A MACHINE VISION SYSTEM FOR GEAR TEETH DEFECT DETECTION

Gears, one of the indispensable components used in the industry, are mechanical elements that ensure efficient energy transmission, altering the speed and torque of rotational movements. The reliability and durability of gears directly affect the overall performance of related systems. Recently, gear manufacturing has been nearly fully automated with the help of advanced technology. However, it is common to assess the quality of a gear via traditional methods. The conventional quality control techniques for gear quality determination cause many difficulties, such as time-consuming and user-dependent measurement errors. In short, these conventional measurement methods decrease manufacturing speed. Today, Machine Vision Systems (MVS) offer the possibility to advance automated quality control systems. In this paper, to save time and reduce user-dependent errors, an automated gear evaluation system was developed for integration into a mass production line. The developed system has a rotating table, with gears progressing on the table at a controllable rotating speed. The gears are inspected for common defects such as missing teeth, rough surfaces, incorrect diameters, and other flaws. The detection process uses an MVS, programmed to differentiate perfect gears from defective ones through a vision system. The detected defective gears are automatically separated by pushing from the production line using compressed air via a pneumatic valve. This system enhances the efficiency of the production line and prevents defective gears from advancing to subsequent stages of production or assembly. As a result of the experiment, the standard deviation of both defective and perfect gears was measured below 1%, which is an indication of high measurement precision. The developed system provides high-speed quality control in mass production processes, thus aiming to increase efficiency by minimizing user-dependent measurement errors on mass production lines.