What Is a Visual Inspection System?
A visual inspection system can inspect the appearance of mass-produced goods. For a large number of manufactured products, the introduction of equipment can improve productivity and reduce prices rather than human inspection.
The system inspects and instantly determines the appearance of foreign matter, dirt, burrs, cracks, breaks, and chips on the surface of manufactured parts and products. Processes for visual inspection include inspection at the manufacturing stage, after surface treatment, and after product assembly. Inspection by the equipment requires clearly defined standards, and a dot gauge is used to determine whether a product has passed or failed.
Uses of Visual Inspection Systems
The visual inspection systems are used in many areas of the manufacturing industry. Food packaging, bottles, electronic devices, medical devices, molded sheets, and many other industries produce so much that they could not keep up with production without visual inspection systems. Improvements in camera and image processing technology have made it possible to determine even the smallest foreign objects, scratches, and defects.
Specifically, it can detect differences in the shape of electronic components, rust, corrosion, and dents in bearings, as well as burrs, cracks, and deformation in bolts and screws. It can also detect burns, dents, and even scratches on food containers.
Principle of Visual Inspection Systems
The image processing inspection method is the most commonly used method for visual inspection. It consists of an image sensor such as a camera, an image processor, and software. The object is illuminated evenly and brightly, and a lens captures the image for inspection.
The system records the shading of each pixel in the image of many passed products in advance. The mean value and standard deviation of the difference in density are determined by dividing the image into certain sections, and the values are compared with the pixel value of the inspected object for each section. If the density differs significantly from that of the past product, it is judged as a flaw or foreign matter. The dimensional criteria for scratches and foreign matter are registered to determine whether the product has passed or failed the inspection. Differential or projection processing may be performed to prevent false positives.
In the case of circumferential objects, polar coordinate transformation processing is used to compare densities and make a judgment.
Image sensors are highly versatile and enable comprehensive inspection, but the initial cost of the equipment itself is pretty high. In addition, it is essential to optimize the software according to the shape of the object and the visual inspection criteria, which can be time-consuming.