coordinate system.

The disclosure is a three-dimensional towered checkerboard for multi-sensor calibration, and a LiDAR and camera joint calibration method based on the checkerboard. The joint calibration method includes: establishing a modeling coordinate system taking the three-dimensional towered checkerboard as a basis, and generating a point cloud of the three-dimensional towered checkerboard; denoising a three-dimensional point cloud obtained by LiDAR, and obtaining an actual point cloud of the three-dimensional towered checkerboard under a LiDAR coordinate system; determining a transformation relationship between the LiDAR coordinate system and the modeling coordinate system; generating a corner point set of two-dimensional checkerboards under the modeling coordinate system in sequence according to actual positions of corners of the two-dimensional checkerboards, and transforming into the LiDAR coordinate system; obtaining a corner point set of the two-dimensional checkerboards on a photo; and calculating a transformation relationship between the camera coordinate system and the LiDAR coordinate system.

A method for vehicle detection based on single-axis geomagnetic sensor is disclosed. The method includes the following: arranging two geomagnetic sensors in tandem at a fixed distance in the same lane, and acquiring, by the two geomagnetic sensors, data when a vehicle passes by; performing data cleansing and denoising on the data acquired by the geomagnetic sensors; detecting, by the geomagnetic sensors, the vehicle by using an state machine algorithm based on a fixed threshold, and extracting moments when the vehicle enters and leaves detection ranges of the sensors; and collecting statistics on traffic flow in a predetermined period of time, and calculating the velocity and length of the vehicle. The presently disclosed method and apparatus can be independently used to obtain vehicle parameters at important traffic sections or intersections, or can be used in combination with other traffic detection devices to obtain more comprehensive traffic information. The presently disclosed method and apparatus has low power consumption, high detection sensitivity, is easy to install and resists the impact of a harsh environment.

A method for vehicle detection based on single-axis geomagnetic sensor is disclosed. The method includes the following: arranging two geomagnetic sensors in tandem at a fixed distance in the same lane, and acquiring, by the two geomagnetic sensors, data when a vehicle passes by; performing data cleansing and denoising on the data acquired by the geomagnetic sensors; detecting, by the geomagnetic sensors, the vehicle by using an state machine algorithm based on a fixed threshold, and extracting moments when the vehicle enters and leaves detection ranges of the sensors; and collecting statistics on traffic flow in a predetermined period of time, and calculating the velocity and length of the vehicle. The presently disclosed method and apparatus can be independently used to obtain vehicle parameters at important traffic sections or intersections, or can be used in combination with other traffic detection devices to obtain more comprehensive traffic information. The presently disclosed method and apparatus has low power consumption, high detection sensitivity, is easy to install and resists the impact of a harsh environment.