Quality control is essential for any manufacturing enterprise
to determine the quality characteristics of a product or process on which product quality depends.
The main task of quality control is to prevent the appearance of defects and other inconsistencies of products with established requirements. Therefore, in the course of control, a constant analysis of deviations of product parameters from established requirements is carried out. As a result of the control, deviations from the requirements are revealed – inconsistencies and defects.
Quality control can be divided into three types:
Incoming control — checking the quality of raw materials and blanks entering production. Control of blanks and components at the initial stage allows you to prevent further production of the product at an early stage, which will save raw materials and time. Also, the constant analysis of the supplied components and materials allows you to control the quality of suppliers, seeking to improve it.
Inter-operational control covers the entire technological process.
Interoperational control is necessary to comply with technological regimes, rules for the manufacture, storage and packaging of products between operations.
Finishing control is used to establish the conformity of the quality of finished products with the requirements of standards or specifications, to identify possible defects and defects. QCD also checks the quality of packaging and the correct labeling of finished products.
Final inspection is used to determine whether the quality of finished products meets the requirements of standards or specifications, to identify possible defects and defects. The QCD also checks the quality of packaging and the correctness of labeling of finished products.
Integrity Inspection
Visual integrity inspection allows for the detection of visible violations of part integrity. Control of product integrity is often necessary in inline production of various materials as well as package integrity. The task is solved by means of vision systems based on video analytics (e.g. Fourier or wavelet analysis) or neural networks in more complex cases.
External Defect Detection
Visual defect inspection has thousands of tasks in various types of production. Video analytics in reflected and transmitted light, neural networks, and 3D laser scanning are used to detect defects depending on the task. Some defects are easily distinguishable for the algorithm, while others require expert evaluation. While in the past it was impossible to assess such defects with the vision system, nowadays, thanks to the use of neural networks with deep learning, this task is relatively easy to solve. At the same time they detect visible errors of surfaces – small and large risks, chafing, scuffing, traces of under-fusion, surface shells of corrosion or cavitation origin, flaking and pitting of fatigue origin, dents, chips, spalling, macro-cracks of different origin, etc.
Application control
Vision systems are ideal for checking the correct application of paint, drawings and texts. The system works on the basis of algorithms for color segmentation and comparison with a template.
