Machine vision has slowly but steadily been applied to a variety of manufacturing challenges over the last two decades, all with the goal of improving manufacturing quality and productivity. Semiconductor and electronics manufacturers were among the first to adopt machine vision, and they now account for roughly half of all machine vision applications on the factory floor. Machine vision systems happen to be now in use in pharmaceuticals, food processing, plastics, wood and paper, metal fabrication, and added industries, and their recognition is growing faster.
System constituents at a glance
The following are collective components:
Cameras — CCD cameras are shrinking in size, weight, and cost. The new dual output cameras produce images twice as fast as previous models, resulting in sharper and more accurate images. A new generation of CCD colour cameras expands machine vision capabilities by allowing systems to better detect and distinguish between objects, remove backgrounds, and perform spectral analysis.
Frame grabbers — Video or still images are converted to digital data using these specialised A/D converters. The majority of frame grabbers are printed circuit boards that are compatible with the most common bus structures, such as PCI, PC-104, ISA, VME, and CompactPCI. Frame grabbers today are more stable and accurate than previous generations, and some can even perform image processing and enhancement on the fly using digital signal processing techniques.
PCs — The PC has had a significant impact on the use of machine vision in manufacturing applications since the introduction of the PCI bus. Up until then, personal computers couldn’t gather data at a fast enough rate to keep up with machine vision’s high I/O demands, which included data transfer rates of 20 MB/second or higher. Instead, the VME bus, a specialised architecture for data acquisition and process control with 40 MB/second bus speeds, became a development standard. With 132 MB/second PCI bus transfer speeds and >100 MHz Pentium microprocessors, today’s PCs can handle machine vision’s demands. PCs are now commonly found embedded in factory equipment. The distributed intelligence enabled by PC technology has made a significant contribution to the speed and efficiency of factory automation.
Software — Machine vision has become a more user-friendly tool thanks to graphical user interfaces and libraries of high-level software modules that run in standard environments like Windows. Leading-edge software vendors have begun to offer object-oriented application development tools, which will help developers develop applications even faster.
New technologies — High-speed serial data ports such as the Universal Serial Bus and FireWire (IEEE 1394) will increase the overall capability of machine vision companies by speeding data transfer and information throughput. PC and peripheral vendors have already adopted USB as an industry standard, making it easier to connect digital cameras to powerful embedded PCs. However, real-time video rates will necessitate the use of the faster Fire Wire.
Manufacturers must determine the best path to take in configuring a system once they have determined that machine vision can be an effective tool for their application. Larger companies with in-house engineering staffs may pursue their own solution, assembling components from various vendors or even employing cutting-edge technology. The in-house approach, however, is largely impractical due to a steep learning curve, a lack of industry standards, and time-to-market pressures. The vision system that is supposed to add value to a product can quickly become a major time, energy, and resource drain. To solve the problem, you’ll need to enlist the help of a professional.
