JP5587030B2 - Peripheral edge detection device and outer edge detection method - Google Patents

Description

  The present invention relates to an outer edge detection apparatus and an outer edge detection method for a raw wood end, and is particularly suitable for use in an apparatus for detecting an outer edge of a raw wood end by image processing.

  While trying to permanently use timber, which is a resource that can be recycled, it is hoped to develop technology for processing small-sized raw wood, such as thinned wood that has been neglected as forest land, into useful components with added value. It is rare. As one of them, a single plate laminated material (LVL) or plywood, which is made by cutting a sheet-like veneer veneer from a raw wood with a cutting device called a veneer lace and laminating and bonding a plurality of veneer veneers, is attracting attention. .

  In veneer laces, both ends of the log are pivotally supported by a spindle with a claw (chuck) at the tip, and while the log is rotated, the table with the knife attached is moved toward the axis of rotation of the log at a specified speed. By moving, a veneer veneer having a constant thickness is manufactured.

  In order to process a veneer veneer from a raw wood having a complicated shape with a high yield, a process of determining a cutting axis that is a rotation axis when cutting a raw wood with a veneer race is important. Specifically, it is necessary to measure in advance the shape of the end of the log and obtain the cutting axis for obtaining the veneer veneer with the highest yield from the measured shape. For that purpose, it is required to grasp the shape of the mouth more accurately.

  Conventionally, various methods have been provided as methods for measuring the shape of the mouth. For example, there is known a method of measuring the outer peripheral shape of a log using a non-contact or contact sensor while rotating the log around a temporary rotation axis (see, for example, Patent Document 1). As described above, in the measurement method that captures the outer periphery of the raw wood in a circular shape, it is necessary to increase the number of measurement points on the outer periphery in order to accurately reproduce the outer peripheral shape of the lip, and the amount of data to be processed becomes enormous. Moreover, since the log must be rotated once in order to measure the outer peripheral shape at several measurement points of the log, the measurement takes time.

  On the other hand, a technique for detecting the shape of the lip by processing images of both lip faces of the raw wood photographed by the camera has been proposed (see, for example, Patent Document 2). In the technique described in Patent Document 2, the contours of the heartwood parts of both the mouthpiece surfaces with respect to the rotation center axis are measured based on the colors of the images of the both mouthpiece images taken by the camera, and the overall shape of the heartwood part is estimated. Has been made. Although the contour of the heartwood portion of the mouth end surface is measured here, it can be applied to the measurement of the outer peripheral edge of the end end surface.

JP 2004-338391 A JP 2010-46937 A

  However, the technique described in Patent Document 2 has a problem in that it is not always possible to accurately detect the outer peripheral edge of the lip of all the logs. In other words, like the sapwood part of cedar that has not been cut for a long time, the outer edge of the lip is white, and for the timber lip where a clear difference is obtained with respect to the background color, The outer peripheral edge can be accurately detected. However, when the color of the outer periphery of the lip is not clearly different from the background color, the outer edge cannot be accurately detected.

  In other words, in the case of a lip that has changed its color to black due to the invasion of discolored bacteria after cutting, or a lip that is dark in color, it can detect a clear difference from the background color. Since this is not possible, the outer edge of the butt cannot be detected. In that case, the operator has to manually set the cutting axis. For this reason, there is a problem in that the production efficiency is deteriorated, the accuracy of obtaining the cutting axis is lowered, and the product yield is lowered.

  The present invention has been made to solve such a problem, and an object of the present invention is to make it possible to accurately detect the outer peripheral edge of the lip regardless of the color of the lip of the original wood.

  In order to solve the above-described problem, in the present invention, two images captured with respect to the mouth of the raw wood so that the position of the raw wood in the image is different are defined as a first image and a second image, respectively. A parallax map is created in which the parallax representing the displacement amount of the corresponding pixel with the second image is represented as the pixel value of each pixel. Then, in the parallax map, a pixel area in which a pixel value in a predetermined central area or a difference between the average value and the pixel value falls within a predetermined range is specified as a lip area, and an outer edge of the lip area is specified as a lip edge. Like to do.

  According to the present invention configured as described above, the parallax values (pixel values on the parallax map) of the objects to be imaged included in the first image and the second image have the same distance from the imaging device. ) Have the same value, and parallax values are different for objects to be imaged having different distances from the imaging device. That is, the greater the distance from the imaging device, the smaller the parallax, and the smaller the distance from the imaging device, the greater the parallax. Therefore, the value of the parallax is almost the same for the end face of the original wood, and the value of the parallax is different for the background portion around the end face. Therefore, it is possible to specify an area of the lip surface having substantially the same parallax value as the lip area, and to specify an outer edge of the lip area as the rim edge. As a result, it is possible to accurately detect the outer edge of the lip based on the parallax, even for a timber that has conventionally been unable to detect the outer edge due to the color state of the original lip.

