Vision Systems

Vision systems at the industrial level are increasingly important to guarantee the quality of the products and the efficiency of the lines in which they are inserted. These systems generally consist of cameras, lighting systems, filters and control software. The vision system to choose for a project and the way in which it will be applied will depend on many factors.

The stability of these systems is easily influenced, as there are many variables to take into account. The ambient light on the factory floor varies not only during the day, but also during the year. In this way, an attempt should be made to close the pieces in a controlled lighting environment, but this is not always possible.

The color of a part significantly affects the checks to be carried out. In these situations, color filters can be applied, which, combined with colored LED lighting, allow creating visual contrast and counteracting the effects of colors on surfaces. Another solution involves the use of lighting systems outside the visible light spectrum, such as infrared or ultraviolet systems.

The reflection of light on shiny surfaces, such as metal or plastic parts, can be attenuated with the use of polarized filters. In general, this type of filter only allows polarized light to pass through the lens, that is, light reflected by the object in a specific direction, depending on the filter. Another alternative to reduce light reflection is to place the cameras at an angle to the part, in order to avoid the beam of greater intensity reflected by the part (img1).

‍

‍

‍

The angle and intensity of the lighting that, for certain analyses, may be correct, for others it may generate shadows, which reduce the clarity of the image. Once again, changing the camera angle in relation to the part may be a solution, but it's not always possible.

Many vision systems already have a tool called High Dynamic Range (HDR), which consists of a digital photo in which the areas with the highest lighting are darkened and the darkest areas are illuminated (img2).

This system significantly reduces the effect of shadows and the contrast between light and dark in images.

‍

‍

Taking into account all these factors, it is possible to choose the cameras and filters for each application.

But the success of the vision system also depends on the associated software. To carry out different tests on the same parts, more complete equipment with greater processing capacity will be required, which makes them significantly more expensive compared to simpler systems.

‍

Below is an example, taken from a Keyence vision system, that demonstrates testing for patterns. In this type of test, the system is taught a pattern that serves as a basis for searching for identical components. This type of test is widely used for the detection and counting of parts.

‍

We have as an example the Keyence CV-X series controllers, which are very complete systems, with dozens of filters and visual tests, and which also allow text detection and reading and distance measurement. Another example of a very complete controller would be the Omron FH series controller.

It should be noted that these controllers, by themselves, do not do all the work. It is necessary to choose the cameras and filters to use, depending on the situation.

A more affordable solution are the Keyence IV series cameras, such as the IV500. Although it is more limited, it allows you to identify pieces by contours or pixel density.

If the objective is to read barcodes, QR codes or Datamatrix, you can use your own systems, such as the Keyence SR1000.

To have a consistent and reliable vision system, external factors must be taken into account, such as the lighting of the surrounding environment, the color and constitution of the parts, the type of analysis to be performed, and also the hardware and software necessary for each application.

‍