Convergent binocular vision guided robot positioning method in high dynamic range

Abstract

The invention discloses a convergent binocular vision guided robot positioning method in a high dynamic range, and the method comprises the following steps: firstly carrying out the quick calibration of a camera response curve of a sequence low dynamic image after distortion correction, building a mapping relation between a pixel value and exposure, and synthesizing a high dynamic range image; then introducing a chrominance space color correction algorithm, and carrying out linear interpolation between image chromatic aberration and achromatism to achieve an effect of adaptively correcting image chrominance information; correcting the three-dimensional coordinates of the workpiece model by calibrating the pose relationship between the convergent binocular camera and the robot; and finally, realizing the final target of guiding the robot to accurately position the to-be-machined area of the workpiece in combination with the motion path planned by the off-line programming technology. The synthesized high-dynamic image is clear in surface detail and balanced in color, the positioning precision of the robot on the target workpiece placed at any pose in the view field range is further improved, and a foundation is laid for follow-up visual guidance robot machining.

Description

technical field [0001] The invention belongs to the technical field of robot processing and positioning guided by machine vision detection, and relates to a convergent binocular vision algorithm under extended imaging dynamic range, in particular to a convergent binocular vision guided robot positioning method under high dynamic range. Background technique [0002] Usually the dynamic range of the complex scene where the target artifact is located is within 10 8 Left and right, the human eye can make the dynamic range reach 10 through the adjustment of its own visual system 7 about. However, the effective dynamic range of images acquired by ordinary industrial cameras is only 256. It can be seen that the display of target workpiece information is limited, which can easily lead to the loss of some detailed information after the camera is imaged, reducing the positioning accuracy of the robot. In order to improve the dynamic range of camera imaging, scholars at home and abro...

AI Technical Summary

Problems solved by technology

[0003] In view of this, the present invention integrates visual sensing technology, and proposes a robot positioning method guided by convergent binocular vision under high dynamic range, which effectively improves the camera imaging field of view, and solves the problem of insufficient dynamic range of the camera and image The problem of serious color loss further improves the positioning accuracy of binocular vision-guided robots

Images