- This article is filed under:
- » View All
What’s Happening in the World of General-Purpose 3D-Based Machine Vision – Part 1
by Nello Zuech, Contributing Editor - AIA Posted 12/01/2004
The market for general-purpose 3D-based machine vision technology is expanding rapidly. By far the largest generic application is metrology. The non-contact data collection ability of machine vision-based implementations makes it possible to make measurements on objects at very high-speeds when compared to contact-based approaches. The computing power available today makes it possible to process the Z-axis data along with the two-dimensional data and map it directly to CAD files fast enough to keep up with production rates in some applications. The result is that 3D-based machine vision is rapidly migrating onto shop floors providing realtime process control.
There are many different approaches to acquiring 3D data. All 3D-based machine vision systems ultimately acquire and operate on image data. Acquisition can be based on collecting the Z-axis data using linear, area or other image sensing techniques, laser radar or other laser scanning techniques and point detectors, or other approach. These systems incorporate the computing power to manage, process and analyze the data acquired, as well as make decisions relating the data to the application without operator intervention. This characterizes what is meant by the term ‘‘3D-based machine vision.’‘
The literature describes numerous approaches: triangulation, structured light, interferometry, moiré interferometry, phase shift interferometry, accordion fringe interferometry, confocal holography, laser ranging, stereo photogrammetry, etc. Some of the approaches are more suitable for small envelope, some for medium envelope and some for large envelope applications. Not all approaches are suitable for all measurement volumes. Some approaches are better suited for finer measurement resolutions – sub-nanometer, while some are better suited for applications requiring micron resolution. It is important that buyers understand their application requirements so that they map the most appropriate approaches to them.
Input for this article was canvassed from virtually all the companies known to offer 3D-based machine vision products. Because of the different perspectives we have written two articles. Part 1 (this article) reflects the input from companies offering a more general-purpose flavor while Part 2 reflects the input from suppliers of more of a solution. For this article the following responded and their input follows as responses to the respective questions posed.
Jim Spare, Vice President of Marketing - Canesta
David Miller, Marketing Manager, IQ Systems
Len Metcalfe, President/CEO – LMI Technologies Inc.
Freddy Paz , Vice President of Marketing and Sales – Optimet
1. What are some specific 3D-based machine vision applications that your company addresses with your 3D machine vision technology?
[Jim Spare, Canesta] ‘‘In the automotive market, our technology addresses the following applications: advanced airbag systems need to classify the occupant of the seat (child/adult/empty/etc.) and identify the location of the head (how far from the airbag) in order to adapt the deployment of the airbag so that children and small adults are not injured or killed while adults seated properly get the full benefit of a robust airbag deployment. Parking assistance systems need increasingly detailed understanding of nearby environments to provide greater accuracy and, eventually, automation, so our 3D vision systems are being considered over today's ultrasonic and radar-based approaches. Legislation has been passed in Europe and Japan that require automakers to deploy pedestrian detection systems to address the high number of pedestrian traffic deaths in crowded city streets there. Vision systems are needed to differentiate pedestrians from other types of obstacles (newspaper blowing across the road or bridge poles that might be off the road but nearby) and 3D is needed to understand distance to apply the appropriate countermeasure.
Similar to these automotive applications, our technology is being applied in the robotics industry in autonomous robots and unmanned vehicles for navigation and obstacle avoidance.
In the security market, 2D vision systems are being applied to automate surveillance activities. However, these systems are often easily fooled by lighting or other environmental conditions, are difficult to install (due to the number of assumptions which must be built-in to the system), are expensive (require high-end PCs and hard drives) and don't scale because they involve sending huge amounts of video around a facility. We are addressing this by 3D vision for distributed intelligent electronic surveillance systems which are more robust in operation through understanding 3D location information, require less processing power (to derive location and remove the effects of shadows), and enable scalable systems because intelligence can be put at the camera, avoiding the need to transfer full frame video around a facility and process it at a central location.’‘
[David Miller – IQ Systems] ‘‘Some specific applications that our 3D technology addresses include precision airfoil measurement, machine tool gauging, automotive stamping gauging, automotive glazing measurement, and mold and die gauging.’‘
[Len Metcalfe – LMI] ‘‘LMI 3D vision applications focus on four primary markets: wood products, automotive, foundry, and industrial sensors. In these markets, our major applications are quality monitoring and process control. LMI provides 3D imaging primarily for embedded OEM applications, developing application-specific solutions for machine builders or system integrators in specific industries.’‘
[Freddy Paz – Optimet] ‘‘Our main activities are applications in the in-process inspection of automotive parts, including laser related applications measuring through the same lens as the laser, QC of turbine blades and plastic molds and reverse engineering of several objects from teeth to jewels.’‘
