CN112129773B - Wood surface defect detection method, device, equipment, system and storage medium - Google Patents

Abstract

The invention discloses a wood surface defect detection method, device, equipment, system and storage medium. Wherein, the method includes: acquiring a laser line image of a conveyor belt under the action of a laser line generated by a laser device; extracting a laser line height position and a laser line width of the laser line based on the acquired laser line image; based on the laser line Line height position and/or laser line width and preset laser line model, identify whether there is wood on the conveyor belt; if there is wood on the conveyor belt, identify the laser line based on the laser line height position and laser line width of the laser line. surface defects of the wood. Due to the tracheid effect on the surface of the wood, the surface defects of the wood can be accurately identified based on the height position of the laser line and the width of the laser line. To meet the identification needs of rapid identification, it is suitable for the field of automatic sawing of wood surface defects.

Description

Wood surface defect detection method, device, equipment, system and storage medium

technical field

The invention relates to the field of wood detection, in particular to a wood surface defect detection method, device, equipment, system and storage medium.

Background technique

Solid wood board is a decorative material formed by drying and processing natural wood. It has the advantages of natural beauty, safety, environmental protection and durability, and electrical and thermal insulation. It has been widely used in furniture manufacturing, indoor and outdoor decoration, etc. In wood science and engineering, the quality grade of wood determines its usefulness in production applications. And wood surface defects such as dead knots, joints, holes, bug eyes, etc. not only affect the aesthetics of the finished wood in the wood industry, but also affect the quality of the wood itself and its load-bearing capacity. Therefore, the detection and selection of wood surface defects is an important process in wood processing. one.

In the related art, wood defect detection methods are mainly based on X-rays, stress waves, ultrasonic waves, infrared, lasers, optical cameras, and the like. In addition, digital image processing technology, computer vision technology, and pattern recognition technology can also be applied to the detection of wood defects, and the fractal theory, wavelet multi-resolution analysis and artificial neural network pattern recognition technology can be combined to study the texture of wood surface defects. Segmentation, feature extraction, pattern recognition and other issues form a new class of detection methods.

With the rapid development of high-end solid wood processing machinery and equipment, automatic identification of wood surface defects and automatic sawing have been widely used in the field of automation industry, but there are generally problems such as low precision or slow speed. In the related art, the detection and positioning of the defect area is realized by the difference in the physical characteristics of the normal area and the abnormal area on the surface of the wood. The need for rapidity, but the lack of precision. In addition, an optical camera can be used to collect pictures, and a high-accuracy image processing algorithm can be designed to detect and locate wood surface defects. Although it has the advantages of high accuracy, it often ignores the needs of industrial rapidity, and the algorithm requires a large amount of calculation. Time consuming.

SUMMARY OF THE INVENTION

In view of this, the embodiments of the present invention provide a wood surface defect detection method, device, equipment, system and storage medium, aiming at improving the defect identification efficiency on the basis of satisfying the defect identification accuracy.

The technical solution of the embodiment of the present invention is realized as follows:

In a first aspect, an embodiment of the present invention provides a method for detecting surface defects of wood, including:

acquiring a laser line image of the conveyor belt under the action of a laser line generated by a laser device, the laser line extending along the width direction of the conveyor belt on a horizontal plane;

Based on the acquired laser line image, extract the laser line height position and the laser line width of the laser line;

Identifying whether there is wood on the conveyor belt based on the laser line height position and/or the laser line width of the laser line and a preset laser line model;

If there is wood on the conveyor belt, the surface defects of the wood are identified based on the laser line height position and the laser line width of the laser line.

In some embodiments, the identifying whether there is wood on the conveyor belt based on the laser line height and/or laser line width of the laser line and a preset laser line model includes:

Compare the laser line height position and laser line width of the laser line with the preset first laser line model and second laser line model to obtain a first similarity degree and a second similarity degree, and the first similarity degree represents similarity between the laser line and the first laser line model, and the second similarity represents the similarity between the laser line and the second laser line model;

It is determined that the first similarity is greater than the second similarity, and it is determined that there is wood on the conveyor belt; or,

It is determined that the laser line width of the laser line exceeds the scattering width value of the laser line in the second laser line model and reaches a first set threshold, and it is determined that there is wood on the conveyor belt; or,

It is determined that the height position of the laser line of the laser line exceeds the height position of the laser line in the second laser line model and reaches the second set threshold, and it is determined that there is wood on the conveyor belt;

Wherein, the first laser line model represents the height position and scattering width value of the laser line acting on the surface of the wood on the conveyor belt, and the second laser line model represents the height position of the laser line acting on the surface of the conveyor belt and the scattering width value.

In some embodiments, the method further includes:

acquiring a first laser line image of the surface of the wood on the conveyor belt under the action of the laser line generated by the laser device, and constructing the first laser line model based on the first laser line image;

A second laser line image of the surface of the conveyor belt under the action of the laser line generated by the laser device is acquired, and the second laser line model is constructed based on the second laser line image.

In some embodiments, the identifying surface defects of the wood based on the laser line height position and the laser line width of the laser line includes:

constructing a relative height image of the wood based on the difference between the laser line height position of the laser line and the height position in the first laser line model;

constructing a relative scattering image of the wood based on the difference between the laser line width of the laser line and the scattering width value in the first laser line model;

determining whether the wood has a first defect based on the relative height image;

determining whether the wood has a second defect based on the relative scattering image;

Wherein, the first defect includes at least one of the following: missing edge and worm's eye; the second defect includes at least one of the following: dead joint and live joint.

In some embodiments, the constructing the first laser line model based on the first laser line image includes:

acquiring multiple frames of the first laser line images;

determining the height position of the laser line on the wood based on the multiple frames of the first laser line image;

determining a scattering width value of a laser line on the wood based on a plurality of frames of the first laser line images;

The constructing the second laser line model based on the second laser line image includes:

acquiring multiple frames of the second laser line images;

determining the height position of the laser line on the conveyor belt based on the plurality of frames of the second laser line images;

A scattering width value of the laser line on the conveyor belt is determined based on a plurality of frames of the second laser line images.

