Machine vision training for industrial vision systems

Making a machine vision system work reliably requires more than configuring a software tool. In practice, almost every good inspection starts with the image itself. The product must be sharp, large enough and visible with sufficient contrast. Only then does it make sense to configure measurement tools, barcode reading, OCR or defect detection.

Machine Vision Academy offers practical machine vision training for companies that want to better understand, configure and maintain vision systems themselves. The training courses are intended for operators, technicians, engineers, maintenance teams, quality departments and machine builders who work with vision systems or want to start using them.

During the training, we do not look at camera, lens, lighting and software as separate components. A camera with sufficient resolution does not automatically result in a good inspection if the lens does not image sharply enough, the lighting does not make the defect visible or the trigger does not occur at the right moment. That is why machine vision is always treated as a complete system.

The training courses can be given with Datalogic IMPACT and EVT EyeVision. Datalogic IMPACT is well suited to smart cameras, vision processors and compact industrial vision solutions. EVT EyeVision is interesting when more flexibility is required in software, camera interface or PC based system architecture. Which route fits best depends on the application, the desired integration and the way the vision system has to communicate with the machine or PLC.

General machine vision training

The general machine vision training covers inspection tasks that are common in many industrial applications. Examples include presence and absence checks, position checks, dimensional checks, barcode reading, DMC code reading, OCR, contour checks and simple pass or fail inspections.

These applications often seem simple, but they only become reliable when the product is presented in a repeatable way and the image is sufficiently stable. A barcode can sometimes still be read with varying image quality, but OCR or dimensional inspection is much more sensitive to contrast, sharpness, lens distortion and product position.

During the training, participants build an inspection step by step. First, we determine which product feature needs to be checked. Then we look at how this feature can be made visible using camera, lens and lighting. After that, the correct tools are configured in Datalogic IMPACT or EVT EyeVision, tolerances are selected and reject images are assessed.

The general training can be given in a classroom setting as a practical two or three day training course. Training on site is also possible. This is especially useful when own products, reject images or machine settings can be used, because the training then directly connects to the daily production environment.

More information about the training offer can be found on the Machine Vision Academy page.

Application oriented machine vision training

In addition to general training courses, Machine Vision Academy offers application oriented training. These courses do not start from a standard course, but from a specific inspection task. This can be label inspection, but also barcode inspection, DMC code reading, OCR on packaging, dimensional inspection, position inspection, defect detection, assembly inspection, connector inspection, metal stamping inspection or end of line inspection.

In label inspection, for example, the checks may include presence and position of the label, barcode reading, DMC code verification, OCR of an expiry date or batch code and checking the correct promotion code. Each additional check places higher demands on camera resolution, lens, lighting and software settings. What is sufficient for a simple presence check is often not enough for OCR or accurate positioning.

In metal stamping inspection, the focus is on fast inspection of stamped metal parts. Examples include contour checks, position deviations, edge quality, scratches and burrs. Because metal surfaces reflect and parts are often produced at high speed, camera, lens, lighting, trigger and software must be properly matched. A small change in lighting angle can determine whether a burr becomes visible or disappears in reflection.

The goal of application oriented training is that participants learn how to technically approach their own inspection problem. Do not start with the software, but first determine which defect, feature or difference needs to be made visible.

Training with Datalogic IMPACT

Datalogic IMPACT is used in the training for industrial vision applications with smart cameras, vision processors and Datalogic vision systems. The training can focus on image settings, inspection structure, tool selection, calibration, result processing and communication with a PLC or machine control system.

Datalogic is especially interesting when an application fits well with a robust smart camera or vision processor route. This may be the case for compact inspection stations, barcode or DMC inspection, simple measurement tasks, product inspection or applications where integration into the machine is more important than maximum programming freedom.

During the training, we look not only at the software environment, but also at the system choice. A smart camera may be sufficient when the inspection is clearly defined and can be processed locally. With multiple cameras, higher data streams or more complex communication, a vision processor or another architecture may be more logical. Training helps to technically substantiate that choice before components are selected.

More information about this software can be found on our page about Datalogic IMPACT machine vision software. For the broader system route, the page about Datalogic Impact vision systems is also relevant.

