IEN: What innovations are in store for users?
Voosen: In the areas of:
- Color -- The biggest challenge with color image processing is the vast amount of information that must be acquired and processed by a PC or intelligent vision system. New technologies, like PCI-Express, and faster processors will allow more users to take advantage of color inspection.
- Intelligent Sensing -- Right now, intelligent sensors (or smart cameras) are popular in simple, low-cost, single camera inspections. Recently, intelligent vision systems have blurred the lines between the simplicity of smart cameras and the power of a PC. These Intelligent Vision Systems (like the NI Compact Vision System) often support multiple cameras, run on much faster processors, and use more complex vision algorithms.
- High-end Edge Detection -- To locate an edge most accurately, machine vision software packages will often interpolate single pixels in subpixels. These algorithms can provide edge locations down to the hundredth of a pixel. However, these values are just approximations; the most accurate measurements come from systems with high-resolution cameras and excellent lighting to provide contrast.
- Data Acquisition -- NI is the world leader in PC-based data acquisition. Recently, NI released DAQmx, a more simple and powerful way to program data acquisition systems. These DAQ systems integrate naturally with NI image acquisition hardware and software for tightly integrated image and data processing applications.
- Software -- Machine vision software can be measured in two ways: algorithm power and ease of programming. The most powerful vision functions are useless if no one can figure out how to implement them. National Instruments Vision tools are among the most powerful in the world. What puts our software above other libraries out there is how easy it is to program. We focus on end users, so our programming libraries are easy to understand, and our configuration software is quite intuitive.
- Ruggedized -- Depending on the application, having rugged hardware can be a primary concern. Industrial PCs with a frame grabber will cover most applications. However, if an environment is wet, dirty, or extremely clean, special hardware or enclosures are necessary. The Compact Vision System is designed with such environments in mind. It contains no vents, fan, or moving parts of any kind. It can also withstand temperatures up to 55ºC and 50 g of shock.
IEN: Can obstacles to wireless applications be overcome?
Voosen: Obstacles in wireless applications will certainly be overcome in the near future. Whether engineers are using 802.11 or Bluetooth, the same three issues apply: bandwidth, security, and distance. Oftentimes, images are far too large to transmit in real-time. To get around this, users need to make sure their software can compress images before streaming them over wireless. Security and distance obstacles are common in most wireless applications, not just machine vision, so they will most likely be addressed by other application areas first.
IEN: How are sensing and machine vision being used in collaborative manufacturing? The extended enterprise (logistics, etc.)?
Voosen: Machine vision is a relatively small part in larger collaborative manufacturing systems. Usually inspection data (and sometimes images) are logged to servers to allow for long-term quality trending. This data is often part of a total test architecture that includes in-circuit and functional test results.