FI121400B - Close-up photogrammetry using structural light - Google Patents

Description

BACKGROUND OF THE INVENTION USING STRUCTURAL LIGHT PHOTOGRAPHY

Photogrammetry utilizes remote sensing based on photographs taken of a desired target. The near-field photogrammetry can be applied to systems based on computer vision. The resulting photometric measurement results can then be widely used in various applications.

10 There are numerous applications where 3D vision imaging is very useful. In general, such systems include a photocell, a camera, a lighting device, a computer, and a connection to a control system 15 that manages machine vision equipment. In such a system, the photocell triggers the camera when a suitable subject arrives near the photocell. The lighting should be arranged according to the ambient lighting conditions so that the subject is well illuminated to allow effective post-processing of the 20 images. The image can be processed on a computer, for example, by removing noise from the image or by simplifying the image in some known manner. The system can then calculate, measure or identify the desired parts of the object in the image. Finally, classification based on measurements or identifications can be made so that the desired performance can be performed on the subject. One example of such equipment is a bottle recovery machine.

Many different machine vision systems have been introduced. For example, U.S. Patent No. 7,032,458 to Tanaka discloses a mobile measurement system capable of measuring the three-dimensional structure of a desired object. There is a floodlight and camera mounted on the mobile rack. The location of the equipment can be changed, and with the help of location information and light and camera directions, the computing means can produce a 3D structure of the object. Accelerometers and direction-angle sensors are used in conjunction with wheel rotation sensors to locate the rack.

5 Note that lighting conditions have two effects on photogrammetric applications. While specially projected light toward a desired object can be very useful when measuring light with a sensor, ambient light generally complicates the measurement task, since ambient light can be considered as noise. For example, if the projected light is visible broadband light used during the daytime outdoors, measuring the projected light becomes virtually useless. This means practically the same thing as 15 outdoor lighting a flashlight on a bright sunny day.

In addition, the analysis of the image itself poses significant problems if the image quality is poor, for example due to other noise effects on the image. Proper filtering, lighting, and the choice of camera and settings will become crucial in such a situation.

Ambient conditions also require extensive knowledge of imaging techniques in addition to photogrammetric knowledge.

SUMMARY OF THE INVENTION

The present invention provides a method of collecting subject information using machine vision, the method comprising the steps of: projecting a pattern of substantially monochromatic electromagnetic radiation onto a surface of the subject, wherein the pattern is pulsed by at least one LED; 35 measuring at least two optical sensors on the surface of the object; and 3 filtering the spectrally measured pattern to obtain information about the subject.

In one embodiment of the invention, near-infrared radiation is used in said projected pattern.

In one embodiment of the invention, said pattern is produced by pulsed operation of at least one LED.

In one embodiment of the invention, said pattern is projected on a reflective projector.

In one embodiment of the invention, said image is projected using Digital Micromirror Device (DMD) technology.

In one embodiment of the invention, said pattern is measured simultaneously by at least two cameras having fixed locations relative to the projector.

In one embodiment of the invention, the projector and sensors are controlled such that they are oriented at substantially the same direction.

In one embodiment of the invention, said at least one LED is triggered in conjunction with said optical sensors.

In one embodiment of the invention, method 25 further comprises: measuring the target with said at least one LED on; re-measuring the target with said at least one LED off, using the same measurement parameters; and subtracting the second image from the first image to minimize the effect of ambient light.

In one embodiment of the invention, the measured and bandpass filtered image is converted to a binary 35 image.

In one embodiment of the invention, radial lines, parallel lines, a grid pattern or a predetermined set of dots are used as the projected pattern.

In one embodiment of the invention, the method comprises: changing the projected pattern during use; and re-measuring the pattern on the surface of the object.

In one embodiment of the invention, the information obtained is used in the photogrammetric analysis of the subject.

The data collection method disclosed in one embodiment of the invention is applied to a multifunctional forest machine apparatus.

In one embodiment of the multifunctional forest machine, the projector and said sensors are mounted on the harvester head of the multifunctional forest machine.

In one embodiment comprising a multifunctional forest machine, said pattern is measured from the surface of a tree 20 as the harvester head of the multifunctional forest machine approaches said tree.

