Measurement Study on Geometric Parameters of Medium Module Straight Tooth Cylindrical Gears

Aiming at the problems of low efficiency, expensive equipment, and limited accuracy in traditional contact measurement of medium modulus gears, this paper proposes a non-contact gear geometric parameter measurement method based on machine vision. The method first performs preprocessing operations such as Gaussian smoothing filtering and Sobel sharpening on gear end-face images to enhance edge quality; then establishes a least squares-based center circle fitting algorithm to achieve sub-pixel accuracy gear center positioning; employs the Canny edge detection algorithm to extract tooth profile edge curves and calculate radius parameters of addendum and dedendum circles. Experiments were conducted using four spur gears with a modulus of 5 as samples. Results show that the system achieves a measurement resolution of 0.0357 mm, with a maximum error of the addendum circle radius of 0.342 mm. This system enables automatic and efficient measurement of geometric characteristic parameters for medium modulus gears, providing a low-cost, high-precision technical solution for quality control in gear manufacturing processes, demonstrating significant engineering practical value and application potential.