A wood board detection system and method for furniture production

Abstract

The purpose of the present invention is to provide a wood board detection system and method for furniture production. The method mainly includes the following steps: S1: transporting the raw material board to the painting line; the painting process is continuous; S2: cutting the painted wood board; transporting the cut target board to the next process line; sending the cut waste board to the color detection line; wherein the waste board cut from the same target board and the corresponding target board are all marked with the same number x; S3: color detection of the cut waste board, and judging whether the paint color of the whole wood board is consistent through the color detection of the waste boards of multiple raw material boards. The wood board detection of the present invention is aimed at the cut waste board, and the paint of the target board is reversely deduced by whether the waste board is painted properly. When the detection is performed, the target board does not need to stop, and can directly enter the next process, thereby improving production efficiency. The color detection method of the present invention is simpler and more efficient than the prior art.

Description

Technical Field

[0001] The present invention particularly relates to a wood board detection system and method for furniture production. Background Art

[0002] With the development of modern industrialization, furniture production has also entered the stage of automatic production on assembly lines. Wood boards are the main material for furniture production. Wood boards may have various defects, such as uneven surface, poor quality, low strength, easy to rot, uneven color, etc. If unqualified wood boards are mixed into actual production, the furniture produced will also be unsightly, unqualified, and not meet market requirements.

[0003] Therefore, in order to produce qualified furniture products, it is necessary to test the wood boards before production.

[0004] Traditional wood board inspection is done through visual observation, which requires a lot of human resources, relies too much on workers’ experience, is inefficient, and does not meet the needs of modern industrialization.

[0005] With the development of computer computing, the technology of applying machine vision to wood board detection has also emerged. For example, the invention patent ZL201410642068X discloses a method for sorting defects in solid wood boards. It first extracts 25 features of three categories, namely, geometric and regional features, grayscale texture features, and invariant moment features of the defect image; secondly, the number of features is fused through the LDA method to reduce the feature dimension; finally, the data dictionary is constructed using the training samples, the optimal solution of the data dictionary is solved in the sense of L1 norm, and the linear programming solution is completed using the least squares to achieve defect classification. This method has high detection accuracy, but is also relatively complex and takes a long time.

[0006] At present, the inspection of wooden boards mainly includes the inspection of raw wood before painting, and the inspection of wooden boards after painting is less. Natural wooden boards have different colors due to their growth characteristics. In order to pursue beauty, modern homes will paint wooden boards to make their colors uniform. The uneven color of wooden boards after painting is mainly caused by the failure of dot matrix nozzles. Summary of the invention

[0007] The purpose of the present invention is to provide a wood board detection system and method for furniture production, mainly to realize the detection of furniture wood boards after painting.

[0008] To achieve the above-mentioned purpose, the present invention adopts the following technical scheme: a wood board detection system for furniture production, which comprises: a paint spraying line, on which a dot matrix nozzle for spraying paint on the raw material board is arranged; when spraying paint, the movement direction of the raw material board on the conveyor belt is perpendicular to the dot matrix nozzle; the painting process is continuous; a cutting line, which is arranged after the paint spraying line, and an electronic saw is arranged on the cutting line, and the electronic saw is controlled by an industrial computer; the painted wood board is transported to the electronic saw, the industrial computer is turned on, and the cutting line is input. Cutting data, cutting the painted wooden board; transporting the cut target board to the assembly line of the next process; sending the cut waste board to the color detection assembly line; wherein the waste board cut from the same target board and the corresponding target board are all marked with the same number x; the waste board at least includes the head and tail edge boards of the painted wooden board; and a color detection device is arranged on the color detection assembly line, the color detection device performs color detection on the cut waste board, and determines whether the paint color of the target board is consistent through the waste board color detection of multiple raw material boards.

[0009] In one embodiment of the present invention, the color detection device includes: an image sensor, which is used to collect images of waste boards; and transmit the collected images to a color processing module; a color processing module, wherein the color processor is used to preprocess the collected images; and extract color features from the preprocessed images; and transmit the extracted color features to a color feature comparison module; and a color feature comparison module, which compares the color features of different waste boards to determine whether the paint colors are consistent; if all the colors are consistent, the paint is determined to be qualified; the cut target board is usable; if the colors are inconsistent, it is further determined whether the numbers of the waste boards with inconsistent colors are the same. If the numbers of the waste boards with inconsistent colors are the same, the corresponding target boards are unqualified for paint.

