- This article is filed under:
- » View All
3D, Advanced Technologies Make Food Inspection Palatable. . . .by Winn Hardin, Contributing Editor
by Winn Hardin, Contributing Editor - AIA Posted 06/08/2006
Inspecting and analyzing organic products, like food, may seem like a repetitive task, but in reality, the job is always changing. Organic products, from lumber to chicken, vary from unit to unit; nature sees to that, and it's this variation that has challenged automated sorting, inspection, analysis, and process control in the food processing industry.
To meet this challenge, companies are packaging advanced machine vision technologies such as 3D and color into easy-to-use systems that OEMs and end users can treat as commodity items – a term that once was looked down on by the vision industry but has come to mean an integration success and chance for OEM product penetration.
Move from 2D to 3D
The food processing industry has used 2D vision systems for sorting fruits and vegetables at high line speeds for many years. Food sorting systems are still among the most powerful vision systems, moving from VME-based systems with high-bandwidth backplanes (compared to ISA and PCI buses), to the latest generation high-bandwidth PC-host systems. New applications are emerging, however, that require more data at high speeds.
‘‘What's the drive in food inspection?’‘ asks SICK, Inc.'s (Minneapolis, Minnesota) Karl Gunnarsson Business Development Manager for Vision, ‘‘mainly it’s the need to see things you normally can't. It could be a surface feature that's hard to pick up in 2D or volume or height information. 2D systems would have to work with shadows, multiple cameras, or special lighting arrangements, and it's still difficult to get an accurate 3D quality measurement using these methods. For instance, a 2D system looking at a piece of food on a conveyor for height measurements might see a little slice of food sticking up, or get a bad angle and you get a bad reading. A cookie that's 1/16th of an inch too thick is no big deal, until you have a bunch of them and try to fit them in a package. You end up breaking the product and that's not good. (Mostly due to clean up efforts/production stop/slowdown not necessarily product lost.)’‘
In response to the need for precise volumetric measurements – sometimes referred to as portioning – companies are combining 3D laser measurement systems with 2D visual inspection systems to optimize food production processes and meet increasingly more stringent visual requirements for packaged foods. These systems typically use area array CCDs to capture the reflected laser line. Image processing algorithms use triangulation to determine the Z (height) axis based on the displacement of the laser line as a product moves underneath the laser line. When combined with X and Y information – also captured by the laser inspection system – the vision system can determine shape and volume of an object.
‘‘There hasn't been any fundamental changes to these systems – other than moving from a scanning laser dot to laser line projector and making the systems easier to use,’‘ Gunnarsson explains. ‘‘However, there is more education of how to use 3D systems [in food processing.]’‘
While the basic methods of laser triangulation may not have changed significantly in recent years, vision suppliers are making inroads into OEM dominated food processing markets by offering highly-integrated vision systems that offer multiple features from a single 'box'.
‘‘Machine builders are not experts in machine vision, especially in the protein portioning market, which has really only used vision in large numbers during the past 3 years,’‘ explains Dan Howe, Business Development Manager, at 3D vision specialist LMI Technologies (Royal Oak, Michigan). ‘‘People have had to go outside their core expertise. While it's not that difficult to use a line-generating lens, a laser pen and a camera, it is hard to make a 3D laser triangulation system that is accurate and fast. We're also helping OEMs deal with obsolescence and other operational and maintenance issues. We're offering them solutions that they can just bolt on and go.’‘
LMI is offering a new quality-assurance sensor for the food processing industry that uses laser line projection in addition to standard 2D image processing systems to automatically measure product length, width, height, area, volume and a new twist for the portioning application space – color vision.
‘‘We're concentrating on protein portioning and quality-assurance at this time because as far as optimization and yield, it has the highest commodity costs, although we're certainly looking at traditional food processing applications in baking, fruit and vegetable sorting, etc.,’‘ Howe says.
Shape and volume measurements are critical to the protein portioning and quality assurance niches of the food processing industry. The largest customers, such as fast-food chains McDonalds and Wendy's, have strict requirements on the shape and size of chicken and beef patties. ‘‘They don't want a bun without meat, for instance,’‘ Howe says. ‘‘Where the optimization is most effective is upstream in the cutting process. Large portioning machine manufacturers use vision already, but it's mostly a home-grown variety. With our systems, they don't have to recalibrate the machine every time a component goes down. We offer plug-and-play vision solutions. And if someone wants to add color, they don't have to task an engineer to study color. We already have that capability at our disposal.’‘
LMI's food quality assurance sensor can measure non-singulated protein products with millimeter resolution at speeds of 100 feet per minute. ‘‘There are other solutions for this problem on the market, but they all require singulation of product. We can measure multiple products in the same field of view without need multiple sensors. We estimate this improves our throughput by 4x over existing solutions,’‘ Howe says.
Just as 3D systems provide food processing machines with additional data that helps them be more accurate and efficient, Howe says color systems can also help improve the contrast of 2D food inspection and is being more readily adopted by the food industry as a result. ‘‘There may be other uses for 'color,' such as fluorescence or other methods that could provide redundant contamination checks to existing x-ray and other safety inspection systems for the food industry,’‘ Howe predicts.
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.