It is a figure which shows the structural example of the whole edge detection system to which the outer periphery edge detection apparatus of the raw wood clerk by this embodiment is applied. It is a perspective view which shows the example of a partial structure of the edge detection system shown in FIG. It is a figure which shows the function structural example of the computer with which the outer periphery edge detection apparatus of the raw wood clerk by this embodiment is provided. It is a figure which shows the example of the 1st image and 2nd image of this embodiment. It is a figure for demonstrating the difference process by the moving object area extraction part of this embodiment. It is a figure for demonstrating the calculation method of the parallax by the parallax map creation part of this embodiment. It is a figure for demonstrating the calculation method of the similarity degree at the time of calculating | requiring a parallax by the parallax map creation part of this embodiment. It is a figure which shows the relationship between the parallax and similarity of this embodiment. It is a figure which shows an example of the parallax map created by the parallax map creation part of this embodiment. It is a figure which shows the example of the lip | edge end edge specified by the lip end edge specific | specification part of this embodiment. It is a figure which shows the positional relationship of the 1st image before and behind the movement of the support stand by this embodiment, and a 2nd image. It is a figure which shows the positional relationship of the raw wood before and behind the movement of the support stand by this embodiment, a support stand, and a camera. It is a flowchart which shows the operation example of the outer periphery edge system by this embodiment. It is a flowchart which shows the detail of the parallax map creation process by this embodiment. It is a figure which shows the structural example in the case of installing the support stand of this embodiment so that reciprocation is possible in a Y-axis direction.

  Hereinafter, an embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a diagram showing a configuration example of an entire edge detection system to which an outer edge detection device for a log end according to this embodiment is applied. FIG. 2 is a perspective view illustrating a partial configuration example of the edge detection system illustrated in FIG. 1. FIG. 3 is a diagram illustrating an example of a functional configuration of a computer included in the apparatus for detecting an edge of the log edge according to the present embodiment.

  As shown in FIGS. 1 and 2, the edge detection system according to the present embodiment includes cameras 1A and 1B, illumination devices 2A and 2B, support bases 3A and 3B, left and right moving devices 4A and 4B, and up and down moving devices 5A and 5B. An interface board 6 and a computer 7 are provided. Of these, the computer 7 corresponds to an outer edge detection device for a log mouth.

  The cameras 1 </ b> A and 1 </ b> B are imaging devices installed on both ends of the log 100, and image the wood ends 101 </ b> A and 101 </ b> B of the log 100. The images captured by the cameras 1A and 1B include not only the mouths 101A and 101B of the log 100 but also the surrounding background. The cameras 1A and 1B are connected to a computer 7 through an interface board 6, and images taken by the cameras 1A and 1B are stored in an image memory of the computer 7. For convenience of explanation, the horizontal axis on the imaging surface is the X axis, the vertical axis is the Y axis, and the optical axes of the cameras 1A and 1B are the Z axis.

  The illuminating devices 2A and 2B are installed toward both the mouth ends 101A and 101B of the log 100, and illuminate the mouths 101A and 101B of the log 100 with sufficient brightness when imaged by the cameras 1A and 1B. The illuminating devices 2A and 2B may be light emitting devices such as an electronic flash.

  The support bases 3A and 3B are support members that support the log 100 at two locations in the longitudinal direction (Z-axis direction). The support bases 3A and 3B have support portions having a V-shaped cross section so that the center of the substantially circular cross section of the log 100 is positioned directly above the V-shaped break point due to the weight of the log 100. It has become.

  The left and right moving devices 4A and 4B are provided on the two support bases 3A and 3B, respectively, and are actuated by a movement command signal from the computer 7, and are supported in a direction orthogonal to the optical axis direction (Z axis direction) of the cameras 1A and 1B. The bases 3A and 3B are moved. In the present embodiment, the left and right moving devices 4A and 4B are configured to move the support bases 3A and 3B in the X-axis direction from the first position to the second position. As a result, the log 100 also moves from the first position to the second position along with the movement of the support bases 3A and 3B.

  The cameras 1A and 1B described above take images of the wood ends 101A and 101B of the log 100 at the first position and the second position, respectively. FIG. 4 is a diagram illustrating images captured at the first position and the second position (referred to as a first image and a second image, respectively). FIG. 4 shows a captured image on the side of the one end 101A, and illustrations of backgrounds other than the support base 3A, the left and right moving device 4A, and the log 100 are omitted.

  In the present embodiment, the first position is a temporary center position where both the mouth ends 101A and 101B of the log 100 are located substantially at the center of the imaging surface. The second position is a position moved from the first position by an arbitrary distance s (mm) in the X-axis direction. In the present embodiment, the left and right moving devices 4A and 4B are constituted by air cylinders, and the support bases 3A and 3B are moved by a distance s in the X-axis direction.

  When the first and second images are captured by the cameras 1A and 1B, the vertical movement mechanisms 5A and 5B position the log 100 to a position (temporary center position) where the throats 101A and 101B are substantially at the center of the imaging surface. In order to do this, the support bases 3A and 3B are moved in the Y-axis direction. In the present embodiment, the support bases 3A and 3B are raised to the temporary center position by the vertical movement mechanisms 5A and 5B in accordance with the temporary diameter of the log 100 that has been measured in advance by a laser distance meter or the like with rough accuracy.

  As shown in FIG. 3, the computer 7 includes an image processor 10, a control unit 20 such as a microcomputer, an image memory 30, and a ROM 40. The control unit 20 outputs a movement command signal to the left and right moving devices 4A and 4B and the vertical moving mechanisms 5A and 5B, and moves the support bases 3A and 3B. In addition, the control unit 20 outputs an imaging command signal to the cameras 1A and 1B to capture the first image and the second image. That is, the control unit 20 controls the cameras 1A and 1B to capture the first image when the support bases 3A and 3B are in the first position, and moves the support bases 3A and 3B from the first position to the second position. After the movement, the camera 1A, 1B is controlled to take a second image.

  The image memory 30 stores images captured by the cameras 1 </ b> A and 1 </ b> B and captured by the interface board 6 into the computer 7 and images being processed by the image processor 10. The image processor 10 performs the following image processing on the image stored in the image memory 30 in accordance with the lip edge detection program stored in the ROM 40. The butt edge detection program may be stored in another recording medium such as a RAM, a CD-ROM, a hard disk, an optical disk, a magneto-optical disk, a DVD, or a semiconductor memory instead of the ROM 40.

