Method for Identifying Timber Surface Properties

The present invention relates to a method for identifying the surface properties of timber, of sawn and planed timber and plywood in particular, when automatically determining defects and quality.

Sawn timber and plywood products to be used for various purposes are classified according to the surface properties of the product. The most important factors affecting the quality (and the strength) are knots, rot, bluestain, rough edges, skewness of grains, grub passages and mechanical defects. The price of board and plywood products is determined by quality classification.

The methods used today for determining timber surface properties or defects are based on tracing the intensity plane of radiation scattered from or gone through the surface to which electromagnetical radiation is directed by means of special geometrical arrangements. Depending on the radiation source, various light and radiation detectors as well as video and diode matrice cameras have been used for the detection. When cameras are used, image processing and pattern recognition algorithms setting great demands on computer capacity are often resorted to. In order to trace different defects, several wavelengths as well as different wavelength filters have been used in connection with the detectors and the cameras.

Specially suitable radiation sources for examining surfaces are lasers which yield a strong, monochromatic, collimated, coherent and often linearly polarized beam of light. It is easy to deviate a laser beam so that the entire surface to be examined can be analysed. A laser beam can also be focused into a very small spot, which enables also the examining of small surface parts having a size of about a micrometer.

The present invention improves the method disclosed in the U.S. Pat. No. 4,482,250. In the method of said patent the timber surface is scanned with a lineary polarized light beam the polarization plane of which is fixed according to the timber surface, and from the radiation reflected from the surface is measured the intensity component I.sub.max that has maintained the polarization plane and the intensity component I.sub.min that is perpendicular to it, and a depolarization variable is formed: ##EQU1##

The variable I.sub.min discloses darkness variations of the surface and the variable P discloses the fiber structure. By means of these variables a very far going identification of defects can be performed without a large computer capacity.

Behaviour of the variables I.sub.min and P depends on the used wavelength or wavelength range. An especially suitable wavelength for the inspection of a timber surface is 632,8 nm of a helium-neon laser, with which a darkness contrast in relation to a healthy surface can be achieved for all central defect types on timber surface.

The method according to the present invention improves the ability of the above mentioned method to separate defects between themselves by means of observing the behaviour of the variable I.sub.min while using two or several wavelengths or wavelength ranges and at the same time the behaviour of the variable P while using one or several wavelengths or wavelength ranges. When moving within the wavelength range of e.g. visible light information is received of intensity signals connected to the colour tone of the timber surface instead of contrast information of one wavelength. Accordingly different types of defects having possibly the same darkness contrast and fiber structure value can be identified on the basis of the colour tone. This feature can be used in identifying especially the following defects: rot, bluestain, colour defect and dirt.

The invention is described in more detail in the following with reference to the accompanying drawings, in which

FIG. 1 shows the principle of the equipment for one wavelength/wavelength range used in carrying out the method according to the invention,

FIG. 2 shows a simplified diagram illustrating the examining of the surface of sawn timber, and

FIG. 3 shows the use of information about the colour tone in the identification of different types of defects (rot and bluestain) when using two wavelengths.

In the measurement device shown in FIG. 1, reference number 1 refers to a light beam projected from a laser and directed towards the surface 2 of the object to be examined. The direction of the incident beam and the direction perpendicular to the surface form an angle .alpha., which is preferably smaller than 30.degree.. Part 3 of the diffused scattered radiation passes through a filter 4 corresponding to the used wavelength (interference) and through a polarization filter 5 to a detector 6 and part 7 (interference) through a filter 8 and a polarization filter 9 to a detector 10. The measuring directions of the scattered radiation and the direction perpendicular to the surface form angles .beta..sub.1 and .beta..sub.2, which are preferably smaller than 30.degree.. The transmission directions of the polarizing filters are at an angle of 90.degree. and the directions are defined by turning the filter 9 so that the intensity signal given by the detector 10 is at its largest (I.sub.max) and correspondingly the value of the detector 6 at its lowest (I.sub.min). The polarization of I.sub.max is the same as the polarization plane of the incident light if the length direction of the fibers of the timber surface is the same as the growth direction of the wood, as is normally the case. The intensity signals I.sub.max and I.sub.min are transmitted to a processor 11 where they are, in a manner known per se, transformed to a form suitable for determining the quality of timber.

The principal arrangement of the equipment intended for examining a board 12 is shown in FIG. 2. The light beam 1 is by means of a swivelling deviation mirror 13 caused to scan over a timber surface 14 in the cross direction of the board along a course 15. The board is moved in the longitudinal direction. The radiation scattered from the board surface is detected by the means according to FIG. 1. The light beam 1 includes two or several wavelength components.

FIG. 3 discloses a typical behaviour of intensity signals that are standardized as regards their level and received during a single light beam scan on a healthy (A), rotten (B) and bluestained (C) timber surface when the red wavelength of a HeNe-laser and one wavelength of a blue green laser (e.g. ARGON or HeNe) are used. The figure also illustrates how the knowledge about the colour tone can be used to separate between themselves rot and bluestain in the present case, although the depolarization (P) is the same for said surface types.

The invention is not limited to the above embodiment but it can be modified and applied within the inventive concept defined by the claims.