Machine vision applications -  Successful examples from many industries

MACHINE VISION APPLICATIONS

Successful examples from many industries

Machine vision system opimises lumber processing in changing daylight conditions

March 2016

A machine vision system mounted onto a grapple loader and a Smart Vision Lights outdoor illumination enable reliable and efficient measurements of tree trunks before they are further processed into boards.

Throughout the world, businesses are highly competitive with every organisation maximising efficiencies and the quality of their processes. This is true of automotive, packaging, food, and even raw materials industries. Lumber processing is no different. Accurately knowing the quantity of wood produced and provided by timber purveyors is critical. That is why lumber producer Pike Lumber contacted an expert in vision and imaging to automate this process.

When trees are sold to lumber processing companies, the chords, or large central trunk portions, are cut to specific lengths and their diameter is measured and noted. The combination of the chord length and diameter are used to calculate the number of 25.5 mm thick boards can be made-up. Measuring the diameter of the chords is a bit of a labour intensive and time consuming process that can potentially fall victim to the element of human error. The American company eSolutions developed an automated solution based on a machine vision system to measure the chords during the production process.

The goal was to integrate an extremely efficient, robust, and accurate measuring process without sacrificing time. As most of the loading and storage of the chords is done outdoors, a vision based system would require proper protection against the elements and repeatability with varying sunlight and weather conditions. The idea was to mount the entire vision system, including the camera, illumination, and processing hardware to the Tigercat grapple loader used to pick and move the chords from place to place. This allows the operator to do their job as usual and let the imaging system take the measurements for them while remaining seated inside of the machine.

Once a chord is grasped by loader it is briefly held in a position and briefly presented for imaging. Two laser sensors check for presence, absence, and distance. A Programmable Logic Controller (PLC) triggers the lasers and the information is sent to a personal computer mounted inside the machine controlling processing for the entire system. From there, a GigE camera acquires an image.

The real key to this application is providing repeatable illumination with the challenge of ever-changing outdoor lighting conditions. An ultra-high intensity XR256 series from Smart Vision Lights emitting 850 nm Infrared Illumination, was selected due to its extreme brightness output.

Infrared was selected, because high powered LEDs are widely offered in the wavelength of 850 nm and the spectral region of light produced by the sun is a bit less than other visible wavelength choices. The camera was matched with a Smart Vision Lights infrared bandpass filter. This filter only allows the wavelength of light produced by the XR256 through to the camera while blocking all others thus reducing the amount of ambient light. The pulse width of the strobe light was set for 100 ms to eliminate any potential for pixel blur during the motion of the chord.

This entire process is activated by a push-button system. All the operator has to do is present the chord and press a button for the system to engage and take an automated measurement and calculation using blob analysis tooling through the software. Greg Hilbert, President of eSolutions and head of this automation project, explains, "Much of the outside bark of the wood is eliminated from the image, providing the lumber company with a better estimate of board feet measure."

This vision based automated solution greatly aids in measuring the amount of raw lumber material on hand and dramatically increasing accuracy of order fulfillment. Knowing exactly how many boards can be cut from a single chord, without the added labour cost and potential error associated with manual measurement, was vital in creating a competitive advantage while maximising efficiency of inventory and process.

Smart Vision Lights

Muskegon, United States

Smart Vision Lights (SVL) was founded in 2007 in West Michigan and quickly developed into a leading developer and manufacturer of highperformance LED illumination systems with integrated current drive and IP rating.

Smart Vision Lights XR 256
  • High speed LED strobe light for up to 5000 strobes per second
  • Smart illumination without the need for an external controller
  • Ideal for high speed applications