[0010] Preferably, the color detection processing module uses the following method to extract color features: convert the color space of the preprocessed image, and perform experimental operations in different color spaces to extract color features; convert the color coordinates originally in the RGB color space into the HSV color space and the Lab color space, and then evenly divide the color space into 255 small color intervals, each small interval becoming a bin of the histogram; then, the color histogram can be obtained by calculating the number of pixels whose colors fall in each small interval; then, according to the following formula for calculating the color moment, the first-order color moment and the second-order color moment of the image are extracted;

[0011]

[0012] ;

[0013] in For the image at coordinates The pixel value at , N represents the number of pixels in the image; is the first-order moment of the image; is the second order moment.

[0014] Furthermore, image preprocessing includes filtering and noise reduction.

[0015] Furthermore, the color detection device also includes a counter, which is connected to the output of the color feature comparison module; when the color feature comparison module determines that the target board x has failed the paint spraying for two consecutive times, the count value A of the counter is increased by 1. When A is greater than a preset value, it can be determined that the corresponding nozzle has a fault, the corresponding nozzle is stopped from being used, and relevant personnel are notified to perform maintenance.

[0016] The present invention also provides a method for detecting wooden boards for furniture production, which includes the following steps: S1: transporting raw material boards to a paint spraying line; a dot matrix nozzle for painting the raw material boards is provided on the paint spraying line; the movement direction of the raw material boards on the conveyor belt is perpendicular to the dot matrix nozzle during painting; the painting process is continuous; S2: transporting the painted wooden boards to the electronic saw, turning on the industrial computer, inputting the cutting data, and cutting the painted wooden boards; transporting the cut target boards to the assembly line of the next process; sending the cut waste boards to a color detection assembly line; wherein the waste boards cut from the same target board and the corresponding target boards are all marked with the same number x; the waste boards at least include the head and tail edge boards of the painted wooden boards; S3: color detection of the cut waste boards, and judging whether the paint color of the entire wooden board is consistent by color detection of the waste boards of multiple raw material boards.

[0017] In one embodiment of the present invention, the color detection in S3 includes the following specific steps:

[0018] S31: collecting an image of the waste plate through an image sensor;

[0019] S32: preprocessing the collected images;

[0020] S33: extracting color features from the preprocessed image;

[0021] S34: Compare the color features of different waste boards to determine whether the paint colors are consistent; if all the colors are consistent, the paint is determined to be qualified; the cut target board is usable; if the colors are inconsistent, further determine whether the numbers of the waste boards with inconsistent colors are the same. If the numbers of the waste boards with inconsistent colors are the same, the corresponding target board is unqualified for painting.

[0022] Preferably, S33 includes the following specific steps: converting the color space of the preprocessed image, and performing experimental operations in different color spaces to extract color features; converting the color coordinates originally in the RGB color space into the HSV color space and the Lab color space, and then evenly dividing the color space into 255 small color intervals, each small interval becoming a bin of the histogram; then, a color histogram can be obtained by calculating the number of pixels whose colors fall within each small interval; then, according to the following formula for calculating the color moment, the first-order color moment and the second-order color moment of the image are extracted;

[0023]

[0024] ;

[0025] in For the image at coordinates The pixel value at , N represents the number of pixels in the image; is the first-order moment of the image; is the second order moment.

[0026] Furthermore, S3 also includes the following steps:

[0027] S35: After judging that the target board x fails to be painted for two consecutive times, the count value A is increased by 1. When A is greater than the preset value, it can be judged that the corresponding nozzle is faulty, the corresponding nozzle is stopped from being used, and the relevant personnel are notified for maintenance.

[0028] Preferably, the preprocessing in S32 includes filtering and noise reduction.

[0029] Compared with the prior art, the wood board detection of the present invention is targeted at the waste board after cutting, and the quality of the target board's painting is inferred by whether the waste board is painted properly. When performing the detection, the target board does not need to stop and can directly enter the next process, thereby improving production efficiency. The present invention only needs to detect whether the color of the waste board is consistent, and does not need to subdivide the color. Therefore, the color detection method used is simpler than the prior art and more efficient. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a schematic diagram of the main process of the present invention.

[0031] Figure 2 for Figure 1 Annotated schematic diagram.