The inventive idea of the invention comprises a corresponding system for gathering information from a subject using machine vision, the system comprising a projector arranged to project a pattern of substantially monochromatic electromagnetic radiation on a subject surface, wherein the projector is arranged to pulse the pattern with at least one LED; at least two optical sensors arranged to measure the pattern on the surface of the object; and a filter arranged to filter the spectrally measured pattern to obtain information about the subject. In other embodiments of the invention, the system components are arranged to perform the above-described process steps, as applicable.

In a further embodiment of the system, it further comprises a controller configured to perform various control functions for triggering, selecting, filtering, measuring, calculating, moving or orienting different parts of the system, or to be performed in said parts of the system.

In one embodiment of the invention, the data acquisition system is implemented, for example, in multifunctional forest machine equipment used by the forest industry.

In one embodiment of the invention, the data acquisition system 10 is used in a three-dimensional scanning application.

The inventive idea also encompasses the use of a computer program in which the program code controls various parts of the hardware according to the previously defined method steps. In one embodiment of the invention, the computer program is implemented on a computer readable medium.

The present invention solves the above-mentioned problems caused by ambient light disturbances in measurements. The use of monochromatic light source, reflective projector technology and image bandpass filtering solves the difficulties that have been present in prior art solutions.

Further, the ability to use an infinite number of different patterns combined with pulsed action to ensure high temporal intensity on the target surface results in a widely modifiable system that allows for realization on a wide front in a variety of applications in the near range photogrammetry.

BRIEF DESCRIPTION OF THE FIGURES

Figure 1 is a side view of an example of an apparatus according to the invention; Figure 2 shows an example of the apparatus according to the invention in front view; 6

Figure 3 shows the spectral reflectance of a softwood bark;

Figure 4 shows the spectral reflectance of conifer needles; Figure 5 illustrates the frequency response of an IBIS5 sensor;

Figure 6 illustrates the spectral intensity of H2W5-905 LED;

Figure 7 illustrates the spectral transmittance of a band pass filter at 905 nm; Fig. 8 shows an example of an LED control circuit;

Figure 9 shows a test in which ambient light is eliminated;

Figure 10 illustrates a coniferous frame mounted on a Valais-15 tu DLP projector; and

Figure 11 shows an example of a binary image for on-line measurement.

DETAILED DESCRIPTION OF THE INVENTION

Reference is made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings.

The present invention describes a novel method for collecting data for a photogrammetric measuring system. In this example, the data acquisition system comprises two cameras and a projector. In one embodiment of the invention, the projector is based on a so-called. For use with DLP (Digital Light Processing) technology. Generally speaking, the core of this technology is the so-called. a digital micromirror device (DMD) capable of creating different colors and shades with a variety of light sources or a color wheel combined with light reflected from small mirrors. However, in one embodiment of the present invention, DMD technology is applied to a monochrome source, that is to say, a narrowband electromagnetic radiation source. In the preferred 7 embodiments, the projected pattern is created from a source of substantially monochromatic electromagnetic radiation. By keeping a record of camera locations and orientations, and keeping a constant distance between them, it is possible to use a projector to project a predetermined pattern onto the surface of the subject and to use the image data recorded by the cameras for photogrammetric measurement using known techniques.

In an alternative embodiment, the projector 10 may be included in the bearing equipment of the system. This approach allows you to create a virtual point of view at the projector location and also use it for a measurement task. This creates an efficient data acquisition system with three cameras.

15 Reliable automatic target measurement requires either high processing power or appropriately coded targets. In one embodiment of the method, a set of images is produced from a subject in binary form, wherein the points representing the object are one and the background points 20 are zeros. This approach is similar to that of laser line scanning, where the pixels forming part of the line are one and the other zeros. In this sense, it is possible to extend the known methods of static structured light to a dynamic system. This is possible with a DMD-based projector because it is capable of projecting a random pattern, whereby the line sweep is only one of the infinite number of patterns.

In one embodiment of the invention, it is possible to manually change the projected pattern or to change the pattern sequentially. Such pattern variation can even be made during operation of the data logger.