In some embodiments, constructing the relative height image of the wood based on the difference between the laser line height position of the laser line and the height position in the first laser line model includes:

For the acquired multiple frames of the laser line images, construct a relative height image on the corresponding area of the wood, wherein the value of each pixel in the relative height image is based on the height position of the laser line and the first The difference in height position in the laser line model is determined.

In some embodiments, constructing a relative scattering image of the wood based on a difference between a laser line width of the laser line and a scattering width value in the first laser line model includes:

For the acquired multiple frames of the laser line images, a relative scattering image on the corresponding area of the wood is constructed, wherein the value of each pixel in the relative scattering image is based on the width of the laser line and the first laser beam. The difference in the scattering width values in the line model is determined.

In some embodiments, the method further includes:

If the first defect exists, determine the defect position corresponding to the first defect based on the relative height image; and/or,

If the second defect exists, a defect position corresponding to the second defect is determined based on the relative scattering image.

In a second aspect, an embodiment of the present invention also provides a wood surface defect detection device, including:

an acquisition module for acquiring a laser line image of the conveyor belt under the action of a laser line generated by a laser device, the laser line extending along the width direction of the conveyor belt on a horizontal plane;

a feature extraction module, configured to extract the laser line height position and the laser line width of the laser line based on the acquired laser line image;

a first identification module for identifying whether there is wood on the conveyor belt based on the laser line height position and/or the laser line width of the laser line and a preset laser line model;

The second identification module is configured to identify the surface defect of the wood based on the height position of the laser line and the width of the laser line if there is wood on the conveyor belt.

In a third aspect, an embodiment of the present invention further provides a wood surface defect detection device, comprising: a processor and a memory for storing a computer program that can be run on the processor, wherein the processor is used to run a computer During the program, the steps of the methods described in the embodiments of the present invention are executed.

In a fourth aspect, an embodiment of the present invention also provides a wood surface defect detection system, including:

Conveyor belts for the online transfer of wood;

a laser device for generating a laser line acting on the wood;

Image acquisition equipment for acquiring laser line images;

The wood surface defect detection device according to the embodiment of the present invention is connected to the image acquisition device, and receives the laser line image acquired by the image acquisition device.

In a fifth aspect, an embodiment of the present invention further provides a storage medium, where a computer program is stored on the storage medium, and when the computer program is executed by a processor, the steps of the method described in the embodiment of the present invention are implemented.

The technical solution provided by the embodiment of the present invention is to obtain the laser line image of the conveyor belt under the action of the laser line generated by the laser equipment, and extract the laser line height position and laser line width of the laser line in the laser line image. Laser line height position and/or laser line width and preset laser line model, identify whether there is wood on the conveyor belt, if there is wood on the conveyor belt, identify based on the laser line height position and laser line width of the laser line The surface defects of the wood; due to the tracheid effect on the surface of the wood, the surface defects of the wood can be accurately identified based on the height position of the laser line and the width of the laser line, and can support missing edges, bug eyes, dead knots, live knots, etc. Accurate identification of defects, and can meet the identification needs of rapid identification, so it is suitable for the field of automatic sawing of wood surface defects.

Description of drawings

Fig. 1 is the schematic flow chart of the wood surface defect detection method of the embodiment of the present invention;

2 is a schematic structural diagram of a wood surface defect detection system according to an embodiment of the present invention;

3 is a schematic diagram of the principle that the reconstructed image is consistent with the actual proportion of the wood in the embodiment of the present invention;

4 is a schematic diagram of the tracheid effect of the laser line in the normal area of the wood surface according to the embodiment of the present invention;

5 is a schematic diagram of a laser line image when the laser line scans the wood with missing edges according to an embodiment of the present invention;

6 is a schematic diagram of a laser line image of a laser line in the wood knot sub-region according to an embodiment of the present invention;

7 is a schematic structural diagram of a wood surface defect detection device according to an embodiment of the present invention;

FIG. 8 is a schematic structural diagram of a wood surface defect detection device according to an embodiment of the present invention.

Detailed ways

The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments.

Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terms used herein in the description of the present invention are for the purpose of describing specific embodiments only, and are not intended to limit the present invention.

Before introducing the wood surface defect detection method according to the embodiment of the present invention, the tracheid effect involved in the embodiment of the present invention is explained. The embodiments of the present invention detect wood surface defects based on the tracheid effect of laser lines on the wood surface. The tracheid effect is an observable optical phenomenon. When the laser line is projected on the surface of wood, part of the light directly shines on the surface, and part of the light penetrates the surface and is conducted in the wood of semi-mirror material. Softwood fibers (ie, tracheids) conduct light better, while defects such as wood knots conduct light differently than normal areas. The embodiments of the present invention utilize this feature to detect wood defect areas by combining a laser device with an image detection device (such as an array camera), thereby improving the accuracy and speed of detection of wood surface defects, which is of great significance to industrial applications.

The embodiment of the present invention provides a wood surface defect detection method, as shown in FIG. 1 , the method includes:

Step 101, acquiring a laser line image of a conveyor belt under the action of a laser line generated by a laser device, the laser line extending along the width direction of the conveyor belt on a horizontal plane;

Exemplarily, in this embodiment of the present invention, a laser device and an image acquisition device may be provided near the conveyor belt, the image acquisition device is connected to a wood surface defect detection device, and the wood surface defect detection device implements the wood surface defect detection method of the embodiment of the present invention.

As shown in FIG. 2 , in an application example, the wood surface defect detection system includes: a conveyor belt 201 , a laser device 202 , an image acquisition device 203 and a wood surface defect detection device 204 . The laser device 202 may be a line laser, the image acquisition device 203 may be an array camera, and the wood surface defect detection device 204 may be a processing device with image data processing capability. The line laser can be located at the lower left of the array camera shown in Figure 2, above the conveyor belt, and the inclined angle with the horizontal plane is about 60°; the array camera is located half a meter above the conveyor belt, and its field of view can cover the entire width of the conveyor. , to ensure that the wood on the conveyor belt can be completely photographed.

In practical applications, the AOI (Automated Optical Inspection, automatic optical inspection) parameters of the image acquisition device 203 can be set to make the image resolution as low as possible under the condition that the laser line can be completely photographed, that is, the image is as small as possible to ensure that the The image processing and data analysis of each frame of image is smaller and the calculation speed is faster.

