Study On Horizontality Measurement Methods For Scale Platform Based On Computer Vision

Aiming at the real-time levelness measurement problem of the existing elevating steel weighing platform in the steel plant,this paper proposes a computer vision-based weighing platform levelness measurement method,which uses line structured light and LED feature marks as the cooperation target.The content of this research is as follows:For the camera model calibration problem,the traditional camera model calibration and virtual image plane camera model calibration methods are used.The calibration of the conventional camera model is performed by using the Zhang Zhengyou calibration method to obtain the internal and external parameters and lens distortion coefficients of the camera,and the obtained parameters are used to calibrate the virtual image plane camera model,and to establish the coordinate system of the virtual image plane and generate the virtual image of the scale.For the feature extraction problem of feature marks and line structured light,an image processing and shape description parameter approach is used.The virtual image of the weighing platform is subjected to basic image processing operations,and the shape description parameters are set to extract features from the line structured light and feature marks on the virtual image based on the consideration of the imaging shapes of the feature marks and line structured light.The least squares method is used to fit the linear equation of the line structured light for the equation solution problem.For the problem of solving the center point coordinates of the feature marker,the center point of the minimum enclosing rectangle of the vertical boundary of the point set of the feature marker region is used as the center point of the feature marker on the virtual image.For the problem of measuring the vertical height of feature marker point-to-line structured light,a monocular vision-based solution method is used.The corresponding mathematical expressions are solved by establishing the corresponding mathematical expressions through the known geometric and perspective projection relationships.For the problem of calculating the level of the scale platform plane,the method of fitting the equation of the scale platform plane using the least squares method for the four characteristic mark center points and calculating the angle between the fitted scale platform plane and the reference plane is used.And the experimental results and conclusions show that the proposed method of measuring the level of the scale platform based on computer vision has the validity.The innovations of this topic are:(1)a device can be designed that can be used to detect the levelness of the scale platform detection,and this topic uses the light strip projected by the line structure light projector on the side of the scale base as the measurement reference,which does not need to install additional fixed reference marks on the steel scale:(2)this topic proposes a computer vision-based method for the levelness of the elevated steel scale platform,and the process of the method to measure the level of the scale platform is a non-contact measurement process,which does not affect the weighing process of the steel scale,and has objectivity and realtime.