In one embodiment of the invention, the image signal received by the camera-35 sensors is subjected to spectral filtering. Spectral filtering involves determining the wavelength range where a combination of sensor response, object reflectance, and ambient light intensity produces the best possible image. This provides the best possible contrast between the projected data points and the background.

5 The data acquisition system preferably comprises three components including two cameras and a projector. It is also possible to use at least one additional camera. In one example, the data acquisition device is arranged as a movable and rotatable device 10 which can be moved either manually (such as a hand-held scanner) or by machine control, whereby the data acquisition device can be mounted at one end of a controllable "arm". The projector and cameras can also be mounted on a head unit comprising 15 functional tools (for example, a gripping device). Such a device end can then be positioned at the end of such an arm, both of which can be guided. It is an object of one example of the invention to provide the widest possible distance between cameras in order to obtain good geometric accuracy for the observations. The projector can be positioned between cameras or otherwise within close range of cameras, but its exact location is not critical unless it is configured to function as a virtual camera by including it in system calibration. Placing the camera on the head of the unit gives a good view of the area. Because the head can be moved freely, the imaging system can be rotated, providing multiple angles for better geometric accuracy. On the other hand, the mechanical endurance of the head must be very good in order to provide a solid base for the system. This improves the stability of relative alignments, resulting in less need for calibration. However, the head of the device together with the data acquisition device may be susceptible to receiving rather severe shocks and vibrations during the use of some applications. These types of environmental conditions require that the moving parts of the imaging equipment 9 be kept to a minimum and must withstand vibrations and shocks. In addition, the system must be able to withstand high humidity and temperature variations in outdoor conditions.

In one embodiment of the invention, at least one infrared LED and DMD technology ("digital microplate device") are used as a source and reflector of electromagnetic radiation. It forms the core of the so-called. About the DLP system (digital light-10). The DMD chip includes an assembly of mirrors that can contain up to 2 million microscopically small mirrors mounted on hinges. Each mirror can either reflect light through the optic onto the surface of the object or reflect light to the heat sink. Essentially, this means that by using a DMD chip projector, it is possible to direct an arbitrary binary pattern of 2 million elements onto the surface of an object. Because the mirrors are reflective, the system has minimal light loss. In one embodiment of the invention, a consumer-level projector having 800x600 elements is used.

In the consumer market, the use of reflection technology enables the production of a projector using LED backlights. This approach eliminates the need for 25 color filter wheels to produce colors, since it is possible to use three LEDs with suitable spectral features to produce colors. Due to the relatively monochromatic nature of the LEDs, an ideal projector screen can be produced that reflects only narrow wavelength ranges of selected LEDs.

This results in improved contrast.

The same principle can be used in scientific imaging. The consumer-level projector LED backlights can be replaced by suitable LEDs when the camera uses optical bandpass filtering to suppress ambient light outside the selected band. This results in sufficient contrast between the projected data points of 10 J and the background illuminated by the excess ambient light.

Considerations such as the harmful nature of the radiation used and its effect on the surrounding environment 5 must also be taken into account. Visible light can be used, but dense flashes of light on the surface of the subject during use can distract the scanner user. This makes the use of the near infrared wavelength range (NIR) sensible because it is invisible to the human eye. Using LEDs as a light source instead of lasers is also important for eye safety reasons. The projector, equipped with near-infrared LED backlights, projects a non-coherent, non-interfering pattern that is also safe to use.

As a result, in one embodiment of the invention, the available wavelengths of LEDs suitable for scanning applications are in the near infrared range of electromagnetic radiation. In other words, the desired wavelength range of LEDs used as radiation sources is 800 nm to 1100 nm.

In one embodiment of the invention, the controller is arranged to control parts of the system. This includes controlling the projector's light source pattern, controlling the cameras used, and controlling the movement of the various components in the environment used. In one embodiment, the invention is implemented as a computer program that controls the various method steps of the invention.

In one embodiment of the invention, the data acquisition device is implemented in a forestry machine device designed to handle tree trunks in the forest.

