Machine vision technology may have cut its teeth in the electronics and semiconductor industries, but increased ease of use and reduced cost is now driving rapid expansion into all industries and a wide range of new areas. As manufacturers across the board raise the bar for process efficiency and product quality, more and more factories are reducing scrap costs and inventory problems by integrating machine vision onto their production lines. Machine vision is extensively used for:
Gauging—to measure the part or examine its critical dimensions
Inspection—to indicate if a part is good or bad based on its physical characteristics
Guidance—to accurately locate or place a part
Identification—to determine whether the right part is present by inspecting its physical characteristics or reading a marked code
Though machine vision is well established in the medical, pharmaceutical, aerospace, automotive, and consumer products industries, awareness of this relatively new technology continues to grow into other areas such as general production and appliance manufacturing. Whether you’re brand new to machine vision, or an experienced user, this article can help you understand what machine vision is, and where it might make sense to use it in your production process.
A machine vision system is essentially a computer equipped to see and spot defects and other problems in manufactured parts or assemblies. Vision systems look for manufacturing flaws using a combination of microprocessor technology and image analysis software to interpret video images and generate information about them. The systems can then communicate this information to other equipment, such as PLCs (Programmable Logic Controllers) or robotic arms that can remove bad parts from the manufacturing line.
Hardware, software, and an application development environment make up the core of any vision system. Vision hardware includes a camera that captures an image of the item to be inspected, lighting to enhance the contrast of features of interest, and optics, which accurately represent the image to the camera by minimizing distortion and loss of resolution. This hardware, works with a processor, or “vision engine” to capture, digitize, and display images for analysis to generate answers, such as whether a part is defective.
Vision software “tools” are the backbone of the vision engine. By comparing specific features of interest within the image to stored data that comprises a standard, these vision tools perform image processing and/or analysis of the captured image. There is a wide assortment of vision tools available for performing many different types of inspection operations that enable vision systems to make decisions about a part’s quality, location, size, and identity. Vision tool selection and capability depends on the type of vision system platform used.
PC-based system or sensor based?
Generally, today’s vision systems are divided into two groups: PC-based and sensor-based. The key differentiators include development environment, capability, architecture, and cost. The development environment allows users to “build” (set up, and program or configure) vision applications to meet specific needs. While PC-based systems have a programmable environment, sensor-based systems generally provide a configurable environment that’s easier to use.
PC-based machine vision systems offer the most capable vision tools, and provide the fastest performance because they rely on the latest CPU architectures. Performance increases with each boost in PC processor speed, and as a result, they are generally used for more complex or mathematically intensive applications. However, applying PC-based systems requires more vision expertise and knowledge of low-level programming languages such as VisualBasic.
In contrast, vision sensors generally require no programming, and provide more user-friendly interfaces. For purpose of this article, vision sensors are systems that are self-contained and don’t require the use of a PC, VME, PCI, or similar architecture to run vision tools. While low cost and ease of deployment remain the key attributes of sensor-based platforms, over the last several years vision sensors have become increasingly sophisticated while simultaneously the cost of PC-based systems has come down.
As this gap between PC-based and sensor-based platforms continues to narrow, new users more often than not initiate feasibility studies using a vision sensor. Though ultimately, application complexity and other variables will dictate the final hardware and software requirements, vision sensors offer a price tag that makes the investment more easily cost-justifiable. In addition, vision sensors are stand-alone systems that are easily integrated with any machine to provide single-point inspections with dedicated processing, and most vision sensors offer built-in Ethernet communications for factory-wide networkability.
Applications for machine vision in appliance manufacturing are far too numerous to cover completely in this article. However, any place in the process where significant value is added, presents a potential opportunity for vision. Some users may require a random inspection during a particularly troublesome production step to eliminate costly rework or equipment damage. Automated inspection can verify assembly prior to soldering, stamping, or riveting. It can also be used for tending press or moulding operations, preventing damage to costly moulds and dies by ensuring that the part is properly ejected before they close again for the next cycle.
For 100% verification at the end-of-the-line, vision is commonly used to spot defects in the application of graphics or labels, or to verify finish colour and quality defects such as embosser pick off, furnace debris, paint misses, oven scrapes, blisters, or water staining. A vision system can also make sure that packaging materials are properly positioned in a shipping container before it is closed.
Where can vision benefit production?
Whether you choose random inspection, end-of-the-line inspection, or multiple in-line inspection systems to scrutinize several processes is highly dependent on company goals. Higher quality, increased productivity, production flexibility, tighter process control, lower production costs, scrap rate reduction, and inventory control are some common goals that machine vision can help achieve.
Once company goals are established, the next thing a new machine vision user will want to do is assess a range of potential applications. Keep in mind that with end-of-the-line only inspection, since defects are not detected at the point of occurrence and the appliance is fully assembled, rework can be more time consuming and costly. Housings and any other associated parts such as gaskets, rivets, and shock pads may have to be scrapped.
That’s one reason why, over the past several years, there has been an increasing trend in general manufacturing toward the use of machine vision technology for 100% automated inspection at multiple points in the production process. For example, mobile-phone manufacturers and their suppliers use machine vision to perform a number of operations including inspecting LCD displays for single pixel defects, verifying the correct placement of keypads, gaskets, and other components. Other applications include automatically reading bar code labels or direct-marked parts for component tracking, assembly verification, and inspecting product enclosures for scratches, dents, and other defects.
Because of the affordable nature of many vision sensors available today and improvements in processor technology and networking, manufacturers can justify the cost of distributing vision throughout the process to catch defects at the source. A company that had only used vision at the end of the line to check display functionality may now be able to add dedicated vision sensors upstream to check the placement of gaskets, buttons, and printed circuit boards as they are applied. That way, if components are not positioned precisely, corrective action can be taken on the spot. In the packaging industry, cameras can be positioned at each stage of the production process, from filling to case packaging, see figure 1.
When shopping around for a vision sensor, it is important to keep in mind that while a sensor may perform with a high degree of accuracy and reliability in laboratory-like conditions, it may ultimately fail on your production line. Thus, an extensive, thorough application analysis and product evaluation is essential. It’s a good idea to make sure that the representative you are working with is a full-time machine vision specialist, so that he or she is able to effectively evaluate your application and help you anticipate any unforeseen issues that may arise. Finally, it is also wise to work with a company that offers a wide range of global product support and educational services to help ensure that your investment in machine vision is a successful one.
Evaluating vision sensor part location tools
Part location tools, available with virtually all vision sensors, are software tools used to find parts within the vision camera’s field of view. This is typically the first step in any vision application, and the one that usually determines whether or not the application succeeds or fails.
While it sounds simple enough, locating parts in today's production environments can be extremely challenging for vision sensors. This is because many variable conditions exist that can alter the way a part appears to a vision sensor, which is trained to recognize parts based on a reference or "model" image of the part. Variable conditions include:
Part rotation
Changes in optical scale
Inconsistent lighting conditions
Normal variations in part appearance
Written by Cliff Fitzgerald, Senior Manager, Worldwide Education Services for Cognex and
John Lewis, a staff writer for Cognex