In this embodiment of the present invention, the direction of the laser line generated by the laser device 202 is perpendicular to the longitudinal direction of the wood (that is, the growth direction) on the horizontal plane, and the image acquisition device 203 collects the image of the laser line based on the set camera frame rate and transmits it to the surface defects of the wood The detection device 204, in this way, the wood surface defect detection device 204 can acquire the corresponding laser line images during the working process of the conveyor belt in real time online.

Step 102, extracting the laser line height position and the laser line width of the laser line based on the obtained laser line image;

Here, the wood surface defect detection device 204 performs feature extraction corresponding to the acquired laser line image, and extracts the laser line height position and the laser line width of the laser line.

Step 103, based on the laser line height position and/or the laser line width of the laser line and the preset laser line model, identify whether there is wood on the conveyor belt;

Here, the wood surface defect detection device 204 can use a pre-built laser line model to judge the laser line height position and/or the laser line width of the extracted laser line, identify whether there is wood on the conveyor belt, and determine whether there is wood on the conveyor belt. After that, step 104 is executed.

Step 104 , if there is wood on the conveyor belt, identify the surface defects of the wood based on the height position of the laser line and the width of the laser line.

Here, the wood surface defect detection device 204 uses the multiple frames of laser line images obtained online, and based on the laser line height positions and laser line widths of the laser lines extracted from each frame of the extracted laser line images, it can reconstruct the information representing the wood surface height information. Relative height image and relative scattering image of wood surface structure information, and then realize the detection and localization of surface defects. Due to the tracheid effect on the surface of the wood, the surface defects of the wood can be accurately identified based on the height position of the laser line and the width of the laser line. To meet the identification needs of rapid identification, it is suitable for the field of automatic sawing of wood surface defects.

In practical applications, the laser line height position and laser line width of the extracted laser lines need to be used to construct a relative height image representing the height information of the wood surface and a relative scattering image of the wood surface structure information, so as to realize the detection and localization of surface defects. In the actual industrial sawing task, it is necessary to provide accurate position coordinates for the sawing equipment, so the aspect ratio of the reconstruction map must be consistent with the actual wood. In the process of reconstructing the relative scattering map and relative height map of wood, the pixel value of each line of pixel points in the reconstructed image is assigned by the information contained in each frame of laser line image, that is, one frame of laser line image can reconstruct one or more lines. Timber reconstruction drawing. When one frame of laser line image reconstructs one row of pixel values, the reconstructed image has the highest resolution; when one frame of laser line image information reconstructs multiple rows of images, the reconstructed image resolution becomes lower. When the wood is transported on the conveyor belt, it passes through the image acquisition device at a certain speed. The image acquisition device shoots at a certain frame rate and stores it in the computer memory, and the computer processes the stored image with another thread. When the shooting frame rate is fixed, the faster the conveyor belt rate is, the shorter the time for the wood to pass through the field of view of the image acquisition device, and the less the number of shooting frames; the slower the conveyor belt rate, the longer the wood passes through the field of view of the image acquisition equipment. , the more frames you shoot. Therefore, the resolution of the reconstructed image is related to the artificially set number of reconstructed image lines (image reconstruction step size) corresponding to one frame of image, the conveyor belt rate, and the shooting frame rate.

Based on this, in order to satisfy the ratio between the width and height of the image in the relative height image and the relative scattering image and the ratio of the width and height of the real wood, and the resolution of the reconstructed image is reasonable, the speed and image acquisition of the conveyor belt 201 need to be The shooting frame rate of the device 203 is appropriately set.

Exemplarily, as shown in Figure 3, l and h are the width and height of the real wood, respectively, and x and y are the width and height of the wood part in the reconstructed image, respectively. Keeping the ratio of the reconstructed image and the real wood to be consistent, it can be obtained:

Among them, the height value y of the reconstructed image is determined by the line laser image under the field of view of the image acquisition device, and the red scattering field excited by the line laser on the wood and the conveyor belt is obviously different. The data calculates the height y of the wood within the field of view. It can be seen from the above that the height value y of the reconstructed wood image is only related to the height value h of the real wood object and the field of view of the image acquisition device. The reconstructed image width x is related to the preset number of reconstructed image lines (reconstruction step size) corresponding to one frame of image and the total number of frames of images collected by the image acquisition device during the process of the wood passing through the range of the image acquisition device, namely:

x=d·λ

Among them, d is the image reconstruction step size (that is, the number of reconstructed image lines corresponding to one frame of image), and λ is the total number of frames of laser line images collected during the process of uniformly passing through the field of view of the image acquisition device for the wood to be inspected. The relationship between λ and the shooting frame rate of the image acquisition device and the conveyor belt rate is as follows:

Among them, l is the actual width of the wood, v is the speed of the conveyor belt, and _fps is the shooting frame rate of the image acquisition device.

Finally get:

Therefore, it is possible to measure the actual height h of a piece of wood and the height y of the reconstructed image determined by the field of view of the camera, and adjust the shooting frame rate of the image acquisition device to correspond to the number of reconstructed image lines (step length) and the speed of the conveyor belt for one frame of image. Control the aspect ratio of the reconstructed image to be consistent with the actual wood aspect ratio, so as to provide accurate defect coordinates for the wood cutting and sawing equipment in the future.

In some embodiments, the pre-built laser line models include: a first laser line model and a second laser line model. The first laser line model represents the height position and scattering width value of the laser line acting on the surface of the wood on the conveyor belt, and the second laser line model represents the height position and the scattering width of the laser line acting on the surface of the conveyor belt. Scatter width value.

In some embodiments, pre-building the laser line model includes:

acquiring a first laser line image of the surface of the wood on the conveyor belt under the action of the laser line generated by the laser device, and constructing the first laser line model based on the first laser line image;

A second laser line image of the surface of the conveyor belt under the action of the laser line generated by the laser device is acquired, and the second laser line model is constructed based on the second laser line image.

