How does quality control by machine vision work?

On any assembly line there is always a margin for error when a part is manufactured or assembled. 

Quality control by machine vision is a system which works by taking multiple images of each part. This allows any deviation from previously stipulated parameters to be identified instantly. 

Each part is scanned by a sensor or camera, allowing precise details to be checked for variations in dimensions, imperfections in form, colour, missing parts, or superficial and internal defects.

Unlike the human eye, a machine can carry out this task automatically and without any interruptions. Moreover, it can take decisions immediately based on the information received and is consequently more efficient and much more cost-effective. 

Machine vision is used in a wide variety of applications, depending on the sector in question and the type of product manufactured on each production line,

• Inspection: This type of visual system is frequently used in industrial applications to ensure homogeneity in production. This system detects missing parts, deformations, measures that the same quantity of liquid is contained in each package etc. 

• Production line monitoring: Hyperspectral cameras are used to detect unwanted elements on the assembly line by measuring the wavelength of each part. 

• Checking: Machine vision can also be used to identify almost imperceptible defects in the final product, such as a different colour in a piece of fruit, a label which has not been perfectly applied or the slightest of deformations in the finished article. 

• Measuring: Certain products are also measured using machine vision to corroborate that they are in perfect condition and that there are no manufacturing errors such as variations in depth, distance, width, length etc. 

• Barcode verification: The use of this technology to read barcodes is habitual as this confirms that they can be read by other devices.