2. Can you provide a general description of the approach your products employ to arrive at 3D image data?
[Dave] ‘‘IQ Systems' CMMaster 3D Optical Surface Gauge employs unique, proprietary moiré imaging techniques and advanced software analysis to provide high-speed, high-resolution, large, full field measurements of complex surfaces.’‘
[Len] ‘‘LMI’s major expertise is in laser triangulation and related technologies, providing depth measurement combined with 2D information to provide appropriate 3D imaging for a specific application area. All LMI products are integrated sensors, including cameras, lighting, image processing, and communications, integrated into an appropriate package.’‘
[Freddy] ‘‘We use our patent technology called conoscopic holography.’‘
[Jim] ‘‘Canesta has developed and patented new, low-cost, semiconductor-based methods for forming electronic images of nearby objects in three dimensions. Unlike the sensors in digital still and video cameras that see the world as flat images, Canesta's CanestaVision™ sensors can additionally compute the distance from the sensor of every single pixel in the image, in real time. The distance value is generated at each pixel simultaneously and in real-time through a time-of-flight approach where invisible infrared light is emitted from a nearby light source and each pixel calculates the time it takes the light to travel from the object and into the pixel. The time yields the distance.’‘
3. What are critical 3D machine vision system performance criteria for each of the applications that you address?
[Len] ‘‘Each of the markets we address has its own specific requirements, with several common elements. Typically these elements are speed of operation, simplicity of integration, and affordable price. Providing 3D vision packaged for operation in hostile environments is a specialty of LMI. Where 3D vision is used in process control, extremely robust sensing, software and environmental protection are critical for user satisfaction.’‘
[Freddy] ‘‘Customers choose our sensors mainly because of their capability to measure at steep angles and because of the good working range to precision ratio.’‘
[Jim] ‘‘Applications require that the vision systems be low cost, to deploy in high volume and have capabilities similar to human vision in terms of speed, depth resolution and detail.’‘
[Dave] ‘‘CMMaster can produce full field analysis of areas over 3 1/2 feet in diameter in a single imaging cycle. Results are typically completed in less than 15 seconds while maintaining depth resolution on the order of 0.001’‘ or better with an included minimum inspection angle of +/- 45 degrees.’‘
4. What changes have been taking place in the underlying technologies (that are the basis of your 3D machine vision systems) that has resulted in improved performance?
[Jim] ‘‘Availability of low cost infrared light sources (e.g. from DVD and CD players), continued advances in CMOS silicon fabrication technology.’‘
[Dave] ‘‘New Camera technology provides higher data density and improved dynamic range. Affordable, high-performance computers provide faster image manipulation.’‘
[Len] ‘‘Advances in component performance, coupled with lower cost, driven by consumer goods such as digital cameras, have allowed 3D vision sensor performance to significantly improve, without a related increase in cost. The result has been ability to provide more cost effective high performance sensors.’‘
5. Where do you see breakthroughs coming in the specific technologies (that are the basis of your 3D machine vision systems) that will result in further improvements in the near future – next three years?
[Len] ‘‘Further advances in sensor performance, laser and LED light sources coupled with more effective costs, should open new markets for 3D vision.’‘
[Freddy] ‘‘We do plan to increase dramatically the measuring speed.’‘
[Jim] ‘‘Continuance of the above trends: Availability of new LED light sources that are near-infrared (785nm), high power, low cost, and can be modulated at high speed. Plus, additional innovation at Canesta will yield additional performance and cost improvements.’‘
[Dave] ‘‘Recent technology advancements will enable our CMMaster to operate at line speed, enabling 3D gauging to move into online, real-time inspection applications.’‘
6. Are there market changes associated with those applications that you address that are driving the adoption of 3D machine vision?
[Jim] ‘‘Never before has a single-chip 3D vision solution which is low cost in high volume existed. The changes are associated with making people comfortable in doing something completely new.’‘
[Freddy] ‘‘Yes, higher speed than contact measurements.’‘
[Dave] ‘‘Not at the present time, but we believe this will happen as more data becomes available to publicly demonstrate the economic impact of online 3D gauging.’‘
[Len] ‘‘Increasing production speeds, improving productivity and the drive to zero defects in many industries are increasing the opportunities to implement 3D vision. Miniaturization in semiconductor and electronics markets has reached the point where vision is often the only practical solution for measurement and control in manufacturing.’‘
7. How will 3D machine vision systems have to change to meet emerging 3D machine vision applications or to further penetrate their markets more successfully?
