Research On The Online Detection System Of Machine Vision Image Geometry

With the improvement of the automation degree of production technology,the requirements of geometric quantity detection technology are becoming higher and higher,especially the online detection technology.It is not only necessary to consider whether the precision meets the production demand,but also a big test in the detection speed.In order to measure the workpiece's geometric quantity in real time,this paper designs an on-line measuring system for measuring the workpiece's geometric quantity.When the workpiece passes through the image acquisition system,the image information can be obtained immediately.The axial length is measured by the principle of image processing.Under the condition of ensuring the detection accuracy,in order to accelerate the detection speed.In this paper,MATLAB,Halcon and other software to compare and analyze various image processing algorithms and choose the appropriate image processing algorithm.The circuit design and program design are completed by using FPGA real-time pipeline operation and characteristic development and verification of parallel processing tasks.In order to improve the resource utilization and operation speed,the circuit is optimized and improved to reduce the consumption of logic resources.According to the principle of phase grouping and the abrupt change of phase at the inflection point,a method to quickly screen the corner points to be measured is designed.This method reduces the amount of interpolation data,not only satisfies the measurement requirements in precision,but also has some advantages in design cost and speed,and has smaller packaging.The specific content is as follows:1)The image acquisition platform was set up and the appropriate device model was selected.The lighting mode and image acquisition mode are compared and analyzed according to the experimental tasks.According to the Ethernet protocol and the working principle of SDRAM,the state jump steps of its image transmission and storage are analyzed,which lays a good foundation for the subsequent FPGA design.2)The distortion was corrected by camera calibration and the actual size of each pixel was calculated.Then gray statistics and linear transformation are carried out to enhance image contrast.The advantages and disadvantages of various edge detection operators are compared and analyzed through experiments,and the appropriate detection operator is selected to facilitate subsequent measurement.3)The distortion was corrected by camera calibration and the actual size of each pixel was calculated.Grayscale statistics and linear transformation are used to enhance image contrast.The advantages and disadvantages of various edge detection operators are compared and analyzed in the experiment,and the appropriate detection operator is selected to facilitate subsequent measurement.4)When measuring the geometrical quantity of the workpiece,the corner points to be measured can be quickly screened based on the characteristics of phase grouping and phase change at inflection points.The sub-pixel positioning technology of interpolation method is used to locate the corner measurement points more accurately,which avoids the interpolation operation for the invalid pixel points and effectively improves the operation speed.The error range and the source of the error are analyzed by measuring the workpiece many times from different angles.The experimental results show that the measurement system designed in this paper can effectively measure the geometric size of the workpiece in real time,and the average error is within the range of 2 pixels.Compared with the traditional detection method or detection software,the detection speed is also significantly improved.