Distance measurement method based on monocular freespace

Abstract

The invention relates to the technical field of distance measurement and automatic driving and provides a distance measurement method based on monocular freespace. The method comprises steps that a camera coordinate system, external parameters of the ground and internal parameters of a camera are calibrated through a calibration plate arranged on the ground in front of the camera in parallel; inverse perspective transformation is carried out for an image which is shot by the camera and comprises the calibration plate to obtain an inverse perspective projection pattern, and the actual size K represented by a single pixel in the inverse perspective projection pattern is calculated; a distance from the camera to a point actually represented by a point in the middle of a lower edge of the inverse perspective projection pattern is manually measured to obtain a camera blind area distance M; the distance L from a target point to the camera is obtained through a following monocular distance measurement formula of L=M+k*v, and the v is the number of pixels from the target point to the lower edge of the inverse perspective projection pattern. The distance measurement method is advantaged inthat inaccurate measurement caused by height and angle problems in the prior art can be overcome, and errors of measurement results are reduced, moreover, the operation process can be simplified, operation complexity is reduced, props are used strategically, and the freespace distance measurement effect is improved.

Description

technical field [0001] The present disclosure relates to the fields of distance measurement technology and automatic driving technology, and in particular, to a single-purpose freespace distance measurement method. Background technique [0002] In today's mainstream autonomous driving technology, freespace technology has become an indispensable part. Freespace, that is, the area where cars can drive, including avoiding other cars, pedestrians, and roadside areas. This technology not only uses image recognition algorithms, but also involves geometric measurement algorithms in computer vision, that is, to obtain the distance of the target area in the camera coordinate system. Commonly used methods include monocular and binocular. Monocular ranging is simple and effective, does not require a large amount of calculation, and has low power consumption, while binocular ranging has the characteristics of complex structure and long matching time. As far as the current vehicle stat...

AI Technical Summary

Problems solved by technology

Due to the height and angle information, manual measurement is required, which will inevitably introduce measurement errors, resulting in a great impact on the accuracy of the final measurement plan

[0004] In addition, due to the inaccurate measurement of height and angle, the existing technical methods will cause too much error in the result of monocular distance measurement, making it difficult to apply

And the operation is complex and time-consuming

Images