Method for automatically acquiring vehicle training sample based on multi-modal sensor data

Abstract

The invention discloses a method for automatically acquiring a vehicle training sample based on multi-modal sensor data. The method comprises the following steps of: detecting a vehicle based on laser data and positioning data, namely acquiring a two-dimensional coordinate relative to a data acquisition vehicle according to a distance and an angle of the laser data and a laser sensor calibration parameter so as to describe horizontal contour information of an object; and extracting a time sequence of parameters such as the position and direction of a candidate vehicle relative to the data acquisition vehicle by analyzing a shape and detecting and tracking a mobile object; and extracting a vision image sample, namely projecting the candidate vehicle into an image according to the position and direction of the candidate vehicle at each moment on the basis of a geological relation between the laser sensor and image acquisition equipment to produce a region of interest, correcting the region of interest by using a detector, calculating a relative view angle of each candidate vehicle relative to a camera according to the parameters such as position and direction, removing image frame samples with similar view angles, and automatically extracting sample pictures of the candidate vehicle under different view angles.

Description

technical field [0001] The invention relates to the technical fields of computer vision, robot and machine learning, in particular to a method for automatically acquiring vehicle training samples based on multimodal sensor data. Background technique [0002] Vehicle detection is an important problem in the field of automotive driver assistance systems (ADAS). There has been a large amount of related research in the field of vehicle detection, which has proved that vehicles can be detected using lasers, radars, monocular / stereo cameras, and multi-sensor fusion. [0003] Due to the low cost of monocular cameras and the simple calibration problem, detection methods based on monocular vision have been extensively studied in the fields of computer vision and robotics. When using visual sensors, the appearance of the vehicle itself and the appearance of the vehicle at different angles are very different, which brings great difficulties to detection. Recently, more and more resea...

AI Technical Summary

Problems solved by technology

[0009] However, all the samples in this dataset do not contain the pose information of the vehicle. Usually, the training data needs to be manually labeled and classified according to the training requirements, which greatly limits the number and performance of samples.

This has become a bottleneck restricting the development of algorithms

Detector performance, often lacking stability to changes in the environment

Images