In some embodiments, the constructing the first laser line model based on the first laser line image includes:

acquiring multiple frames of the first laser line images;

determining the height position of the laser line on the wood based on the multiple frames of the first laser line image;

determining a scattering width value of a laser line on the wood based on a plurality of frames of the first laser line images;

The constructing the second laser line model based on the second laser line image includes:

acquiring multiple frames of the second laser line images;

determining the height position of the laser line on the conveyor belt based on the plurality of frames of the second laser line images;

A scattering width value of the laser line on the conveyor belt is determined based on a plurality of frames of the second laser line images.

Exemplarily, standard wood (ie, wood without any defects) can be placed on the conveyor belt, and the first laser line model can be constructed by using the first laser line image acquired in this scene. Figure 4 shows a schematic diagram of the tracheid effect of the laser line in the normal area of the wood surface. The laser line is normally scattered in the normal area of the wood, forming a light spot that is bright in the middle and gradually scattered to both sides. The information of the brightest laser line in the middle can be obtained from the green channel of the image, while the red part scattered outward due to the tracheid effect is located in the red channel of the image.

Here, the grayscale image imgR is constructed based on the red channel of the first laser line image, the grayscale image imgG is constructed based on the green channel of the first laser line image, and the scattering width model of the first laser line model can be constructed based on the grayscale image imgR, based on The grayscale image imgG constructs a height model of the first laser line model.

Exemplarily, a Gaussian distribution model can be used to establish a gray value distribution model of the red channel in the laser line image:

Among them, x is the gray value of the image imgR pixel point, f(x) is the probability that the gray value belongs to the laser line bright spot and its scattering area, μ and σ are the pixel gray value of the laser line scattering area in the N images imgR, respectively The average value and variance of the intensity value are used to establish the gray value distribution model of the red channel of the laser line.

Based on the gray value distribution model of the red channel in the first laser line image, if the gray probability f(x) of a certain pixel point is greater than a certain threshold Th, the pixel point belongs to the laser line scattering area (that is, belongs to the laser line within a given threshold range) The pixels in the brighter area in the center are counted), where the probability threshold Th is in the interval [0.5, 0.8]. Each pixel in the area is classified and counted, and the total number of pixels belonging to the laser line scattering area is counted in units of one column of one frame of image, and the sum of the number is used as the laser line width value. For example, the average value of the laser line width values in the N first laser line images is taken as the scattering width value of the first laser line model.

Exemplarily, a grayscale image imgG may be constructed based on the green channel in the first laser line image, a fixed threshold is set to perform binarization processing on imgG, and the threshold range may be set to a value in the [180, 230] interval. For each frame of image, the pixel points of the laser line greater than the threshold are counted in column units, and the height position coordinate of the center is calculated, and the coordinate is used as the height coordinate value of the center position of the laser line. Take N second laser line images, and calculate the average height position of the laser line center as the height position of the first laser line model. Among them, N is a natural number greater than 1, which can be reasonably selected based on the image resolution.

It can be understood that the construction process of the second laser line model is similar to the construction process of the aforementioned first laser line model, the difference is that the second laser line image obtained in the scene where no wood is placed on the conveyor belt is required, The second laser line model is constructed based on the second laser line image. For details, refer to the foregoing generation process of the first laser line model, which will not be repeated here.

In some embodiments, the difference between the height position of the first laser line model and the height position of the second laser line model can also be obtained as the standard deviation ε between the laser line acting on the wood surface and the laser line acting on the conveyor belt surface .

In some embodiments, the identifying whether there is wood on the conveyor belt based on the laser line height and/or laser line width of the laser line and a preset laser line model includes:

Compare the laser line height position and laser line width of the laser line with the preset first laser line model and second laser line model to obtain a first similarity degree and a second similarity degree, and the first similarity degree represents similarity between the laser line and the first laser line model, and the second similarity represents the similarity between the laser line and the second laser line model;

It is determined that the first similarity is greater than the second similarity, and it is determined that there is wood on the conveyor belt.

In some embodiments, the identifying whether there is wood on the conveyor belt based on the laser line height and/or laser line width of the laser line and a preset laser line model includes:

It is determined that the laser line width of the laser line exceeds the scattering width value of the laser line in the second laser line model and reaches a first set threshold, and it is determined that there is wood on the conveyor belt.

Exemplarily, several columns of each frame of laser line image can be extracted, the laser line width of the extracted column can be calculated one by one, and compared with the scattering width value of the second laser line model, and then the existence of wood can be judged, for example, it is uniform from left to right. Take 5 columns and calculate the laser line width. If the laser line width is greater than 1.5 times the scattering width value of the second laser line model, it is considered that there is wood.

In some embodiments, the identifying whether there is wood on the conveyor belt based on the laser line height and/or laser line width of the laser line and a preset laser line model includes:

It is determined that the height position of the laser line of the laser line exceeds the height position of the surface of the conveyor belt in the second laser line model and reaches a second set threshold, and it is determined that there is wood on the conveyor belt.

Exemplarily, several columns of each frame of laser line images can be extracted, for example, 5 columns are taken uniformly from left to right, the average value of the height positions of the centers of the extracted laser lines is calculated, and the average value of the height positions is combined with the second laser line. The height position of the surface of the conveyor belt in the line model is compared to determine whether the wood exists. If the difference between the average value of the height positions of the drawn columns and the height position of the surface of the conveyor belt in the second laser line model is close to 0, then It is determined that there is no wood, otherwise, it is determined that there is wood.

In some embodiments, constructing the relative height image of the wood based on the difference between the laser line height position of the laser line and the height position in the first laser line model includes:

For the acquired multiple frames of the laser line images (that is, the multiple frames of laser line images acquired in real time during the operation of the conveyor belt), construct a relative height image on the area corresponding to the wood, wherein each pixel in the relative height image The value of the point is determined based on the difference between the height position of the laser line and the height position in the first laser line model.