[0032] Figure 3 Schematic diagram of the structure of a color detection device in one embodiment of the present invention.

[0033] Figure 4 Schematic diagram of the main process of color detection in one embodiment of the present invention. DETAILED DESCRIPTION

[0034] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0035] A wood board detection system for furniture production comprises: a painting line, on which a dot matrix nozzle for painting a raw material board is arranged; during painting, the moving direction of the raw material board on a conveyor belt is perpendicular to the dot matrix nozzle; the painting process is continuous; a cutting line, which is arranged after the painting line, and an electronic saw is arranged on the cutting line, and the electronic saw is controlled by an industrial computer; the painted wood board is transported to the electronic saw, the industrial computer is turned on, the cutting data is input, and the painted wood board is inspected. The target board is cut; the cut target board is transported to the assembly line of the next process; the cut waste board is sent to the color detection assembly line; wherein the waste board cut from the same target board and the corresponding target board are all marked with the same number x; the waste board at least includes the head and tail edge boards of the wooden board after painting; and a color detection device is arranged on the color detection assembly line, the color detection device performs color detection on the cut waste board, and judges whether the paint color of the target board is consistent through the color detection of the waste board of multiple raw material boards. Among them, the waste boards of multiple raw material boards at least include the waste boards that have been judged to be qualified for painting (that is, the color is consistent).

[0036] In one embodiment of the present invention, the color detection device includes: an image sensor, which is used to collect images of waste boards; and transmit the collected images to a color processing module; a color processing module, wherein the color processor is used to preprocess the collected images; and extract color features from the preprocessed images; and transmit the extracted color features to a color feature comparison module; and a color feature comparison module, which compares the color features of different waste boards to determine whether the paint colors are consistent; if all the colors are consistent, the paint is determined to be qualified; the cut target board is usable; if the colors are inconsistent, it is further determined whether the numbers of the waste boards with inconsistent colors are the same. If the numbers of the waste boards with inconsistent colors are the same, the corresponding target boards are unqualified for paint.

[0037] Preferably, the color detection processing module uses the following method to extract color features: convert the color space of the preprocessed image, and perform experimental operations in different color spaces to extract color features; convert the color coordinates originally in the RGB color space into the HSV color space and the Lab color space, and then evenly divide the color space into 255 small color intervals, each small interval becoming a bin of the histogram; then, the color histogram can be obtained by calculating the number of pixels whose colors fall in each small interval; then, according to the following formula for calculating the color moment, the first-order color moment and the second-order color moment of the image are extracted;

[0038]

[0039] ;

[0040] in For the image at coordinates The pixel value at , N represents the number of pixels in the image; is the first-order moment of the image; is the second order moment.

[0041] Furthermore, image preprocessing includes filtering and noise reduction.

[0042] Furthermore, the color detection device also includes a counter, which is connected to the output of the color feature comparison module; when the color feature comparison module determines that the target board x has failed the paint spraying for two consecutive times, the count value A of the counter is increased by 1. When A is greater than a preset value, it can be determined that the corresponding nozzle has a fault, the corresponding nozzle is stopped from being used, and relevant personnel are notified to perform maintenance.

[0043] A method for detecting wooden boards for furniture production comprises the following steps:

[0044] S1: transport the raw board to the painting line; the painting line is equipped with a dot matrix nozzle for painting the wood board; during painting, the movement direction of the raw board on the conveyor belt is perpendicular to the dot matrix nozzle; the painting process is continuous;

[0045] S2: transport the painted wooden board to the electric saw, turn on the industrial computer, input the cutting data, and cut the painted wooden board; cut it into the shape of the board we need, hereinafter referred to as the target board.

[0046] The cut target board is transported to the next process line; the cut waste board is sent to the color detection line; the waste board cut from the same target board and the corresponding target board are all marked with the same number x; the waste board at least includes the head and tail edge boards of the painted wood board;

[0047] S3: Performing color detection on the cut waste boards, and judging whether the paint color of the whole wooden board is uniform and the desired color through the color detection of the waste boards of multiple raw boards.