An example of the hardware components and their relative locations is illustrated on page 1. This example relates to a forestry machine application used in the forest industry. However, the apparatus 35 may also be used in other applications where data acquisition is required for photogrammetric purposes. The apparatus of Figure 1 includes an arm 10 of a forestry machine 11. The gripping and handling portions 11 of the apparatus are disposed at the end of the arm 10, hereinafter referred to as harvester head 11. Projector 12 is positioned in line with harvester head 5. The two cameras 13 are positioned near the projector 12 so that there is a certain distance between the cameras. The locations of the projector and cameras themselves depend on the model of the harvester head. The opening angle α, which can be seen by both cameras 13 and the projector, is denoted by 14. The angle α depends on the equipment used, but in the preferred embodiment the angle is between 45 ° and 90 °. The cameras 13 and the projector 12 are attached to the harvester head and, as a result, they are also interconnected and consequently the functions of both the projector and the cameras can be synchronized.

Figure 2 is a front view of the structure of the apparatus. The harvester arm 20 is connected to the head 23 of the harvester 20. The cameras 21a, 21b acting as sensors are attached to the head 23 of the harvester so that there is a certain distance between the cameras 21a, 21b. The projector device 22 is connected near the cameras, in this example below them, so that the location of the projector 25 22 is fixed relative to the harvester head 23. The harvester head 23 is rotatable about the arm 20, but the control device synchronizes the cameras and the projector to the movements of the harvester head. The cameras and the projector are each connected to a control device so that they can be triggered in cooperation with each other.

The motivation of the device according to this embodiment of the invention is that the device should function properly in challenging forest conditions. The core of the research is to be able to implement a co-35 vision-driven forestry machine system that can operate in any environmental condition without sacrificing the efficiency of the forestry tool already achieved. The goal is to be able to measure all possible characteristics of a single tree trunk before touching it. If the 3D model of the tree can be measured as it approaches, the mal-5 can be used to assist in deciding how to cut and treat the tree trunk. Model data can also be used for a database containing all truncated tree trunks. Using such a data set to plan your next woodworking job is extremely useful.

The proposed approach to obtain measurement data suitable for photogrammetric techniques becomes challenging due to the environmental conditions in the working range of the forestry machine. Perhaps the 15 most challenging aspects to overcome are lighting because the pattern to be projected onto the surface of the object must be stronger than the ambient light. This results in the most difficult situation where the intensity of the projected light should exceed the intensity of sunlight on a bright summer day. Other environmental variables include rain, snow, leaves, twigs, vibrations, etc. These problems also mainly exist in other outdoor applications of the scanning device. In one embodiment of the present invention, the problem is solved by pulse shape operation of the projected pattern and spectral filtering of the images. The projector employs LEDs driven at high electrical current during pulse-mode operation to achieve high intensity over time on the target surface. 30 DMD-based projection technology is used in this context because its reflection-based operation allows for minimal light loss during projection. This approach is important because the forestry machine must maintain its normal operation. As a result, data collection has to be done in 35 real-time, resulting in short integration times of a fraction of a second to effectively stop motion. The short integration time also improves the relative intensity of the structured light compared to the ambient light, resulting in more efficient spectral filtering. This procedure can be compared to fast flash photography, which is often used to produce a black background in macro photography.

In this embodiment of the harvester application, implementing such a DMD projector based system in a challenging outdoor environment is useful because the DMD chip is in fact very durable. It is capable of withstanding 20g oscillations at 2000Hz and mechanical shocks up to 1500g (see Douglass, M., 2003, DMD Reliabil ity MEMS success story, Texas Instruments). Used cameras must also be durable. This can be achieved by using an electronic shutter. It eliminates the need for a mechanical shutter that could fail in challenging conditions. In addition, the optics should be secured with retaining screws to ensure the stability of the optic system and to reduce the need for system recalibration that is too frequent.

In one embodiment of the invention, suitable input frequencies are 850 nm, 870 nm, 905 nm and 950 nm. In this particular example, which is discussed in the following figures and description, a wavelength of 905 nm is used. In this case, the wavelength is selected by comparing the frequency response of the sensor and the reflectance of the rod and needles. The ideal situation is that the sensor sensitivity and bark reflectivity are high and the needles reflectivity is low. Preliminary tests have been measured on an ASD Field Spec Pro FR and show the spectral reflectance of a conifer bark and needles in Figures 3 and 4, respectively.

