Outer parameter calibration device and calibration method of laser radar and visual camera

Abstract

The invention relates to an outer parameter calibration device and calibration method of a laser radar and a visual camera. The device comprises a support with universal wheels, a telescopic sleeve, an expansion link inserted into the upper end of the telescopic sleeve and of which the position is fixed by an expansion link fixing nut, a ball cup fixed at the upper end of the expansion link, a ball support connected with the call cup and a calibration plate connected with the ball support arranged in sequence; the calibration plate is a circular black plastic plate with a smooth surface, and acircular hole is formed closed to the edge position of the circular black plastic plate; square slots are milled in the surface of the calibration plate according to a checkerboard, and white squarealuminum alloy sheets are embedded into the slots, so that the aluminum alloy sheets and the calibration plate are formed into a rectangular black-white checkerboard form with the smooth surface. Compared with the prior art, the outer parameter calibration device and calibration method have the advantages of being practical, simple, convenient and strong in universality.

Description

technical field [0001] The invention relates to a calibration technology between sensors, in particular to an external parameter calibration device and a calibration method for a laser radar and a visual camera. Background technique [0002] In recent years, artificial intelligence technology has been continuously improved, and some functional algorithms have gradually been applied to the field of driverless car technology. Especially in computer vision technology, deep convolutional neural networks bring solutions to many problems that cannot be modeled. The neural network can process a large amount of collected experimental data and use supervised learning to complete the fitting of the mapping function. However, since the neural network needs to be supported by big data and requires high hardware configuration, the current intelligent driving system cannot replace human driving in complex traffic environments through a purely visual solution. At the same time, the visua...

AI Technical Summary

Problems solved by technology

[0004] First, the calibration device of single-line laser radar is too complicated, it is difficult to ensure the accuracy of the structure during processing, and it is easy to deform during use

At the same time, the same calibration device has not yet been used to simultaneously calibrate the external parameters of single-line lidar, multi-line lidar and visual camera

[0005] Second, the calibration method based on the natural environment is largely affected by the uncertain environment, especially in the environment with high light intensity during the day, the shade of trees and the positive and negative sides of buildings are likely to cause non-convergence of the parameters

A large amount of collected data is required for iteration, which increases the workload and calculation

Images