每一帧激光图像提取的像素值向量共有width个元素，将每个元素标准化映射到[0，255]的区间上，并依次为重建相对高度图对应一行或多行的width个像素点赋值，若干帧激光图像依次提取像素值向量并依次为重建图像赋值即可重建出包含了相对高度信息的相对高度图。显然，当激光扫描缺边区域时，激光线中心位置与在传送带上的激光线标准位置更为接近，偏差值较小，ε_i＜ε，在重建图像上表现较黑。Exemplarily, FIG. 5 is a schematic diagram of a laser line image when the laser line scans the missing edge of the wood. When scanning the missing edge area of the wood, the laser line is offset from the normal area of the wood. Based on the first laser line model, threshold segmentation is performed on the pixels in the green channel grayscale image of each frame of laser line image to obtain a binary image of the spot position of the green channel laser line. Threshold segmentation is performed on the pixel points of the green channel to obtain a binary image of the spot position of the laser line in the green channel. The height position of the center of each column of laser lines in the binary image is counted to obtain the relationship between each column of laser lines in each frame of the image and the standard laser line. The deviation information ε _i of the line height position, i is the number of laser line image columns per frame. The wood image was reconstructed with the standard deviation ε of the laser line between the wood and the conveyor belt as the standard value. If the image resolution is set to H×width, the pixel value vector can be calculated for each frame of the laser image

The pixel value vector extracted from each frame of laser image has a total of width elements, and each element is standardized and mapped to the interval of [0, 255], and then assigns the width pixels corresponding to one or more lines of the reconstructed relative height map in turn, A relative height map containing relative height information can be reconstructed by sequentially extracting pixel value vectors from several frames of laser images and assigning values to the reconstructed images in turn. Obviously, when the laser scans the missing edge area, the center position of the laser line is closer to the standard position of the laser line on the conveyor belt, the deviation value is small, ε _i <ε, and the reconstructed image appears darker.

Here, based on the relative height image, it may be determined whether the wood has a first defect; wherein, the first defect includes at least one of the following: missing edge and bug eye.

In some embodiments, constructing a relative scattering image of the wood based on a difference between a laser line width of the laser line and a scattering width value in the first laser line model includes:

For the acquired multiple frames of the laser line images (that is, the multiple frames of laser line images acquired in real time during the working process of the conveyor belt), construct a relative scattering image on the area corresponding to the wood, wherein each pixel in the relative scattering image The value of the point is determined based on the difference between the laser line width and the scattering width value in the first laser line model.

每一帧激光图像提取的像素值向量共有width个元素，将每个元素标准化映射到[0，255]的区间上，并依次为重建散射图对应一行或多行的width个像素点赋值，若干帧激光图像依次提取像素值向量并依次为重建图像赋值即可重建出包含了木材缺陷先验信息的木材散射图像。显然，当激光扫描节子或其他缺陷区域时，激光线宽ω＜ω_μ，像素值较小，在重建图像上表现较黑；当激光线宽ω＞ω_μ时，认为激光线处于正常区域，重建图像对应像素点赋值标准化映射区间最大值255。Figure 6 shows the schematic diagram of the laser line image of the laser line in the wood knot sub-region. In the wood defect area, the scattering effect caused by the tracheid effect of the laser line is obviously different. In the node sub-region, the laser scattering ability is obviously weaker than that in the normal region, the area where the laser spot scatters to both sides becomes smaller, and the overall laser linewidth becomes narrower. Based on the aforementioned first laser line model, the pixels in the red channel grayscale image in each frame of laser line image are classified to obtain a binary image of the red channel laser linewidth. The pixel points are classified to obtain a binary image of the laser line width of the red channel, and the width of each column of laser lines in the binary image is counted to obtain the width information of each column of laser lines in each frame of image ω _i , i is the laser line of each frame The number of image columns. The wood image was reconstructed using the standard laser line width ω _u in the laser line scattering area as the standard value. If the image resolution is set to H×width, the pixel value vector can be calculated for each frame of the laser image

The pixel value vector extracted from each frame of laser image has a total of width elements, and each element is standardized and mapped to the interval of [0, 255], and then assigns the width pixels corresponding to one or more lines of the reconstructed scatter map in turn. The frame laser image extracts the pixel value vector in turn and assigns values to the reconstructed image in turn to reconstruct the wood scattering image containing the prior information of wood defects. Obviously, when the laser scans knots or other defect areas, the laser line width ω < ω _μ , the pixel value is small, and the reconstructed image appears darker; when the laser line width ω > ω _μ , the laser line is considered to be in the normal area , the corresponding pixels of the reconstructed image are assigned a maximum value of 255 in the normalized mapping interval.

Here, based on the relative scattering image, it may be determined whether the wood has a second defect; the second defect includes at least one of the following: dead joints and loose joints.

In some embodiments, the method further includes:

If the first defect exists, determine the defect position corresponding to the first defect based on the relative height image; and/or,

If the second defect exists, a defect position corresponding to the second defect is determined based on the relative scattering image.

In the embodiment of the present invention, since the ratio of the width to height in the relative height image and the relative scattering image is consistent with the ratio of the width and height of the real wood, the defect position can be determined on the relative height image or the relative scattering image, and It is converted to the corresponding position on the real wood, which is convenient for wire sawing of the defects on the wood.

Exemplarily, firstly, by reconstructing the relative height image, it is determined that the wood surface is uneven and has a certain height difference from the normal area, such as defects in the wood surface caused by missing edges, bug eyes, etc., and such defects have obvious features on the relative height map. For example, set a threshold to binarize the relative height map. The threshold value ranges between [30, 50]. The lower the gray value on the relative height map, the greater the deviation between the actual height value and the normal wood height. Large, that is, the greater the probability of defects such as missing edges and bug eyes in this area, set a reasonable threshold to deal with defects such as missing edges and bug eyes on the relative height map, and perform threshold segmentation. Results; then, by reconstructing the relative scatter map of the wood, other defects such as dead joints, live joints, etc. are located. The scatter map does not have the defect of height difference with the normal area, so it has a complementary effect with the relative height map. Set a threshold to binarize the scatter map. The threshold range can be between [30, 50]. After the threshold is divided The scattering binary map results can be used as the localization results of other wood defects.

