Abstract <p>Due to Western sanctions, deliveries to Russia of high-quality, high-strength fasteners for the engine assembly are currently halted. Bolts that were previously imported underwent continuous surface quality control at the manufacturers. Bolts produced by domestic enterprises, as a rule, are not subjected to proper inspection. It is necessary to detect surface discontinuities in bolts and to reject parts containing unacceptable defects: cracks, gouges, tears, scoring, scratches and dents. In service, engine components, including fasteners, operate under elevated reversing (alternating) loads. The presence of manufacturing defects leads to accelerated wear and fatigue damage. To prevent premature bolt failure, rejection nondestructive testing (NDT) must be performed before installation. Bolt surface defects are generally detected using magnetic NDT methods, among which the most promising and widely used is the magnetic particle inspection method. Its main drawback is the appearance of an indicator pattern on “false” defects (false indications). This paper presents an effective solution to this problem using automated NDT technology and machine vision.</p>