Research On The Online Examination Technology Based On Optical Characteristic Point Imaging

The computer vision measurement technology is a new bionic measurementtechnology, simulates the function of human eyes, using camera to collect the measuredtarget images, the camera image after image processing, the information acquired, themeasured object and the purpose of measurement. Computer vision measurementtechnology has been widely used in three dimensional various parts of the modernizationof industrial production in the processing, the assembly process and fixture positioningetc. Through in-depth study of the existing detection methods, analysis the special needsof detection technology in the practical production process, this paper studied the visioncoordinate measurement technology based on optical image, in order to solve the threedimensional precision on-line detection problem in industrial production.The vision measurement technology is realized by measuring the analysis ofmeasured target image, system uses the camera as the image acquisition device. Themeasured target imaging process caused by lack of depth information of feature points,5optical system using the optical probe position known as the actual imaging target, themissing depth information by using the constraint relationship between optical featurepoints in the process of image restoration. Systems using optical measuring any head oftwo optical feature points and camera image point to form a triangle constraint, establishthe measurement model of the system through the triangle cosine theorem, theleast-squares optimization method to calculate the spatial coordinates of opticalcharacteristic point. In the known optical feature points in the camera coordinate systemis the spatial coordinates and measurement coordinates head coordinates, the space poseoptimization algorithm, to determine the camera coordinate system and thetransformation relationship between head coordinates, calculated at the cameracoordinate system using the measuring tip in the probe coordinate system that is, themeasured coordinates of the object, measure.System through the measured point position determination on the optical image coordinates of the feature points, so the optical feature points of the image coordinates ofposition accuracy directly affects the precision of the measuring system. System of imagegray distribution according to the optical characteristic point——the infrared lightemitting diode is accord with the characteristic of Gauss function, determine the imagingcenter optical feature points by Gauss fitting location algorithm, increasing the number ofeffective calculation point by fractal algorithm, improve the positioning precisionpositioning algorithm, so as to improve the measuring precision of the whole system.Finally, the feasibility and accuracy of measurement method are verified by theexperimental results presented in this paper. The stability of the system by single pointrepeatability verification system, the measurement standard single point repeatability ofdifference is x0.011mm， y0.013mm， z0.021mm. The measurementaccuracy of the length of gauge block measurement verification system has beencalibrated,5groups of length measurement standard deviations are within0.045mm.The experimental results show that, the proposed measurement method is feasible, hashigh accuracy, can meet the demands of the precision measurement of three-dimensionalsize some occasions in the industry.