Quality inspection method of counterbore based on machine vision

Aiming at the problems of low inspection efficiency and unstable inspection accuracy of counterbore quality, a quality inspection method of countersunk hole based on machine vision was proposed.In this method, the image of counterbore is collected by an industrial camera, and the image is processed by Gaussian filtering and the required eigenvalues are extracted.By establishing the mathematical model of parallel projection of counterbore and modifying it with perspective projection, the vertical normal vector deviation and dimple depth of countersunk hole are obtained.Finally, the measurement test is carried out to verify the effectiveness of the method.The test results show that the error of verticality vector deviation is about 0.04° and the error of dimple depth is about 0.03 mm, which can meet the engineering requirements.