  The image processing processor 10 includes an image acquisition unit 11, a moving object area extraction unit 12, a parallax map creation unit 13, a lip region specifying unit 14, a lip end edge specifying unit 15, and a lip distance calculating unit 16 as functional configurations. Yes. The image acquisition unit 11 is a camera in a state where the relative position of the log 100 with respect to the cameras 1A and 1B is different in the X-axis direction (that is, in a state where the mouths 101A and 101B of the log 100 are in the first position and the second position). Two images (a first image and a second image) captured with respect to the end portions 101A and 101B by 1A and 1B are acquired.

  As shown in FIG. 5, the moving object area extraction unit 12 calculates the difference between the first image and the second image, so that the object (object to be imaged) moved by the left and right moving devices 4A and 4B is moved before and after the movement. An area on the existing image is extracted as the moving object area 50. In the present embodiment, the moving objects in the captured image are the support bases 3A and 3B, the cylinder portions of the left and right moving devices 4A and 4B, and the wood ends 101A and 101B of the log 100, and other portions move. Not.

  Specifically, the moving object area extraction unit 12 generates a difference processed image by subtracting each pixel value of the second image from each pixel value of the first image. In this difference processed image, the difference value is zero in all areas where the object is not moving, such as the background reflected in the vicinity of the mouth ends 101A and 101B. On the other hand, the difference value is a value other than zero in the contour portion of the moving object. Thereby, the moving object area extraction unit 12 extracts, as the moving object area 50, an area obtained by combining the area where the moving object exists on the first image and the area where the moving object exists on the second image. Can do.

  The parallax map creation unit 13 calculates, for each pixel of the first image, a parallax indicating the amount of displacement (number of pixels) of the corresponding pixel between the first image and the second image, and the parallax is calculated for each pixel of the pixel. A parallax map expressed as a value is created. The parallax takes different values depending on the distance between the cameras 1A and 1B and the object to be imaged. That is, the parallax becomes smaller as the object to be imaged has a larger distance to the cameras 1A and 1B, and the parallax becomes larger as the object to be imaged has a smaller distance from the cameras 1A and 1B. Accordingly, among the imaging objects included in the first image and the second image, the imaging objects having the same distance from the cameras 1A and 1B have the same parallax (pixel values on the parallax map), and the cameras 1A and 1B. The object to be imaged having a different distance from each other has a different parallax value.

  Here, how to obtain the parallax by the parallax map creation unit 13 will be specifically described with reference to FIGS. FIG. 6 is a diagram for explaining a parallax calculation method. FIG. 7 is a diagram for explaining a method of calculating similarity when obtaining parallax. FIG. 8 is a diagram illustrating the relationship between parallax and similarity.

  First, as shown in FIG. 6, the parallax map creation unit 13 adds 3 × 3 pixels to a target pixel (coordinates are (m, n)) and a neighboring pixel adjacent to the target pixel of the first image as a reference. Window 51 is set. Similarly, a window 52 of 3 × 3 pixels is set at the same coordinates (m, n) of the second image to be compared. Then, the similarity R between the two windows 51 and 52 is calculated.

The similarity R is an index that represents the degree of coincidence of two compared points. As illustrated in FIG. 7, the target pixel value of the window 51 of the first image as a reference is a, and the neighboring pixel values are a _{1 to} a ₈ . Further, the comparison pixel value of the window 52 of the second image to be compared is b, and the neighboring pixel values are b _{1 to} b ₈ . In this case, for example, the parallax map creation unit 13 obtains the similarity R by the equation R = Σ ( _ak− a) (b _k −b).

Next, as illustrated in FIG. 6, the parallax map creation unit 13 shifts the window 52 of the second image to be compared by one pixel in the X-axis direction (the direction in which the log 100 is moved from the first position to the second position), The similarity R is calculated. Similarly, the second image window 52 is sequentially shifted by one pixel to calculate the similarity R, and this is repeated until the comparison pixel coordinate becomes (m + d _max , n). The result is shown in a graph as shown in FIG.

As shown in FIG. 8, the window 51 of the first image as a reference is similar to the window 52 of the second image to be compared, and the target pixel (m, n) of the window 51 and the comparison pixel (m + d, n) is a point corresponding to each other, the similarity R indicates a maximum value. In this case, the number of pixels d indicating the amount of displacement in the X-axis direction is the parallax of the pixel of interest at the coordinates (m, n) of the first image. Note that d _max is an upper limit setting value of parallax, and is set as appropriate in advance depending on the resolution of the cameras 1A and 1B, the system configuration, and the allowable length of the log 100 to be input. For example, d _max = 40 pixels.

  Here, FIG. 8 will be described in detail. A solid line 53 represents the degree of similarity R when a pixel in the central area of the wood ends 101A and 101B of the log 100 is the target pixel. A broken line 54 represents the similarity R when a pixel in the region of the support bases 3A and 3B is a target pixel. Since the displacement amount of the pixel at the point where the maximum value of the similarity R is the maximum is parallax, the parallax d is 35 pixels in the central area of the lip end 101A, 101B, and the pixels in the area of the support bases 3A, 3B are parallax d is 25 pixels.

  The magnitude of the parallax d changes according to the distance from the cameras 1A and 1B to the object to be imaged. When the parallax d is large, the object to be imaged is closer to the cameras 1A and 1B, and as the parallax d becomes smaller, the object to be imaged is farther away. It can be said that it is located. Therefore, the example of FIG. 8 shows that the support bases 3A and 3B are located farther from the wood ends 101A and 101B of the log 100.