It can be seen from Figures 3 and 4 that the reflectance of the needles remains quite similar in the NIR-14 wavelength range, while the reflectivity of the bark appears to be constantly increasing. From this observation we can conclude that it would be more feasible to use a wavelength of 950 nm rather than 905 nm. However, the transducer sensitivity drops rapidly as a function of wavelength, as can be seen from the transducer response curve provided by the manufacturer in Figure 5. Also, filtering at 950 nm band proved to be more difficult than filtering at 905 nm. The selected cameo-10 frame, SMX-150M-E, used in this example uses an IBIS5B-NIR sensor with an improved frequency response in the NIR range due to a thicker epitactic silicon layer.

To find the right bandpass filter for this application, the frequency response of the Roithner Lasertechnics 15 H2W5-905 LED is measured. The results are shown

In Figure 6.

In one embodiment of the invention, it is feasible to make a suitable band-pass filter by combining a short-wave and a long-wave interference filter. Edmund optics # 48561 and # 47-588 filters are used for the 905 nm wavelength and the measured transmittance is shown in Figure 7. Figure 7 shows that the center of the pass band is located quite accurately at 905 nm and thus the filter-25 combination is very suitable for the selected LED .

The spectral intensity of the light in any outdoor environment also influences the band selection. Atmospheric absorption strongly influences the ambient light. The absorption properties of water vapor are particularly important because they are very pronounced at 905 nm. This results in the attenuation of ambient light on the surface of the subject, resulting in more efficient use of structured light.

A special driver circuit is required to operate the LED and the camera in pulse-mode operation. The purpose is first to turn on the LED, and when it reaches its maximum intensity, integration should begin.

15

Shortly after the set integration time is reached, the LED is allowed to go off. For this application, the following embodiment of the driver circuit prototype is implemented and is shown in Figure 8 of Figure 5. However, the parameter values are not limited to the values shown in the figure. The driver circuit may also be any other functional circuit capable of generating the desired radiation source depending on the application selected.

Figure 8 shows such a circuit comprising two transistor switches. The MJE3055 controls the charge stored in the condenser, while the BC337 controls the signal required to saturate the MJE3055 to release the charge and turn on the LED. Circuit 15 is connected to a computer providing power through a USB bus and a timing signal through the LPT port. The LPT port is connected to Dn_out as indicated in the circuit diagram. A 5V USB power supply line is used to charge the capacitor and also to saturate the MJE3055. The BC337 is saturated through the LPT port data connector. The same data connector can be used as an external shutter release for the SMX-150M-E. This will synchronize the camera and LED operation. The capacitance Cl may vary according to the lighting requirements. In order to maintain high current driving on the LED, the current prototype uses a combination of four capacitors to achieve a total capacitance of 19 mF.

The theoretical peak control current of an LED can simply be calculated using Ohm's law: 30

! = J

where I = current, U = voltage and R = resistance.

The circuit according to the present embodiment 35 has been tested with a theoretical peak current up to 8.3 Ar using an Rl value of 0.6 Ω. This 16 causes the LED to be controlled by a peak current well above the specification which produces the highest possible time intensity, improving the visibility of the structured light on the target surface. With the current implementation, the shortest possible LED pulse duration is 50 ms, which ensures that the LED is illuminated throughout the camera integration time, but still short enough to minimize the risk of temperature-induced LED failure. 10 On the other hand, camera parameters can be freely modified within specifications, giving a minimum integration time of 3 με. In practice, the most reasonable operating range is 1-10 ms.

Pulse mode operation has been tested using the introduced pulse control circuit in combination with 905 nm spectral filtering and a SMX-150M-E camera. The H2W5-905 LED is operated in a pulsed fashion with a theoretical peak current of 8.3 A and a 50 ms flash.

20 The camera's exposure value is determined by directing it to take a picture of a tree through the window glass. The integration time is varied until the image is properly exposed to obtain the highest luminance value of 240 from 1 to 255 at a 3 ms integration time. The corresponding pixel is a reflection of a leaf on a deciduous leaf on a sunny summer day.

Next, the camera is pointing towards a black plate, illuminated by a spotlight for 1000 W photo use. Exposure is varied by changing the spotlight distance from the black plate until the 30 highest stored luminance value is 240. Camera parameters are kept the same. This gives the same intensity on a black plate as the aforementioned reflection from a leaf.