Training with EVT EyeVision

EVT EyeVision is used when flexibility in software, camera interface or system architecture is important. The training can focus on 2D inspection, 3D inspection, camera settings, tool structure, communication with external systems or a combination of these.

EVT EyeVision is well suited to machine builders and technical teams that want to keep control over the inspection structure themselves. The software can be interesting when different cameras, industrial PCs or more complex inspection tasks need to be combined.

In the training, participants learn how to build an inspection logically, use test images, assess reject images and reliably send results to the machine. The relationship between software settings and image quality is always considered. If a measurement result varies, the cause is not always in the tool setting. Lens, lighting, product position and trigger timing can be just as decisive.

More information about this software can be found on our page about EVT EyeVision machine vision software.

Camera, lens and lighting as the basis of the training

An important part of the training is assessing camera, lens and lighting. A camera with more megapixels is not automatically better. If the lens does not image the required details sharply, the lighting is too weak or the product moves during the exposure, a higher resolution mainly produces larger images without making the inspection more reliable.

The lens determines field of view, working distance, sharpness, distortion and how much detail actually reaches the sensor. That is why lens selection is always linked to sensor size, pixel size, camera resolution and the smallest detail that needs to be measured or recognized.

The lighting determines whether the relevant difference becomes visible. A backlight is suitable for contour and dimensional checks, while dome light or diffuse lighting may work better on reflective surfaces. Dark field or low angle lighting can make small scratches, edges or surface structure visible. The right lighting makes the software simpler, because the contrast is already present in the image.

For component selection, participants can further explore our pages about machine vision cameras, machine vision lenses and machine vision lighting.

PLC communication and integration into the production line

A vision system must not only recognize pass or fail. The result must also be sent at the right moment to the machine, PLC, robot, end of line controller or logistics system. That is why the training also covers triggering, output signals, processing time, reject timing and communication with the line.

In a production line, timing is often just as important as image quality. If the trigger comes too early or too late, the wrong product is inspected. If the lighting is not synchronized with the exposure, brightness varies. If software processing takes longer than the cycle time, the result arrives too late.

Training makes it clear that vision integration is always a combination of optics, electronics, software and machine control. This is especially important for companies that want to modify or maintain vision systems themselves.

Training on site

Training on site is suitable when the team wants to work with its own products, own reject images or an existing machine setup. This makes the training more practical, because participants immediately see how camera, lens, lighting and software respond to their own parts and production variation.

Training on site can also help to better understand existing vision systems. Systems are often installed at some point, but the team does not always know exactly why certain settings were chosen. By reviewing image acquisition, tools, tolerances and machine communication together, the team gains more control over maintenance, troubleshooting and future changes.

Training on site is especially valuable when the application strongly depends on product handling, ambient light, mechanical mounting or line speed. In such cases, a standard training setup does not show all limitations of the real production environment.

Training as part of self integration

Not every company wants to outsource a complete vision project. Many technical teams want to keep control over their machine, settings and future changes themselves, but do not want to risk choosing the wrong components or the wrong software architecture. Training can then become part of a guided self integration route.

In such a route, the team learns how a vision system is technically built, while Machine Vision Shop can support the selection of camera, lens, lighting and software. This reduces the risk during the first implementation and makes it easier to manage the system independently later on.

Training is then not separate from the webshop. It forms a bridge between product selection and reliable application. A customer then does not only buy a camera, lens, lighting or software licence, but better understands why these components fit together for the inspection task.

Which training fits your application?

Not every machine vision training course needs to be the same. The right training depends on the application, the vision system used, the knowledge level of the team and the way the inspection is integrated into the machine or line.

A team that is working with vision for the first time usually benefits from general training in which camera, lens, lighting and basic software are covered. A machine builder that already works with Datalogic IMPACT or EVT EyeVision often benefits more from application oriented training with own images or a specific inspection task. A company that wants to assess a new application can combine training with consultancy or an application review.

Would you like to know which training fits your situation? Contact us for advice about general machine vision training, training on site or application oriented training with Datalogic IMPACT or EVT EyeVision.

Also view the overview of Machine Vision Academy for training courses and practical knowledge about machine vision.