Optimum Design Of Visual Sensor Structure Based On Scheimpflug Condition

Three-dimensional morphology of line structured light has the advantages of strong anti-interference ability,non-contact and high measurement resolution,and is widely used in industrial measurement.Recently,with the high-speed development of the semiconductor industry,small workpieces such as semiconductor components are massively manufactured and applied among various precision instruments.The demand for accurate three-dimensional measurement of small workpieces under high magnification has also increased.High magnification means a smaller field of view.Since the depth of field is negatively correlated with the system magnification,the measured object in the close-range measurement can only be clearly imaged in a small range,and the part outside the depth of field will out of focus,many characteristic information will be missing,which seriously affects the measurement accuracy.In the field of photography,when the camera meets the Scheimpflug condition of the focus plane,the lens plane and the image plane are intersected,any point on the focus plane can be clearly imaged.In line structured light measurement,if the laser plane coincides with the focus plane,the depth of field of the imaging system can be greatly expanded.Based on the Scheimpflug condition and structured light measurement technology,this paper establishes a sensor measurement model,analyzes and optimizes the layout of structural parameters,builds a line structured light sensor based on this model and writes related algorithms to verify the feasibility of the theory.The main research contents and results of this paper are as follows:Outlines the principle of extended depth of field of line structured light imaging system under the Scheimpflug condition,and a method to refocus the imaging system by focusing control when the imaging system is out of focus is given.Combines Scheimpflug condition and the laser triangulation sensor ideal measuring model is set up,on the basis of the thin lens model directly into the lens thickness,thick lens ideal measurement model is established,and the image coordinate transformation formula is derived in detail.The influence of structural parameters and imaging position on imaging magnification and sensitivity of the sensor is analyzed by MATLAB,and the optimal layout of sensor structural parameters is determined.In the actual measurement,the plane where each device is located in the sensor construction process deviates from the expected position due to the influence of installation error and does not meet the Scheimpflug condition,resulting in the deviation of the final calculation results.In this paper,the modes of the plane excursions are analyzed,and the plane excursions are summarized as the synthesis of several basic spatial motions,and these excursions are substituted into the measurement model as parameters.In the part of image algorithm,this paper proposes an adaptive filter to denoise and enhance the fringe images with low brightness.DBSCAN clustering algorithm is combined with Steger algorithm to extract the center point of the fringe and eliminate the redundant points generated during the extraction process.Based on the measurement model,the homography relationship between object point and image point was established,and the calibration of structural parameters was completed by nonlinear optimization.According to build line structured light vision sensor measurement model,and the system structure parameter calibration,based on the standard measurement to determine the measurement error of sensors,characteristics of small workpiece surface topography measurement to verify the rationality of the theory in this paper.