From the above description, it can be known that the embodiment of the present invention improves the recognition accuracy and recognition speed of wood surface defects by combining machine vision and laser line scanning, and provides accurate information of defect positions for wood cutting and sawing equipment. By means of line laser scanning, we obtained the relative scattering image and the relative height image that the wood contains the defect information on the wood surface. The relative scattering image mainly reflects the dead knots, live knots and other defects on the surface of the wood that cause the defect itself to be different from the normal sheet. Structural information of defect types; the relative height image is not sensitive to most defect types that do not affect the flatness of the board, and can mainly assist in the determination of defects such as wood edge missing and insect eyes. In practical applications, firstly, the defects such as wood edge missing are located by the reconstructed relative height map, and then the threshold (or other classification method) segmentation method is used to segment the connected domain of the defect part of the relative scattering image. A large amount of surface information such as texture and color leaves only the required defect information, so a simple image segmentation method can achieve a good segmentation effect on the defect area. Combining the two reconstructed image information, the localization of wood surface defects can be completed.

In order to realize the method of the embodiment of the present invention, the embodiment of the present invention further provides a wood surface defect detection device, the wood surface defect detection device corresponds to the above-mentioned wood surface defect detection method, and each step in the above-mentioned embodiment of the wood surface defect detection method It is also fully applicable to this embodiment of the wood surface defect detection device.

As shown in FIG. 7 , the wood surface defect detection device 700 includes: an acquisition module 701 , a feature extraction module 702 , a first identification module 703 and a second identification module 704 , wherein the acquisition module 701 is used to acquire the conveyor belt generated by the laser equipment. The laser line image under the action of the laser line, the laser line extends along the width direction of the conveyor belt on the horizontal plane; the feature extraction module 702 is used for extracting the laser line height position of the laser line based on the acquired laser line image and laser line width; the first identification module 703 is used to identify whether there is wood on the conveyor belt based on the laser line height position of the laser line and/or the laser line width and the preset laser line model; the second identification module 704 If there is wood on the conveyor belt, the method is used to identify the surface defects of the wood based on the laser line height position and the laser line width of the laser line.

In some embodiments, the first identification module 703 is specifically used for:

Compare the laser line height position and laser line width of the laser line with the preset first laser line model and second laser line model to obtain a first similarity degree and a second similarity degree, and the first similarity degree represents similarity between the laser line and the first laser line model, and the second similarity represents the similarity between the laser line and the second laser line model;

It is determined that the first similarity is greater than the second similarity, and it is determined that there is wood on the conveyor belt; or,

It is determined that the laser line width of the laser line exceeds the scattering width value of the laser line in the second laser line model and reaches a first set threshold, and it is determined that there is wood on the conveyor belt; or,

It is determined that the height position of the laser line of the laser line exceeds the height position of the surface of the conveyor belt in the second laser line model and reaches a second set threshold, and it is determined that there is wood on the conveyor belt;

Wherein, the first laser line model represents the height position and scattering width of the laser line acting on the surface of the wood, and the second laser line model represents the height position and scattering width of the laser line acting on the surface of the conveyor belt value.

In some embodiments, it further includes: a construction module 705, configured to acquire a first laser line image of the surface of the wood on the conveyor belt under the action of the laser line generated by the laser device, and construct based on the first laser line image the first laser line model; and acquiring a second laser line image of the surface of the conveyor belt under the action of the laser line generated by the laser device, and constructing the second laser line model based on the second laser line image.

In some embodiments, the second identification module 704 is specifically used to:

constructing a relative height image of the wood based on the difference between the laser line height position of the laser line and the height position in the first laser line model;

constructing a relative scattering image of the wood based on the difference between the laser line width of the laser line and the scattering width value in the first laser line model;

determining whether the wood has a first defect based on the relative height image;

determining whether the wood has a second defect based on the relative scattering image;

Wherein, the first defect includes at least one of the following: missing edge and worm's eye; the second defect includes at least one of the following: dead joint and live joint.

In some embodiments, the building module 705 builds the first laser line model based on the first laser line image, including:

acquiring multiple frames of the first laser line images;

determining the height position of the laser line on the wood based on the multiple frames of the first laser line image;

determining a scattering width value of a laser line on the wood based on a plurality of frames of the first laser line images;

The building module 705 builds the second laser line model based on the second laser line image, including:

acquiring multiple frames of the second laser line images;

determining the height position of the laser line on the conveyor belt based on the plurality of frames of the second laser line images;

A scattering width value of the laser line on the conveyor belt is determined based on a plurality of frames of the second laser line images.

In some embodiments, the second identification module 704 is specifically used to:

For the acquired multiple frames of the laser line images, construct a relative height image on the corresponding area of the wood, wherein the value of each pixel in the relative height image is based on the height position of the laser line and the first The difference in height position in the laser line model is determined.

In some embodiments, the second identification module 704 is specifically used to:

For the acquired multiple frames of the laser line images, construct a relative height image on the corresponding area of the wood, wherein the value of each pixel in the relative height image is based on the height position of the laser line and the first The difference in height position in the laser line model is determined.

In some embodiments, the second identification module 704 is also used to:

If the first defect exists, determine the defect position corresponding to the first defect based on the relative height image; and/or,

If the second defect exists, a defect position corresponding to the second defect is determined based on the relative scattering image.

In practical application, the acquisition module 701 , the feature extraction module 702 , the first identification module 703 , the second identification module 704 and the construction module 705 may be implemented by a processor in the wood surface defect detection device. Of course, the processor needs to run a computer program in memory to perform its functions.

It should be noted that: when the wood surface defect detection device provided in the above-mentioned embodiment performs wood surface defect detection, only the division of the above program modules is used as an example for illustration. The program module is completed, that is, the internal structure of the device is divided into different program modules to complete all or part of the above-described processing. In addition, the wood surface defect detection device and the wood surface defect detection method embodiments provided by the above embodiments belong to the same concept, and the specific implementation process thereof is detailed in the method embodiments, which will not be repeated here.

Based on the hardware implementation of the above program modules, and in order to implement the method of the embodiment of the present invention, the embodiment of the present invention further provides a wood surface defect detection device. FIG. 8 only shows an exemplary structure of the wood surface defect detection apparatus but not the whole structure, and part or all of the structure shown in FIG. 8 may be implemented as required.