[0048] Color detection in S3 includes the following specific steps:

[0049] S31: collecting an image of the waste plate through an image sensor;

[0050] S32: preprocessing the collected image; the preprocessing includes noise reduction;

[0051] S33: extracting color features from the preprocessed image;

[0052] S34: Compare the color features of different waste boards to determine whether the paint colors are consistent; if all the colors are consistent, the paint is determined to be qualified; the cut target wood board is usable; if the colors are inconsistent, further determine whether the numbers of the waste boards with inconsistent colors are the same; if the numbers of the waste boards with inconsistent colors are the same, the corresponding target board is unqualified for paint spraying and is picked out. In a specific embodiment, a robotic arm can be used to sort it out.

[0053] In one embodiment of the present invention, S33 includes the following specific steps: converting the color space of the preprocessed image, and performing experimental operations in different color spaces to extract color features; converting the color coordinates originally in the RGB color space into the HSV color space and the Lab color space, and then evenly dividing the color space into 255 small color intervals, each small interval becoming a bin of the histogram; then, the color histogram can be obtained by calculating the number of pixels whose colors fall in each small interval; then, according to the following formula for calculating the color moment, the first-order color moment and the second-order color moment of the image are extracted;

[0054]

[0055] ;

[0056] in For the image at coordinates The pixel value at , N represents the number of pixels in the image; is the first-order moment of the image; is the second order moment.

[0057] In other embodiments of the present invention, S3 further includes the following steps:

[0058] S35: After judging that the target board x fails to be painted for two consecutive times, the count value A is increased by 1. When A is greater than the preset value, it can be judged that the corresponding nozzle is faulty, the corresponding nozzle is stopped from being used, and the relevant personnel are notified for maintenance.

[0059] like Figure 1 As shown, in a specific embodiment of the present invention, a raw material board 1 is sprayed with paint by a dot matrix nozzle 2 to obtain a sprayed painted board 3; an electronic saw 4 cuts the sprayed painted board 3 to obtain a cut board 5, which is divided into a waste board 6 and a target board 7; the waste board 6 and the target board 7 are sorted, wherein the target board 7 enters the next production process, punching or assembly, and the waste board 6 is color-detected by a color detection device 8 to determine whether the sprayed paint is qualified. The X axis is the direction of the assembly line.

[0060] The main structural diagram of the color detection device 8 is as follows Figure 3 See the main process diagram of color detection. Figure 4 .

[0061] like Figure 1 As shown in , the three waste boards ① and the target board I are all numbered 1, and the three waste boards ② and the target board II are all numbered 2.

[0062] The cut waste board is passed through the color detection device 8 for color detection:

[0063] Acquire an image of the waste plate by an image sensor;

[0064] Preprocessing the collected images; the preprocessing includes noise reduction;

[0065] The preprocessed image is converted into a color space, and experiments are performed in different color spaces to extract color features; the color coordinates in the original RGB color space are converted into the HSV color space and the Lab color space, and then the color space is evenly divided into 255 small color intervals, each of which becomes a bin of the histogram; then, the color histogram can be obtained by calculating the number of pixels whose colors fall in each small interval; then, the first-order color moment and the second-order color moment of the image are extracted according to the following formula for calculating the color moment;

[0066]

[0067] ;

[0068] in For the image at coordinates The pixel value at , N represents the number of pixels in the image; is the first-order moment of the image; is the second order moment.

[0069] Table 1 shows the first-order color matrix eigenvalues ​​of six waste boards

[0070]

[0071] Table 2 shows the second-order color matrix eigenvalues ​​of six waste boards

[0072]

[0073] Data 1-3 represent three waste boards ①, and data 4-6 represent three waste boards ②. From the data in Tables 1 and 2, it can be seen that within a certain error (which may be affected by the light and angle of image acquisition), except for data 3, the color feature values ​​of the data of other waste boards are basically the same. Further judging that the number of the waste board corresponding to data 3 is the same as the number of the waste board corresponding to data 1-2, the corresponding target board I is unqualified for painting.

[0074] Because the painting line is continuous, if the paint of the wood board at the same position is always unqualified, it is considered that the nozzle at the corresponding position is faulty, excluding environmental factors and detection accuracy. That is, if the target board I is judged to be unqualified for two consecutive times, the count value A is increased by 1. When A is greater than the preset value (which can be set to 2), it can be judged that the corresponding nozzle is faulty, stop using the corresponding nozzle, and notify relevant personnel for maintenance.

[0075] In order to facilitate the understanding of those skilled in the art, Figure 1 Take the detection principle of this application as an example to briefly explain.