The camera is set to external shutter mode, 35, and fired with a H2W5-905 LED aimed at a 250mm F4 Leitz Telyt-R lens, which produces approximately the same diameter as a black plate. After exposure, the same image is taken again without triggering the LED. The second image is considered black image data and subtracted from the first image.

Fig. 9 shows a black le vy illuminated by an LED. A little extra ambient light is still visible in the form of interference rings. These residues could be filtered off after finishing, but have been left for illustrative purposes. An Otta-10 bat image of the same subject synchronously with two cameras, the resulting information could be used for photogrammetric measurement tasks if the relative and internal directions of the cameras are known.

The unmodified Voigtländer DLP200 -15 projector is used in this example to illuminate the function of the structured light. The projector is used to project a test pattern of vertical white light strips on the surface of a softwood frame in dark conditions. Tracks are recorded with the Olympus E-330 - 20 with an 11-22mm Zuiko Digital lens set at 11mm focal length. This is shown in Figure 10.

The photograph is then enhanced by retrieving the edges of the projected tracks and marking each pixel as part of the 25 tracks with "l" and leaving the others as "0".

For the resulting image, see p. Figure 11, is a binary representation of vertical scan lines. This representation greatly reduces the time required to find similar features on a stereo image pair by automatic search methods. This makes the automated photogrammetric measurement system easier to implement and by allowing the user to select from an infinite variety of pro-35 scalable patterns, making the system highly suitable for a variety of applications.

18

Testing of the entire scanning hardware involves replacing the DLP projector's backlight with high-power 905nm LEDs controlled by the circuit described above and attaching the cameras and the projector to the harvester head. The relative and internal directions of the cameras should be calculated at the time they are mounted and secured.

The main problem with the implementation is perhaps the loss of light caused by the spread of the pattern over a sufficiently wide angle (FOV). The projector should provide the same perspective as the camera to take full advantage of the pictures. However, this would limit the choice of camera focal length. To overcome this problem, the projector can be equipped with a zoom lens, allowing the user to produce a higher intensity per square unit, while still providing the opportunity to illuminate a wide angle. Cameras should not be equipped with zoom lenses as they tend to lose their calibration data quite easily and even in the best 20 cases they require multiple sets of calibration parameters to ensure proper internal bearing when changing focal length.

In this embodiment, the projector does not need to be included in system calibration. Because the met-25 metering machine is used in the woods, the focal length of the projector should be varied according to the distance of the subject to allow a sharp pattern of structured light on the subject's surface. The focal length can be determined by using an existing 30 laser scanner in the forestry machine. By omitting the projector from system calibration, it can be used to field calibrate the camera system by projecting the calibration target on the projector and using it as input to the calibration.

The present invention may also be implemented in a number of other applications. Such applications include three-dimensional Scanning Devices, such as Medical Devices for example 19, which utilize machine vision to examine a patient.

It will be apparent to one skilled in the art that as technology advances, the basic concept of the invention can be implemented in a variety of ways. The invention and its embodiments are therefore not limited to the above examples only, but many modifications are possible within the scope of the inventive idea as defined in the claims.

Claims (32)