As shown in FIG. 8 , the wood surface defect detection device 800 provided by the embodiment of the present invention includes: at least one processor 801 , a memory 802 , a user interface 803 and at least one network interface 804 . The various components in the wood surface defect detection apparatus 800 are coupled together by a bus system 805 . It will be understood that the bus system 805 is used to implement the connection communication between these components. In addition to the data bus, the bus system 805 also includes a power bus, a control bus and a status signal bus. However, for the sake of clarity, the various buses are labeled as bus system 805 in FIG. 8 .

The user interface 803 may include a display, a keyboard, a mouse, a trackball, a click wheel, keys, buttons, a touch pad or a touch screen, and the like.

The memory 802 in the embodiment of the present invention is used to store various types of data to support the operation of the wood surface defect detection apparatus. Examples of such data include: any computer program used to operate on wood surface defect detection equipment.

The wood surface defect detection method disclosed in the embodiment of the present invention may be applied to the processor 801 or implemented by the processor 801 . The processor 801 may be an integrated circuit chip with signal processing capability. In the implementation process, each step of the wood surface defect detection method can be completed by the hardware integrated logic circuit in the processor 801 or the instructions in the form of software. The above-mentioned processor 801 may be a general-purpose processor, a digital signal processor (DSP, Digital Signal Processor), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, and the like. The processor 801 may implement or execute the methods, steps, and logical block diagrams disclosed in the embodiments of the present invention. A general purpose processor may be a microprocessor or any conventional processor or the like. The steps of the method disclosed in combination with the embodiments of the present invention can be directly embodied as being executed by a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software module may be located in a storage medium, and the storage medium is located in the memory 802, and the processor 801 reads the information in the memory 802, and completes the steps of the wood surface defect detection method provided by the embodiment of the present invention in combination with its hardware.

In an exemplary embodiment, the wood surface defect detection device may be implemented by one or more of Application Specific Integrated Circuit (ASIC, Application Specific Integrated Circuit), DSP, Programmable Logic Device (PLD, Programmable Logic Device), Complex Programmable Logic Device (CPLD, Complex Programmable Logic Device), FPGA, general-purpose processor, controller, microcontroller (MCU, Micro Controller Unit), microprocessor (Microprocessor), or other electronic component implementations for executing the aforementioned method.

It will be appreciated that the memory 802 may be either volatile memory or non-volatile memory, and may include both volatile and non-volatile memory. Among them, the non-volatile memory may be a read-only memory (ROM, Read Only Memory), a programmable read-only memory (PROM, Programmable Read-Only Memory), an erasable programmable read-only memory (EPROM, Erasable Programmable Read-only memory) Only Memory), Electrically Erasable Programmable Read-Only Memory (EEPROM, Electrically Erasable Programmable Read-Only Memory), Magnetic Random Access Memory (FRAM, ferromagnetic random access memory), Flash Memory (Flash Memory), Magnetic Surface Memory , CD-ROM, or Compact Disc Read-Only Memory (CD-ROM, Compact Disc Read-Only Memory); the magnetic surface memory can be a magnetic disk memory or a tape memory. The volatile memory may be Random Access Memory (RAM), which is used as an external cache memory. By way of example and not limitation, many forms of RAM are available, such as Static Random Access Memory (SRAM), Synchronous Static Random Access Memory (SSRAM), Dynamic Random Access Memory Memory (DRAM, Dynamic Random Access Memory), Synchronous Dynamic Random Access Memory (SDRAM, SynchronousDynamic Random Access Memory), Double Data Rate Synchronous Dynamic Random Access Memory (DDRSDRAM, Double Data Rate Synchronous Dynamic Random Access Memory), Enhanced Synchronous Dynamic Random Access Memory (ESDRAM, Enhanced Synchronous Dynamic Random Access Memory), Synchronous Link Dynamic Random Access Memory (SLDRAM, SyncLink Dynamic Random Access Memory), Direct Memory Bus Random Access Memory (DRRAM, Direct Rambus Random Access Memory) . The memory described in the embodiments of the present invention is intended to include, but not be limited to, these and any other suitable types of memory.

In some embodiments, embodiments of the present invention further provide a wood surface defect detection system, as shown in FIG. 2 , wherein the wood surface defect detection device 204 may be the aforementioned wood surface defect detection device 800 . The specific defect detection method Referring to the foregoing description, details are not repeated here.

In an exemplary embodiment, the embodiment of the present invention further provides a storage medium, that is, a computer storage medium, which may be a computer-readable storage medium, for example, including a memory 802 for storing a computer program, and the above-mentioned computer program can detect wood surface defects The processor 801 of the device executes the steps described in the method of the embodiment of the present invention. The computer-readable storage medium may be memory such as ROM, PROM, EPROM, EEPROM, Flash Memory, magnetic surface memory, optical disk, or CD-ROM.

It should be noted that "first", "second", etc. are used to distinguish similar objects, and are not necessarily used to describe a specific sequence or sequence.

In addition, the technical solutions described in the embodiments of the present invention may be combined arbitrarily if there is no conflict.

The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention. should be included within the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims (11)

1. A method for detecting surface defects of wood is characterized by comprising the following steps:

acquiring a laser line image of a conveyor belt under the action of a laser line generated by laser equipment, wherein the laser line extends along the width direction of the conveyor belt on a horizontal plane;

based on the obtained laser line image, extracting the laser line height position and the laser line width of the laser line;

wherein, based on the laser line image that obtains, extract laser line height position and laser line width of laser line, include:

constructing a gray level image according to the laser line image;

establishing a gray value distribution model according to the distribution model and the gray image;

obtaining the gray probability of the pixel points according to the gray value distribution model;

screening laser line pixel points according to the relation between the gray level probability and the probability threshold;

carrying out classified statistics on the number of the pixel points to obtain the laser line width;

calculating the pixel points by classification statistics to obtain the height position of the laser line;

identifying whether wood exists on the conveyor belt or not based on the height position and/or width of the laser line and the height position and scattering width value of a preset laser line model, wherein the scattering width value is obtained according to the widths of the laser lines;

wherein the preset laser line model comprises at least one of:

a first laser line model characterizing the height position and scattering width value of a laser line acting on the surface of the wood on the conveyor belt; the second laser line model is used for representing the height position and the scattering width value of the laser line acting on the surface of the conveyor belt;

if wood exists on the conveyor belt, identifying the surface defects of the wood based on the height position and the width of the laser line and the height position and the scattering width value of the preset laser line model;

wherein the identifying of the surface defects of the wood based on the laser line height position and the laser line width of the laser line comprises:

constructing a relative height image of the wood based on a difference between a laser line height position of the laser line and a height position in the first laser line model; determining whether the wood has a first defect based on the relative height image; wherein the first defect comprises at least one of: edge deletion and moth eye;

constructing a relative scattering image of the wood based on a difference value between the laser line width of the laser line and the scattering width value in the first laser line model; determining whether a second defect exists in the wood based on the relative scatter image; wherein the second defect comprises at least one of: a dead knot and a movable joint.

