The Research Of Visual Inspection Method For The Gear Geometry Parameters

As the artificial detection accuracy of spur gears geometrical parameters is low,damaging the tooth surface and low working efficiency,it adopted the method of machine vision to detection gear geometry parameters.It constructed visual inspection platform by using foreground-background comprehensive method of improvement measures.It compared the main pixel edge detection algorithms,which found the traditional Canny edge detection algorithm was lost edge,error detection and filtering,threshold parameters need to manually set.It put forward an improved adaptive Canny edge detection algorithm.Mainly includes:Using the integral image simplified adaptive Gaussian to filter the noise;Increasing 45 ° and 135 ° direction of improved Sobel gradient template to gradient calculation;Otsu method is adopted to realize the double threshold detection.The experimental results show that the improved Canny edge connectivity is better,and improved the effect of image edge detection.In order to achieve the high accuracy of the gear testing and maintain good noise resistance,it adopt the interpolation method to get the pixel and uses the least squares method to extract the gear edge pixels by the parameter calculation and calibration coefficient to get the actual geometry size.The result shows that it realizes the non-contact measurement of gear geometric parameters,the pixel level precision and error is 0.00435 mm.It meets the actual requirements,and it is widely used in the manufacture because of its application value.