  The parallax map creation unit 13 sequentially moves the coordinates (m, n) of the pixel of interest using the method described above, and obtains the parallax d for the entire region of the first image as a reference. Then, a parallax map is created by using the parallax d obtained for each pixel of the first image as the pixel value of each pixel.

  FIG. 9 is a diagram illustrating an example of a parallax map created by the parallax map creating unit 13. In FIG. 9, the numerical value in the image indicates the value of the parallax d. As shown in FIG. 9, the parallax map is in a state in which pixel groups indicating the same parallax d value are mapped (segmented) on the image. That is, the parallax map is characterized in that regions of the imaging target having the same distance from the cameras 1A and 1B show substantially the same pixel value. In FIG. 9, there is a blank area that does not indicate a pixel value, but this is simply omitted for convenience of explanation, and actually there is a pixel value corresponding to parallax (a value close to zero).

  The lip region specifying unit 14 specifies the regions of the lip portions 101A and 101B by using the feature of the parallax map that the regions of the imaging objects having the same distance from the cameras 1A and 1B show substantially the same pixel values. Specifically, in the disparity map created by the disparity map creating unit 13, the lip region specifying unit 14 has a predetermined center area pixel value (a value of the disparity d) or a difference between the average value and the pixel value. An area of a pixel that falls within the range is specified as a lip end area.

  For example, as shown in FIG. 9, the lip region specifying unit 14 sets a central area 60 (area indicated by a broken line) of an arbitrary size at the center of the imaging surface that is around the centers of the lip ends 101A and 101B. The average pixel value Av of the pixels in the central area 60 is obtained. Then, the xylem area specifying unit 14 specifies an area of a pixel in which the difference between each pixel value and the average pixel value Av is within a predetermined allowable range α in the parallax map as the lip area 61 (area indicated by a thick solid line). .

  The size of the central area 60 may be an area smaller than the minimum log diameter to be input to the edge detection system. More specifically, the central area 60 is set in advance according to the minimum log diameter, the resolution of the cameras 1A and 1B, and the approximate distance between the cameras 1A and 1B and the wood ends 101A and 101B of the log 100. In the present embodiment, a circle having a diameter of 100 pixels is set at the central portion of the imaging surface, and this is defined as the central area 60.

  Next, the allowable range α will be described. The end face of the raw wood 100 is not necessarily perpendicular to the outer peripheral surface of the raw wood 100, and most of the raw wood 100 is inclined and cut. For this reason, the distance between the front end of the log 100 supported by the support bases 3A and 3B and the cameras 1A and 1B is not always a constant distance. It is necessary to set a range. More specifically, the allowable range α is set in advance according to the state of the log 100, the resolution of the cameras 1A and 1B, and the approximate distance between the cameras 1A and 1B and the wood ends 101A and 101B of the log 100. In the present embodiment, the allowable range α = 2 pixels.

  As shown in FIG. 9, it is assumed that the average value Av of the pixel values in the central area 60 of the parallax map is “35”. In this case, the range of pixel values (Av−α to Av + α) for specifying the lip region 61 is “33” to “37”, and the lip region specifying unit 14 specifies the region indicating the pixel value in this range as the lip region 61. To do. In addition, since the pixel value of the area 62 of the support bases 3A and 3B is “25”, it exceeds the allowable range α and is not specified as the lip area 61.

  The lip end edge specifying unit 15 uses the lip end of the lip region 61 specified in the moving object area 50 extracted by the moving object area extracting unit 12 among the lip region 61 specified by the lip region specifying unit 14 as a lip. The outer peripheral edges of 101A and 101B are specified. FIG. 10 is a diagram illustrating an example of the lip outer edge specified by the lip outer edge specifying unit 15. As shown in FIG. 10, by narrowing down within the moving object area 50 and specifying the outer peripheral edge of the lip region 61, the noise region 64 of the reflection of the surrounding background outside the moving object area 50 is removed, and the lip outer edge 63. Can be identified accurately.

  The pier end distance calculating unit 16 calculates the distance between the cameras 1A and 1B and the pier ends 101A and 101B of the log 100 using the feature of the parallax map. In other words, the culm distance calculation unit 16 determines the camera 1A based on the parallax of the lip region in the parallax map, the parallax of the support base region, and the actual distance on the machine between the cameras 1A and 1B and the support bases 3A and 3B. , 1B and the timbers 101A and 101B of the log 100 are calculated.

  Hereinafter, a specific method for obtaining the distance between the cameras 1A and 1B and the end portions 101A and 101B will be described with reference to FIGS. FIG. 11 is a diagram illustrating a positional relationship between the first image and the second image before and after the movement of the support bases 3A and 3B. FIG. 12 is a diagram illustrating a positional relationship among the raw wood 100, the support base 3A, and the camera 1A before and after the movement of the support bases 3A and 3B.

  In FIG. 11, the first image is represented by a solid line, and the second image is represented by a broken line. D1 indicates the parallax of the central section 60 in the lip region 61 (d1 = 35), and d2 indicates the parallax of the support base region (d2 = 25). As shown in FIG. 12, in the first image and the second image that are projected onto the imaging surface 71 through the lens 70 of the camera 1A, the parallax d1 of the log pier 101A that is close to the camera 1A is large and is far from the camera 1A. It can be seen that the parallax d2 of a certain support base 3A is small.