A method for gathering information from a subject using machine vision, the method comprising the steps of: projecting a pattern of substantially monochromatic electromagnetic radiation onto the surface of the subject, wherein the pattern is pulsed to be generated by at least one LED; measuring the pattern on the surface of the object using at least two optical sensors; and filtering the spectrally measured pattern to obtain information about the subject. A method according to claim 1, characterized in that the method further comprises 15 steps: using near-infrared radiation in the projected pattern. The method according to claim 1, characterized in that the method further comprises 20 steps: projecting the pattern with a reflective projector. A method according to claim 3, characterized in that the method further comprises the step of: projecting the pattern using digital micro-mirror technology (DMD). A method according to claim 1, characterized in that the method further comprises the step of: measuring the pattern simultaneously with at least two cameras having fixed relative positions to the projector. A method according to claim 1, characterized in that the method further comprises 35 steps: controlling the projector and sensors so that they are oriented at substantially the same angle. A method according to claim 1, characterized in that the method further comprises the step of: triggering at least one LED together with optical sensors. A method according to claim 1, characterized in that the method further comprises the steps of: measuring an object with at least one LED on; re-measuring the target with at least one LED off using the same measurement parameters; and subtracting the second image from the first image to minimize the effect of ambient light. The method of claim 1, characterized in that the method further comprises the step of: converting the measured and bandpass filtered image 20 into a binary image. A method according to claim 1, characterized in that the method further comprises the step of using radial lines, parallel lines, lattice patterns or a predetermined combination of points as the projected pattern. The method according to claim 1, characterized in that the method further comprises the steps of: changing the projected pattern during use; and re-measuring the pattern on the object surface. The method according to claim 1, characterized in that the method further comprises 35 steps: the resolved information is used for photogrammetric analysis of the subject. 13. A method according to claim 1, characterized in that the method further comprises the step of: applying the data collection method to the forestry machine equipment. 14. The method of claim 13, further comprising the step of: attaching the projector and sensors to the harvester head. Method according to claim 13, characterized in that the method further comprises the step of: measuring the pattern on the surface of the wood as the harvester head 15 approaches the wood. A system for collecting information from a subject using machine vision, characterized in that the system comprises: a projector (12, 22) arranged to project a pattern of substantially monochromatic electromagnetic radiation to a subject surface, wherein the projector (12, 22) is arranged to produce a pattern of pulse - formally with at least one LED; at least two optical sensors (13, 21a, 21b) arranged to measure the pattern on the surface of the object; and a filter arranged to filter the spectrally measured pattern to obtain information about the subject. A system according to claim 16, characterized in that the system further comprises: a projector (12, 22) arranged to use near-infrared radiation in the projected pattern 35. The system of claim 16, characterized in that the projector (12, 22) is a reflective projector. The system of claim 18, characterized in that the system further comprises: a projector (12, 22) configured to project the pattern using digital microreflection technology (DMD). The system of claim 16, characterized in that the system further comprises: at least two cameras (13, 21a, 21b) arranged to simultaneously measure the pattern, which cameras (13, 21a, 21b) have fixed relative positions to the projector (13). 12, 22). A system according to claim 16, characterized in that the system further comprises: 20 controllers arranged to control the projector (12, 22) and sensors (13, 21a, 21b) so that they are oriented at substantially the same angle. A system according to claim 16, characterized in that the system further comprises: a controller arranged to trip at least one LED together with optical sensors (13, 21a, 21b). A system according to claim 16, characterized in that the system further comprises: sensors (13, 21a, 21b) arranged to measure an object with at least one LED on; sensors (13, 21a, 21b) arranged to remeasure the target with at least one LED off using the same measurement parameters; and a controller arranged to reduce the second image from the first image to minimize the effect of ambient light. The system of claim 16, characterized in that the system further comprises: a controller arranged to convert the measured and bandpass filtered image into a binary image. A system according to claim 16, characterized in that the system further comprises: a projector (12, 22) arranged to use radial lines, parallel lines, a grid pattern or a predetermined point combination as a projection pattern. A system according to claim 16, characterized in that the system further comprises: 20 controllers arranged to change the pattern to be adjusted during use; and sensors (13, 21a, 21b) arranged to re-measure the pattern on the surface of the object. A system according to claim 16, characterized in that the system further comprises: a controller arranged to use the resolved information for photogrammetric analysis of the subject. System according to claim 16, characterized in that the system for collecting information is implemented in the forestry machine equipment (10, 11, 20, 23). The system of claim 28, characterized in that the system further comprises: a harvester head (11, 23) to which the projector (12, 22) and sensors (13, 21a, 21b) are attached. System according to claim 28, characterized in that the system further comprises: sensors (13, 21a, 21b) arranged to measure the pattern from the surface of the wood as the harvester head (11, 23) approaches the wood. 31. A computer program for gathering information from a subject using machine vision, wherein the computer program comprises program code, characterized in that the program code, when executed in a data processing device, is arranged to perform the steps of: controlling a substantially monochromatic controlling the measurement of the pattern on the object surface using at least two optical sensors; and controlling the spectral filtering of the measured pattern to obtain information about the subject. The computer program of claim 31, wherein the computer program is implemented on a computer readable medium.