Snapshots

Vision peels apple skins; Wood knots unraveled by vision; 3-D system inspects seams; and MORE…

Th 0509vsd Snapshot1

Vision peels apple skins

The tree-fruit industry in the USA is worth about $2 billion a year but must continually enhance its competitive position and profitability. Renfu Lu, an agricultural engineer at the USDA Agricultural Research Service Sugarbeet and Bean Research Unit, partnered with researchers at Michigan State University (both in East Lansing, MI, USA; www.ars.usda.gov) to develop a nondestructive method for grading and sorting fruit such as apples, peaches, and cherries based on the internal qualities of firmness and sugar content.

Th 0509vsd Snapshot1
Click here to enlarge image

Currently, steel probes are used to test fruit samples by punching a hole in the fruit, making it unmarketable. “Our system tests every single fruit, and it can all be sold,” says Lu. The researchers are now sampling apples with a multispectral imaging system that uses four different laser beams scattered by the fruit’s core. An imaging spectrograph, a digital camera, and a computer analyze the laser light absorbed by the apples, which indicates sweetness. The light reflected from fruit tissue indicates firmness.

The researchers are working on the ability to sense acidity and to increase scanning speed to match that of commercial apple conveyors, which is about 10 fruit per second. The system could be merged with existing industry sensors that nondestructively assess superficial visual traits, including size, color, and bruising.

Wood knots unraveled by vision

In the labor-intensive production of plywood, a veneer ribbon of wood is clipped into 4 × 8-ft sheets and passed through a dryer. The sheets are inspected and sorted according to grade and then inserted into a press, layered, and combined with glue, pressure, and steam heat to make a plywood panel. The sheets pass along a conveyor belt at a rate of 1/s, and the task of looking for knots, holes, dirt, gouges, and other defects is challenging for workers. In addition, smoke and dust raise health concerns.

Th 0509vsd Snapshot2
Click here to enlarge image

Ventek (Eugene, OR, USA; www.ventek-inc.com) has developed a machine-vision-based inspection system, the GSC2000, that scans the sheets and calculates a grade with an accuracy rate of 95% or higher. The GSC2000 features a Basler (Ahrensburg, Germany; www.baslerweb.com) L301KC color camera and a Matrox Imaging (Dorval, QC, Canada; www.matrox.com/imaging) Meteor-II/Camera Link frame grabber in a 3-GHz Xeon PC (with hyperthreading technology). Ventek’s True Color Lighting illuminates the inspection area. Image capture, processing, and measurements are all performed by Matrox MIL version 7.

The system relies on Ventek’s Neural Network Classifier software, which must be trained with images of defects. The sheets are assigned a grade or rejected; lower-grade sheets can be used as the central layer of the plywood. More than 40 GSC2000 systems have been sold to major plywood producers, including Boise Cascade, Georgia-Pacific, Martco Plywood, and Weyerhaeuser.

3-D system inspects seams

Weld seams, seals, and rivets are often complex and difficult to inspect with satisfactory assurance during manufacturing processes. In applications such as assembling airplane-cockpit support frames, seam inspection is critical for detecting holes, surface voids and pores, missing seams, and cracks within joints. The cockpit support frame consists of two aluminium half-shells that are welded together over a length of 1300 m in an electron-beam welding facility.

Th 0509vsd Snapshot3
Click here to enlarge image

A 3-D seam-inspection system such as the VIRO developed by Vitronic (Wiesbaden, Germany; www.vitronic.com) can ensure that the necessary quality inspection is performed. The sensor unit on the VIRO contains light-stripe sensors with scan rates of 500 to 2000 profile stripes per second. The robot moves the frame past the sensor and interfaces with the robot controller. A good/no-good signal is transmitted from the sensor to a PLC, which can send the part back for rework and to a PC that tracks the type and frequency of defects and documents the process.

European market to grow

A recent study by Frost & Sullivan (London, UK; www.frost.com) says that the European industrial vision-systems market grew 6.8% from 2001 to 2004 and generated revenue of $500 million in 2004. Sustained support from traditional industrial sectors, newer application areas, and the development of higher-performance vision systems should lead the market to reach $948.44 million by 2011, with a CAGR of 9.6%

The study divided the market into product segments: recognition systems, guidance systems, metrology systems, shape-conformity systems, nonweb-flaw systems, and web-based inspection systems. It found that 1-D/2-D metrology systems represented 22% of the 2004 market, but 3-D metrology systems will erode this market share. Web inspection and nonweb-flaw systems will also gain more acceptance.

Germany had an estimated 36% of European revenues in 2004, while the UK and France each had 13%. The study says that Eastern Europe represents an unexploited region with great potential for growth. However, without government investment, the revenue share in 2011 will be only 3%.

The largest end-user sector is the electrical and electronics industry, which accounted for 28% of revenues in 2004 and is expected to fuel future growth of vision systems. The automotive sector, representing 19% of the 2004 market, is forecast to grow, as well.

More in Cameras & Accessories