As shown in FIG. 12, the movement distance in the X-axis direction of the support base 3A is s, the distance from the lens 70 of the camera 1A to the raw wood end 101A is L1, the distance from the lens 70 of the camera 1A to the support base 3A is L2, Assuming that the focal length of the camera 1A is f, the parallax of the log end 101A on the imaging surface 71 is d1, and the parallax of the support base 3A is d2, the following equation is established from the principle of triangulation.
s / L1 = d1 / f, s / L2 = d2 / f

  From the above two formulas, L1 = L2 × d2 / d1. Here, the distance L2 from the lens 70 of the camera 1A to the support base 3A is measured in advance in order to be determined by the machine configuration. By doing so, the distance L1 from the lens 70 of the camera 1A to the original wood end 101A can be calculated by obtaining the disparity d1 of the end area and the disparity d2 of the support base area. That is, the position in the Z-axis direction of the raw wood end 101A can be calculated. Specifically, the butt end distance calculation unit 16 calculates an average pixel value of pixels in the support base region 62 set in advance in the parallax map shown in FIG. The value is calculated as d1, and L1 is calculated by the above formula.

  Next, the operation of the outer edge system according to this embodiment configured as described above will be described. FIG. 13 is a flowchart illustrating an operation example of the outer peripheral edge system according to the present embodiment. The flowchart shown in FIG. 13 starts when the support bases 3A and 3B are moved to the first position by the left and right moving devices 4A and 4B and the up and down moving mechanisms 5A and 5B.

  In FIG. 13, first, the wood ends 101A and 101B of the log 100 are imaged by the cameras 1A and 1B. The image acquisition unit 11 acquires images captured by the cameras 1A and 1B via the interface board 6 and stores them in the image memory 30 as a first image (step S1). In addition, it is preferable that the cameras 1A and 1B capture images alternately with a time difference in order to prevent the lighting devices 2A and 2B from being reflected in each other.

  Subsequently, the left and right moving devices 4A and 4B move the support tables 3A and 3B together with the log 100 from the first position to the second position by a distance s (30 mm) in accordance with a movement command signal from the control unit 20. After the movement is completed, the cameras 1A and 1B again capture the logs of the log 101A and 101B. And the image acquisition part 11 acquires the image imaged with camera 1A, 1B via the interface board 6, and preserve | saves it in the image memory 30 as a 2nd image (step S2).

  Next, the moving object area extraction unit 12 extracts the moving object area 50 by taking the difference between the first image and the second image stored in the image memory 30 (step S3). Information on the extracted moving object area 50 is stored in the image memory 30. Next, the parallax map creation unit 13 compares the first image and the second image stored in the image memory 30, obtains the parallax at each pixel of the first image, and sets the parallax as the pixel value of each pixel. To create a parallax map (step S4).

FIG. 14 is a flowchart showing details of the parallax map creation processing. In FIG. 14, the parallax map creation unit 13 first sets the value of the Y coordinate n of the target pixel in the first image to “1” (step S11) and sets the value of the X coordinate m to “1”. (Step S12). Further, the parallax map creation unit 13 sets the value of the parameter k representing the displacement amount of the comparison pixel in the second image to “0” (step S13). Up to here is the initial setting. Note that m = 1, 2,..., M _max and n = 1, 2 _,. m _max is the number of pixels in the X-axis direction of the imaging surface, and n _max is the number of pixels in the Y-axis direction of the imaging surface. In addition, k = 0, 1, 2,..., D _max , and d _max is a preset parallax upper limit value.

After the initial setting is completed, the parallax map creation unit 13 calculates the similarity R between the target pixel at the coordinates (m, n) in the first image and the comparison pixel at the coordinates (m + k, n) in the second image. (Step S14). Then, it is determined whether or not the value of the movement amount parameter k exceeds d _max (step S15). If not, the value of the movement amount parameter k is incremented (step S16), and the process returns to step S14. In this manner, the similarity R is repeatedly calculated while increasing the value of the movement amount parameter k by 1 up to d _max for the comparison pixel in the second image.

When the value of the movement amount parameter k exceeds d _max , the value of the movement amount parameter k when the similarity R calculated so far shows the maximum value is used as the parallax d of the target pixel in the first image. Determine (step S17). The parallax map creation unit 13 stores the value of the parallax d obtained for the pixel of interest at coordinates (m, n) in the image memory 30 as the pixel value D (m, n) of the parallax map (step S18).

Thereafter, the parallax map creation unit 13 determines whether or not the value of the X coordinate m of the pixel of interest exceeds m _max (step S19), and if not, increments the value of the X coordinate m (step S20). The process returns to step S13. On the other hand, if the value of the X coordinate m of the pixel of interest exceeds m _max , it is determined whether or not the value of the Y coordinate n of the pixel of interest exceeds n _max (step S21). Is incremented (step S22), and the process returns to step S12.

On the other hand, when the value of the Y coordinate n of the pixel of interest exceeds n _max , the parallax map creation unit 13 uses the pixel value D (m, n) of each pixel stored in the image memory 30 and uses the parallax map. Is created (step S23). In other words, the pixel values D (m, n) of the entire region are obtained by performing the processes of steps S14 to S18 on the entire region of the first image and stored in the image memory 30. The parallax map creation unit 13 creates a parallax map from the pixel values D (m, n) of the pixels stored in this way.

  Returning to FIG. 13, next, the lip region specifying unit 14 on the parallax map created by the parallax map creating unit 13 has substantially the same pixel value as the average pixel value of the central area 60 where the timbers 101A and 101B of the log 100 exist. A lip region 61, which is a pixel region indicating, is specified (step S5).