[Dave] ‘‘CMMaster is engineered to operate at line speeds, therefore, it is an enabling technology for integration of 3D gauging on manufacturing lines, potentially replacing slow, costly, off-line CMM sampling of precision parts.’‘
[Freddy] ‘‘Easier interface to the different platforms and more available analysis software.’‘
[Jim] ‘‘A low cost approach is needed to bring 3D machine vision from the field of academia and high-end complex and costly industrial applications to an everyday convenience that we all can enjoy in the devices/machines we touch in our daily lives. We believe we at Canesta are pioneering that trend.’‘
[Len] ‘‘Cost and performance of 3D vision is moving in the proper direction for future growth. For applications requiring multiple sensors, simplifying system integration in areas such as synchronization will assist broader implementation.’‘
8. As a supplier of 3D machine vision systems what are some challenges you face in
marketing 3D machine vision systems?
[Dave] ‘‘As a new, enabling technology, acceptance has been slow. Our unique moiré imaging techniques and analysis produce results with much greater accuracy than what was previous thought possible from moiré imaging systems, a stigma that we are fighting to overcome.’‘
[Len] ‘‘Market research in identifying new application areas must be done carefully to insure that the best product is offered, with the broadest implementation opportunity. Offering new technology without understanding the requirements of the user is an easy mistake to make. Attention must be given to understanding the real needs of the user, and providing real benefits in the application.’‘
[Jim] ‘‘There is a lot of market hype about the ability for 2D-based systems to provide a general approach to machine vision. The fact is that 2D systems are great for HIGHLY CONSTRAINED applications such as inspecting a label to see if it is aligned right on a pre-programmed assembly line. But 2D systems are impractical for any application that involves understanding how far away an object is or understanding the location of an object in 3 dimensions. If these constraints cannot be pre-programmed into the system than a 2D solution fails. In most general vision applications, these constraints cannot be pre-programmed into the system, or, in the case they can, doing so often involves a level of expense and time that makes it impractical to do so. The existing community marketing/selling 2D systems is hyping and over promising on the capabilities of 2D approaches for general-purpose applications. This presents a challenge, but we find that the challenge is easily overcome after the customer has tried and failed to install one of these overly hyped 2D-based ‘‘general’‘ image solutions.’‘
9. What are your thoughts on the future of 3D machine vision?
[Len] ‘‘We believe the future is looking good for 3D vision. Product maturity is well accepted, and 3D is proving benefits to users. Today, for simple applications, 3D vision is rapidly moving into the commodity position of many 2D systems such as barcode readers, simple ranging sensors, OCR readers, and presence detection.
Our definition of a volume machine-vision solution is what I call a triple 1, a ‘‘111 solution’‘—one integrated mechanical package with one interface cable and a one-page user's manual. When the user's manual reaches only a few pages then you know that a distribution channel can handle the product with little support, and it can become a commodity.’‘
[Freddy] ‘‘As speed and precision increase and as large groups develop analysis software, more and more end users will try to go to non contact 3D.’‘
[Jim] ‘‘Microprocessors and computing technology is now sufficiently integrated into our lives -- computers, cell phones, PDAs, automobile electronics, videogames, robot vacuums, etc. The problem is that their interface to the world is in most cases very crude. The next wave is making all of these devices play with their surroundings -- including us humans -- in ways that dramatically increase their functionality and convenience. 3D machine vision is one of the best ways to solve this problem. A low cost implementation like CanestaVision is necessary to fulfill this promise and make these benefits pervasive.’‘
[Dave] ‘‘On-line 3D gauging could radically improve yields in precision manufacturing applications.’‘
10. What advice would you give to a company investigating the purchase of a 3D machine vision system?
[Freddy] ‘‘Look for machines with upgrade capability since the sensing technology will improve all the time.’‘
[Len] ‘‘The first step is identifying the needs and developing a performance specification that is realistic and effective. Involving a 3D vision supplier early in project definition phase can assist in specification development resulting in a cost effective solution.
Selection of the vision supplier is a critical step – look at the capabilities of the supplier, including depth of experience in vision technology, experience in related applications, knowledge of the true total cost of integration and ability to respond quickly to specific needs.’‘
[Jim] ‘‘Think about your goals and select a technology and platform in which to invest which is most consistent with your goals.’‘
[Dave] ‘‘Investigate new large, full field, high-speed systems as an alternative to expensive, off-line touch probe CMMs. Camera-based systems can generate more data points significantly faster than point-to -point systems. For example, our CMMaster configured for a 12’‘ diameter field of view can generate nearly 2,400 data points per square inch in cycle times of less than 5 seconds in typical applications.’‘
Watch for Part 2 in our Mid-December E-Newsletter!
There are currently no comments for this article.
Leave a Comment:
All fields are required, but only your name and comment will be visible (email addresses are kept confidential). Comments are moderated and will not appear immediately. Please no link dropping, no keywords or domains as names; do not spam, and please do not advertise.