MACHINE vision has come a long way in the last thirty years. In the early days of computing it was difficult to process even moderately large sets of image data. But by the late 1970s technology had advanced to the point where more focused studies could occur and technological spins off for industrial purposes began to appear.
Early machine vision systems often consisted of bulky cameras and separate controllers using processors, such as the Motorola 68000. Even in these early days, equipment developers realised that the possibilities for applying this technology were almost limitless and very quickly appreciated that almost any material handling or logistic process that involved object inspection would suit a vision system.
From rudimentary beginnings, machine vision suppliers have developed machine vision systems into ‘smart cameras’, integrated vision systems which, as well as possessing image capture capability, are able to extract information from captured images and make decisions that are used in wider machine automation systems.
Modern smart cameras are self-contained, standalone units with built-in image sensors, communication interfaces, and industry standard outputs for connection to PLCs where processing of the image takes place. Most units are now no larger than an ordinary industrial or surveillance camera, but rival PCs in terms of their processing capabilities.
Typical applications have grown to include detection of faults and flaws, non-contact measurement, part sorting, code reading, continuous flow dimensional gauging, and position detection and rotation.
Industries where these applications are commonly used are automotive, electronics, food and beverage, logistics, manufacturing, packaging, pharmaceuticals, steel and wood.
Label checking on soap packages
Commonly used techniques used include edge detection, template matching, thresholding, segmentation, blob extraction, pattern recognition, barcode reading, and optical character recognition.
However, most smart cameras remain expensive and require extensive integration by developers to be effective. Custom programs often need to be developed for applications or off the shelf software from third parties must be adapted.
This is why machine vision system suppliers have more recently developed cost effective and user friendly 2D vision systems – now commonly known as vision sensors.
Vision sensors are compact, practical, and have reliable 2D imaging with integrated lighting, image evaluation and Ethernet interfacing. They are designed for trouble-free and fast solving of a diverse range of part inspection applications, are capable of inspecting parts in any orientation and position and have a predictable response time, making them ideal for packaging and small part assembly.
A vision sensor’s robust design is suited for industrial applications and its intelligent and fast evaluation algorithms enable it to keep up with high processing speeds.
In large production plants that have the need to solve a variety of inspection tasks with different set-ups, it is important to have a flexible solution as well as a complete centralised control of the production. This is normally not possible to achieve with a vision sensor. Therefore, customers may have to choose from more complex and expensive smart camera solutions.
However some newer vision sensors offer part inspection at smart camera level of performance. This means high performance inspection, flexible setup and complete production control support – without increasing the complexity and cost of the solution.
More importantly they are able to offer high resolution VGA imaging (640×480 pixels), which provides improved image quality and enables inspection at higher accuracy or on wider areas without compromising speed. Powerful inspection tool sets provide for straightforward, easy-to-configure problem solving abilities, enabling easy solutions for tough targets and a common unit across a number of solutions.
They come in the flexible housing designs for easy lens exchange, and to make them even more adaptable, a number of easy-to-set-up illumination accessories are available. This concept combines built-in illumination with different front windows, resulting in red, green, blue or diffused white illumination – without any extra wiring – illumination has never been easier!
Pre-defined packages for production control within user interfaces are provided and emulator tools provide easy methods of fine tuning and testing of solutions without disturbing production. Ideally, vision sensors should also incorporate image logs for analysing production problems.
Always remember that when choosing and implementing machine vision technology, a simple solution that works is a preferable solution.
Ask yourself the following questions –
What outcome is required? Often the sole drivers of vision projects are the need to increase production speed but it can just as easily be a desire to reduce labour costs, improve quality, and increase production yield.
Is vision needed to do the job? Sometimes a combination of standard sensors can achieve the same outcome albeit with increased complexity.
Is the application solvable? Image quality is important and complex problems can lead to lengthy periods spent engineering suitable illumination with subsequent delays in commissioning.
Which technology should be used? Define your task. Speak to experts who have the experience and knowledge of your application, and who can advise whether standard hardware or a customised solution is applicable.
What does a typical vision project look like? Ask for typical application information and make sure that whomever you engage is fully committed to guiding you with your project.
What problems will be encountered along the way? Often there are heat, vibration, dust and humidity concerns. Mounting space, cycle times and illumination must be taken into account.
Luckily Sick’s Inspector family of vision sensors are perfectly suited to many machine vision applications. With its increased performance and advanced features, the Inspector I40 in particular can perform part inspection tasks just like a smart camera.
The I40 also supports PLC control over Ethernet, with support for reading out detailed results and control of the configuration set-up. Moreover, it supports storage of images to FTP, an extended arm to the image log that makes it unrestricted in memory capacity. Thus, the I40 is the ultimate choice for complete production control.