  Next, the lip end edge specifying unit 15 reads the information of the moving object area 50 from the image memory 30 and specifies the rim edge of the lip region 61 existing in the moving object area 50 as the lip end edge 63 (step S6). ). Finally, the culvert distance calculation unit 16 calculates the distance between the cameras 1A and 1B and the culm 101A and 101B of the log 100 that is the object to be captured using the parallax map created by the parallax map creation unit 13 ( Step S7).

  As described above in detail, in the present embodiment, two images captured with respect to the mouths 101A and 101B of the log 100 are acquired as the first image and the second image, respectively, so that the position where the log 100 is captured in the image is different. To do. And the parallax map which represented the parallax showing the displacement amount of the corresponding pixel between the 1st image and the 2nd image as a pixel value of each pixel is created. Further, in the parallax map, a pixel area in which the difference between the average pixel value and the pixel value of the predetermined central area 60 falls within a predetermined range is specified as a lip area 61, and an outer edge of the lip area 61 is defined as a lip edge 63. I try to identify.

  According to the present embodiment configured as described above, as illustrated in FIG. 9, the parallax values are obtained for the areas of the wooden ends 101 </ b> A and 101 </ b> B of the raw tree 100 among the captured objects included in the first image and the second image. (Pixel values on the parallax map) are substantially the same, and the parallax values are different for the background portions around the mouthpieces 101A and 101B. Therefore, a region having substantially the same parallax value can be specified as the lip end region 61, and the outer peripheral edge of the lip end region 61 can be specified as the lip end outer peripheral edge 63.

  In addition, in the present embodiment, the noise edge 64 of the reflection of the surrounding background outside the moving object area 50 is removed by narrowing down the moving object area 50 to identify the outer edge of the lip area 61, and the lip outer edge 63. Can be identified accurately. As a result, the rim end edge 63 can be detected with high accuracy based on the parallax even for the timber 100 that has conventionally been unable to detect the rim edge due to the color state of the lip end. As a result, the accuracy of the centering operation of the log 100 is improved, and the product yield and work efficiency can be increased.

  Further, according to the present embodiment, it is possible to calculate the distance between the cameras 1A and 1B and the wood ends 101A and 101B of the log 100 by using the parallax map created to detect the outer edge 63 of the mouth end. If the distances from the cameras 1A, 1B to the two piers 101A, 101B can be calculated, the length and position of the log 100 can be known. Thereby, it is possible to use the calculation result of the distance for positioning the log 100 with respect to the subsequent veneer race. Conventionally, a contact-type position measuring mechanism for measuring the position by mechanically contacting the end portions 101A and 101B, a non-contact type laser displacement sensor, and the like are separately provided to measure the positions of the end portions 101A and 101B. According to the present embodiment, it is possible to measure the positions of the end portions 101A and 101B without providing these separate position measuring mechanisms and laser displacement sensors.

  In the above embodiment, an air cylinder having a stroke of 30 mm is used as the left and right moving devices 4A and 4B, but the present invention is not limited to this. For example, a moving device using a crank method using a motor or a screw feed method driven by a motor with a position detection function may be used. In short, it is only necessary that the support bases 3A and 3B can be moved by an arbitrary distance s by the operation signal.

  Moreover, in the said embodiment, although the moving distance s of the X-axis direction of support stand 3A, 3B was 30 mm, this invention is not limited to this. That is, it is only necessary that the images of the lip 101A and 101B before and after the movement are within the range of the imaging surface and that there is a significant difference in the displacement amount between the two images sufficient to specify the lip region 61 on the parallax map. The distance s is not particularly limited to 30 mm.

  Moreover, in the said embodiment, although the support member which has a V-shaped support part is used as support stand 3A, 3B which supports the log 100, this invention is not limited to this. In other words, any structure that can support the log 100 may be used. For example, the structure in which the outer periphery of the log 100 is clamped by an arm with a claw may be used.

  Moreover, in the said embodiment, although the similarity R was used as the parameter | index which specifies a parallax in the parallax map creation part 13, this invention is not limited to this. For example, a value obtained by a known method such as a square residual method, a residual successive test method, or a normalized cross correlation method may be used as an index.

  Moreover, although the said embodiment calculated | required the parallax d about the pixel of the whole area of a 1st image by the parallax map preparation part 13, and demonstrated the example which produces a parallax map from the parallax d of the pixel of the said whole area, this invention is It is not limited to this. For example, the parallax map creation unit 13 calculates a parallax d only for each pixel in the moving object area 50 of the first image, and a parallax map in which the parallax d in the moving object area 50 is expressed as a pixel value of each pixel. You may make it create. In this case, a parallax map is created by repeatedly performing the processes in steps S14 to S18 in FIG. 14 only for the pixels in the moving object area 50 extracted in step S3 in FIG.

  In this way, the processes in steps S5 to S7 in FIG. 13 need only be performed on the pixels in the moving object area 50. Further, since there is only one lip area 61 specified by the lip area specifying unit 14, the outer edge of the lip area 61 may be specified as the lip edge 63 as it is in the process of step 6. According to this method, the area for obtaining the parallax may be only in the moving object area 50, the processing time can be remarkably shortened, and the operating efficiency of the detection processing of the rim outer edge can be increased.

  Further, the parallax d of the parallax map is obtained for the pixels of the entire area of the first image, and then the lip end specifying unit 14 narrows down the moving object area 50 in the parallax map and specifies the lip end area 61. Good. Also in this case, since there is only one lip region 61 specified by the lip region specifying unit 14, the outer edge of the lip region 61 may be specified as the lip edge 63 as it is in the process of step 6.

