Lane detection method based on smartphone sensor

Abstract

The invention discloses a lane detection method based on a smartphone sensor. The method comprises a sampling phase and a positioning and matching phase. The sampling phase comprises acquisition of training data, obtaining of rotation matrix R, and obtaining of corresponding threshold of target lane. Three-axis acceleration sensor data and gyroscope data of a cell phone are acquired on a known driving path, and centripetal acceleration and angular velocity and its radius-time characteristic curve of a vehicle are obtained, and the threshold is selected as the maximum radius value of the corresponding target lane; and the correspondence between the lane and the threshold is uploaded to a server. In the positioning and matching phase, the maximum radius value in the obtained radius-time characteristic curve matches the correspondence between the lane and the threshold obtained by a sampling module so as to determine the target lane. Without additionally arranging hardware equipment in the vehicle, the detection can be realized only by the use of the sensor in the smartphone. Thus, the detection method is low-cost and is easy to popularize.

Description

technical field [0001] The invention belongs to the technical field of road network navigation, and in particular relates to a smart phone-based lane detection method and system thereof. Background technique [0002] As unmanned vehicles become a reality, technologies related to unmanned driving have become a current hot research field, including precise lane positioning technology. The precise positioning of lanes has very important applications in the assisted driving system technology and the field of driverless driving, so it has become a research hotspot. [0003] Existing lane positioning methods still have the following deficiencies: first, some methods need to utilize special enhanced GPS positioning equipment, which is expensive and popular among car owners. For example, the publication number is CN105444770A and the The invention of "Smartphone-based lane-level map generation and positioning system and method" provides a smart phone-based lane-level map generation...

AI Technical Summary

Problems solved by technology

[0003] Existing lane positioning methods still have the following deficiencies: first, some methods need to utilize special enhanced GPS positioning equipment, which is expensive and popular among car owners. For example, the publication number is CN105444770A and the The invention of "Smartphone-based lane-level map generation and positioning system and method" provides a smart phone-based lane-level map generation and positioning system and method, including GPS, acceleration sensor, and direction sensor built into the mobile phone Obtain the real-time GPS, acceleration, and direction data of the vehicle; then the mobile client identifies the action point when the vehicle changes lanes and turns by monitoring changes in the acceleration sensor and direction sensor; the mobile client uploads the obtained GPS data to the server; the server will A map containing lane information is fed back to the user

This invention generates a lane-level map through the GPS information uploaded by the user. Although the accuracy of the map is improved, the cost is relatively high

Second, some methods use cameras to capture road markings and use image recognition technology to locate lanes. Such methods are greatly affected by external factors, such as being blocked by some obstacles in front of the vehicle, or the images captured by bad weather Not very recognizable

Images