2. The method of claim 1, wherein identifying whether wood is present on the conveyor based on the laser line height position and/or laser line width of the laser line and the height position and scattering width values of the preset laser line pattern comprises:

comparing the height position and the width of the laser line with a preset first laser line model and a preset second laser line model to obtain a first similarity and a second similarity, wherein the first similarity represents the similarity between the laser line and the first laser line model, and the second similarity represents the similarity between the laser line and the second laser line model;

determining that the first similarity is greater than the second similarity, and judging that wood exists on the conveyor belt; or,

determining that the laser line width of the laser line exceeds the scattering width value of the laser line in the second laser line model and reaches a first set threshold value, and judging that wood exists on the conveyor belt; or,

and determining that the height position of the laser line exceeds the height position of the laser line in the second laser line model and reaches a second set threshold value, and judging that wood exists on the conveyor belt.

3. The method of claim 1, further comprising:

acquiring a first laser line image of the surface of the wood on the conveyor belt under the action of a laser line generated by the laser equipment, and constructing a first laser line model based on the first laser line image;

and acquiring a second laser line image of the surface of the conveyor belt under the action of the laser line generated by the laser equipment, and constructing a second laser line model based on the second laser line image.

4. The method of claim 3,

the constructing the first laser line model based on the first laser line image comprises:

acquiring a plurality of frames of the first laser line images;

determining the height position of the laser line on the wood based on the plurality of frames of the first laser line images;

determining a scattering width value of a laser line on the wood based on the plurality of frames of the first laser line images;

the constructing the second laser line model based on the second laser line image comprises:

acquiring a plurality of frames of the second laser line images;

determining the height position of the laser line on the conveyor belt based on the plurality of frames of the second laser line images;

and determining the scattering width value of the laser line on the conveyor belt based on the plurality of frames of the second laser line images.

5. The method of claim 1, wherein constructing the relative height image of the wood based on the difference of the laser line height position of the laser line and the height position in the first laser line model comprises:

and constructing a relative height image on the corresponding area of the wood aiming at the acquired multiple frames of laser line images, wherein the value of each pixel point in the relative height image is determined based on the difference between the height position of the laser line and the height position in the first laser line model.

6. The method of claim 1, wherein constructing a relative scatter image of the wood based on a difference between a laser line width of the laser line and a scatter width value in the first laser line model comprises:

and constructing a relative scattering image on the corresponding area of the wood aiming at the acquired multiple frames of laser line images, wherein the value of each pixel point in the relative scattering image is determined based on the difference value between the laser line width and the scattering width value in the first laser line model.

7. The method of claim 1, further comprising:

if the first defect exists, determining a defect position corresponding to the first defect based on the relative height image; and/or the presence of a gas in the gas,

and if the second defect exists, determining the defect position corresponding to the second defect based on the relative scattering image.

8. A wood surface defect detecting device, characterized by comprising:

the device comprises an acquisition module, a display module and a control module, wherein the acquisition module is used for acquiring a laser line image of a conveyor belt under the action of a laser line generated by laser equipment, and the laser line extends along the width direction of the conveyor belt on a horizontal plane;

the characteristic extraction module is used for extracting the laser line height position and the laser line width of the laser line based on the acquired laser line image; wherein the feature extraction module is further configured to: constructing a gray level image according to the laser line image; establishing a gray value distribution model according to the distribution model and the gray image; obtaining the gray probability of the pixel points according to the gray value distribution model; screening laser line pixel points according to the relation between the gray level probability and the probability threshold value; carrying out classified statistics on the number of the pixel points to obtain the laser line width; calculating the pixel points by classification statistics to obtain the height position of the laser line;

the first identification module is used for identifying whether wood exists on the conveyor belt or not based on the height position and/or the width of the laser line and the height position and the scattering width value of a preset laser line model, wherein the scattering width value is obtained according to the widths of a plurality of laser lines; wherein the preset laser line model comprises at least one of:

a first laser line model characterizing the height position and scattering width value of a laser line acting on the surface of the wood on the conveyor belt; the second laser line model is used for representing the height position and the scattering width value of the laser line acting on the surface of the conveyor belt;

a second recognition module, configured to identify the surface defects of the wood based on the height position and width of the laser line and the height position and width of the laser line model, if the wood is present on the conveyor belt, wherein the height position and width of the laser line identify the surface defects of the wood, including:

constructing a relative height image of the wood based on a difference between a laser line height position of the laser line and a height position in the first laser line model; determining whether the wood has a first defect based on the relative height image; wherein the first defect comprises at least one of: edge deletion and moth eye;

constructing a relative scattering image of the wood based on a difference value between the laser line width of the laser line and the scattering width value in the first laser line model; determining whether a second defect exists in the wood based on the relative scatter image; wherein the second defect comprises at least one of: a dead knot and a movable joint.

9. A wood surface defect detecting apparatus, comprising: a processor and a memory for storing a computer program capable of running on the processor, wherein,

the processor, when executing the computer program, is adapted to perform the steps of the method of any of claims 1 to 7.

10. A wood surface defect detection system, comprising:

the conveying belt is used for conveying the wood on line;

a laser device for generating a laser line acting on the wood;

the image acquisition equipment is used for acquiring a laser line image;

the wood surface defect detecting device of claim 9, being connected to the image capturing device to receive the laser line image captured by the image capturing device.

11. A storage medium having a computer program stored thereon, the computer program, when executed by a processor, implementing the steps of the method of any one of claims 1 to 7.

Images