  Further, the processing of the moving object area extraction unit 12 may be omitted. For example, when the noise area 64 in the background of the lip ends 101A and 101B hardly occurs by surrounding the periphery of the lip ends 101A and 101B with a curtain or the like so that there is no reflection from the periphery of the lip ends 101A and 101B, the moving object area Even if the processing of the extraction unit 12 is omitted, it is possible to specify the lip end edge 63.

  In the above embodiment, the support bases 3A and 3B are installed so as to be reciprocally movable in the X-axis direction by the left and right moving devices 4A and 4B, but may be installed so as to be reciprocally movable in the Y-axis direction. That is, in the above embodiment, an example is shown in which the raw wood 100 is moved in the X-axis direction and the parallax in the X-axis direction is obtained. However, the parallax in the Y-axis direction may be obtained by moving the raw wood 100 in the Y-axis direction. Good. Even in this way, the lip outer peripheral edge 63 of the log 100 can be detected in the same manner.

  FIG. 15 is a diagram illustrating a configuration example when the support bases 3A and 3B are installed so as to be reciprocally movable in the Y-axis direction. In FIG. 15, the support bases 3A and 3B are guided by sliding members 75A and 75B on both side surfaces thereof, and are installed so as to be reciprocally movable in the Y-axis direction. The vertical movement devices 76A and 76B detect the positions of movement screws 77A and 77B, one ends of which are rotatably fixed to the support bases 3A and 3B, motors 78A and 78B for driving the screws, and the support bases 3A and 3B. It is composed of encoders 79A and 79B which are support base position detection devices.

  The motors 78A and 78B and the encoders 79A and 79B are electrically connected to the control unit 20 in the computer 7. The control unit 20 controls driving of the motors 78A and 78B based on the position information of the support bases 3A and 3B output from the encoders 79A and 79B. The vertical movement devices 76A and 76B also function as the vertical movement mechanisms 5A and 5B.

  In the case of such a configuration, if the vertical movement devices 76A and 76B are provided to move the log 100 to the temporary position, it is also used as a moving mechanism of the support bases 3A and 3B for creating a parallax map. be able to. Accordingly, the outer edge 63 of the butt end can be detected without newly installing the left and right moving devices 4A and 4B.

  Here, although the vertical movement mechanisms 76A and 76B have a screw-type configuration, the present invention is not limited to this. For example, a method using a hydraulic servo cylinder, a crank mechanism, and a drive servo motor may be used. In short, any mechanism that can move the support bases 3A and 3B in the Y-axis direction while detecting their positions may be used.

  In addition, although the encoders 79A and 79B are used as the support base position detection devices and the configuration in which the positions of the support bases 3A and 3B can be permanently grasped is shown, the present invention is not limited to this. For example, instead of the encoders 79A and 79B, a mechanical switch such as a limit switch is installed in contact with the support bases 3A and 3B, and the support base 3A is moved to the first position and the second position by a signal from the mechanical switch. , 3B may be stopped.

  Moreover, in the said embodiment, although it was set as the structure which images the wooden mouths 101A and 101B before and behind the movement of the support bases 3A and 3B with the cameras 1A and 1B of each one side of the log 100, this invention is limited to this. Not. For example, a method may be used in which two cameras are arranged side by side on both ends of the log 100 toward the ends 101A and 101B, and the first image and the second image obtained by simultaneously capturing the two cameras are processed. .

  The two cameras have their optical axes arranged in parallel so that the distance between the two optical axes is the same as the movement distance s in the above embodiment. In this case, the direction in which the disparity map creating unit 13 searches for the corresponding point for calculating the similarity R is the direction from the optical axis of one camera to the optical axis of the other camera. The optical axis of the camera does not necessarily have to be parallel. That is, if the angle between the optical axes is known, the parallax can be obtained.

  In such a configuration, since the first image and the second image are simultaneously captured, it is not necessary to move the log 100 by the left and right moving devices 4A and 4B and stop it twice, thereby improving work efficiency.

  Moreover, although the said embodiment demonstrated the example which moves the log 100 with support stand 3A, 3B, you may make it move camera 1A, 1B conversely. In short, the configuration may be such that the log 100 and the cameras 1A and 1B are relatively provided with a moving device that moves by an arbitrary distance s.

  In the above embodiment, the method of calculating the parallax d by moving the log 100 from the first position to the second position by an arbitrary distance s and performing image processing on the two images before and after the movement is shown. The present invention is not limited to this. That is, it is sufficient to move the log 100 by a distance that produces a significant difference in obtaining the parallax d, and to capture two images, and it is not always necessary to move the log 100 by the moving distance s and stop for imaging. Absent.

  That is, since the value of the movement distance s is unnecessary in the process of detecting the outer edge 63 of the lip and the process of calculating the distance from the cameras 1A and 1B to the lip 101A and 101B, it is necessary to grasp the value of the movement distance s. There is no. Therefore, for example, the tree mouths 101A and 101B of the log 100 moving in the Y-axis direction are imaged twice at a certain time interval while moving without stopping, and image processing is performed on the two captured images. By doing this, the outer peripheral edge 63 may be detected.

  For example, as shown in FIG. 15, the vertical movement mechanisms 76A and 76B are configured. The vertical movement mechanisms 76A and 76B move the support bases 3A and 3B in the Y-axis direction at a predetermined speed v (mm / second) by the motors 78A and 78B in accordance with a movement command signal from the control unit 20 in the computer 7.

  Based on the signals output from the encoders 79A and 79B during the movement, the control unit 20 determines that the positions of the support bases 3A and 3B are the first positions (the temporary ends in which the wood ends 101A and 101B of the log 100 are located at the center of the imaging surface). When it is detected that the center position has been reached, the cameras 1A and 1B are controlled to image the lips 101A and 101B.

  After the first imaging, the control unit 20 detects that a predetermined time t (seconds) has elapsed, controls the cameras 1A and 1B, and causes the lips 101A and 101B to image again. An image captured at the first position is defined as a first image, and an image captured when a predetermined time t (seconds) elapses after the first image is captured is defined as a second image.

  Here, the moving speed v (mm / second) and the predetermined time t (second) of the support bases 3A and 3B (log 100) will be described. The distance v × t (mm) by which the support bases 3A and 3B move during the imaging of the first image and the second image is a parallax between the two images, similarly to the movement distance s described in the above embodiment. Any distance that produces a significant difference can be obtained, and there is no particular limitation as long as the images of the lips 101A and 101B are within the range of the imaging surface.

  For example, if the mouths 101A and 101B of the log 100 moving at a speed of 30 mm / sec are imaged at 1 second intervals, the moving distance of the log 100 between the two imaging operations is 30 mm. If the desired speed v is increased, the working efficiency is improved, but the sharpness of the image is inferior, and there is a possibility that the required parallax accuracy may be problematic. It is necessary to set a desired speed v (mm / second) and a predetermined time t (second) in consideration of the brightness of illumination by the illumination devices 2A and 2B, the resolution of the cameras 1A and 1B, the exposure time, the imaging processing time, and the like. There is.

  When the first image and the second image are captured at two timings while moving the log 100, a moving image captured by a video camera or the like instead of capturing two still images by the cameras 1A and 1B. Alternatively, two frames may be extracted from and processed as a first image and a second image, respectively. Specifically, when the support bases 3A and 3B reach the first position (temporary center position), a frame is extracted from the moving image as a first image, and a frame after a predetermined time t (seconds) is extracted therefrom. Two images may be used.

  In addition, each of the above-described embodiments is merely an example of implementation in carrying out the present invention, and the technical scope of the present invention should not be construed in a limited manner. In other words, the present invention can be implemented in various forms without departing from the spirit or main features thereof.

DESCRIPTION OF SYMBOLS 1A, 1B Camera 3A, 3B Support stand 4A, 4B Left-right moving apparatus 7 Computer 10 Image processor 11 Image acquisition part 12 Moving object area extraction part 13 Parallax map creation part 14 Kipple area specific | specification part 15 Kiple outer periphery edge specific part 16 Calculation unit 20 Control unit 30 Image memory 50 Moving object area 60 Central zone 61 Kip area 63 Kip outer peripheral edge 100 Log 101A, 101B Kguchi

Claims (5)

Two images captured with respect to the mouth of the raw wood by the imaging device in a state where the relative position of the raw wood with respect to the imaging device is different in a direction orthogonal to the optical axis of the imaging device are acquired as a first image and a second image, respectively. An image acquisition unit to
For each pixel of the first image, a parallax representing the displacement amount of the corresponding pixel between the first image and the second image is calculated, and a parallax map representing the parallax as a pixel value of each pixel is created. A parallax map creation unit for
In the parallax map created by the parallax map creating unit, a lip end specifying unit that specifies a region of a pixel in which a pixel value in a predetermined central area or a difference between the average value and the pixel value falls within a predetermined range as a lip end region When,
An apparatus for detecting an edge of an edge of an original lip, comprising: a lip edge specifying part for specifying a lip edge of the lip area specified by the lip area specifying part as a lip edge. Moving to extract the area on the image of the object whose position is moving between the first image and the second image as a moving object area by taking the difference between the first image and the second image An object area extraction unit;
The parallax map creation unit calculates, for each pixel in the moving object area of the first image, a parallax representing a displacement amount of a corresponding pixel between the first image and the second image, and the parallax 2. The apparatus according to claim 1, wherein a parallax map is created in which each pixel is represented as a pixel value of each pixel. Moving to extract the area on the image of the object whose position is moving between the first image and the second image as a moving object area by taking the difference between the first image and the second image An object area extraction unit;
In the moving object area of the parallax map created for the entire area of the first image by the parallax map creating unit, the lip end specifying unit is configured to calculate a pixel value of a predetermined central area or an average value and a pixel value thereof. 2. The apparatus according to claim 1, wherein an area of a pixel whose difference falls within a predetermined range is specified as a lip end area. Moving to extract the area on the image of the object whose position is moving between the first image and the second image as a moving object area by taking the difference between the first image and the second image An object area extraction unit;
The lip outer edge specifying unit specifies, as the lip outer edge, an outer edge of the lip region specified in the moving object area among the lip region specified by the lip region specifying unit. Item 2. A device for detecting an edge of an edge of a wood log according to Item 1. Two images captured with respect to the mouth of the raw wood by the imaging device in a state where the relative position of the raw wood with respect to the imaging device is different in a direction orthogonal to the optical axis of the imaging device are acquired as a first image and a second image, respectively. A first step to:
For each pixel of the first image, a parallax representing the displacement amount of the corresponding pixel between the first image and the second image is calculated, and a parallax map representing the parallax as a pixel value of each pixel is created. A second step of:
In the parallax map created in the second step, a third step of specifying, as a lip end region, a pixel area in which a pixel value in a predetermined central area or a difference between the average value and the pixel value is within a predetermined range When,
And a fourth step of identifying the outer edge of the lip region identified in the third step as a lip outer edge.

Images