[0076] When a nozzle fails, it may not produce any printout or it may produce uneven printout. The failure may occur at the waste plate position or at the target plate position. Figure 1 The plates are marked, see Figure 2 Regardless of which board a, b, c, or d the fault occurs on, at this time, since the movement direction of the wood board on the conveyor belt is perpendicular to the dot matrix nozzle during painting, and the painting process is continuous, the nozzle fault will not repair itself, so the following situations will generally occur:

[0077] (1) The nozzle has already failed before painting the current raw material board. The colors of the three waste boards a, c, and d at this time are definitely uneven and inconsistent with the waste boards that were previously painted normally.

[0078] (2) The nozzle fails during the painting process of the current raw material board. At this time, the colors of the three waste boards a, c, and d are definitely uneven.

[0079] If the nozzle failure only occurs at the position of the waste board, such as waste boards a, c, and d, it is necessary to rely on an increasing counter to identify it. That is, by setting a counter, the counter is connected to the output of the color feature comparison module; when the color feature comparison module determines that the target board x is unqualified for two consecutive times, the count value A of the counter is increased by 1. When A is greater than the preset value, it is determined that the corresponding nozzle is faulty, stop using the corresponding nozzle, and notify relevant personnel to repair it.

[0080] In very extreme cases, the technical solution of the present application may have misjudgments, for example, the nozzle suddenly fails when spraying the target board b, and suddenly becomes good before spraying the waste board c. That is, the nozzle suddenly fails and then suddenly recovers by itself (accidental event 1), and the failure occurs when the b board is sprayed with color (accidental event 2), and the failure self-recovery occurs before the c board is sprayed with paint (accidental event 3). In other words, this requires multiple accidental events to occur at the same time, which is a low-probability event in actual engineering. The probability of this low-probability event itself is extremely low. Even if it occurs, it will only affect a very few painted boards (in the case of the occurrence of event 3, before the next raw material board is painted, the nozzle failure has been restored, so it will not affect the painting of the next raw material board), and will not affect the detection results of the entire production line. Errors in the detection of one or two boards (missed detection) are acceptable in actual use. For economic considerations and detection efficiency, the above situation can be recorded as an engineering detection error and ignored.

[0081] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0082] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A wood board detection system for furniture production, characterized in that: include: A paint spraying line is provided with a dot-matrix nozzle for spraying paint on the raw material plate; during spraying, the movement direction of the raw material plate on the conveyor belt is perpendicular to the dot-matrix nozzle; The painting process is continuous; A cutting line, which is arranged after the painting line, and is provided with an electronic saw, which is controlled by an industrial computer; the painted wooden board is transported to the electronic saw, the industrial computer is turned on, the cutting data is input, and the painted wooden board is cut; the cut target board is transported to the next process line; the cut waste board is sent to the color detection line; the waste board cut from the same target board and the corresponding target board are all marked with the same number x; the waste board at least includes the head and tail edge boards of the painted wooden board; and a color detection device arranged on the color detection assembly line, the color detection device performs color detection on the cut waste board, and determines whether the paint color of the target board is consistent through the color detection of the waste board of multiple raw material boards; The color detection device comprises: An image sensor, the image sensor is used to collect an image of the waste board; and transmit the collected image to the color detection processing module; A color processing module, wherein the color detection processor is used to preprocess the collected image; extract color features from the preprocessed image; and transmit the extracted color features to the color feature comparison module; and a color feature comparison module, which compares the color features of different waste boards to determine whether the paint colors are consistent; if all the colors are consistent, the paint is determined to be qualified; the cut target board is usable; if the colors are inconsistent, it is further determined whether the numbers of the waste boards with inconsistent colors are the same, and if the numbers of the waste boards with inconsistent colors are the same, the corresponding target board is unqualified for paint; The color detection device also includes a counter, which is connected to the output of the color feature comparison module; When the color feature comparison module determines that the target board x has failed the paint spraying for two consecutive times, the count value A of the counter is increased by 1. When A is greater than the preset value, it is determined that the corresponding nozzle is faulty, the corresponding nozzle is stopped from being used, and the relevant personnel are notified for maintenance. The preset value is 2.

2. The wood board detection system for furniture production according to claim 1, characterized in that: The color processing module uses the following method to extract color features: Convert the preprocessed image to a different color space and perform experiments in different color spaces to extract color features. Convert the original color coordinates in the RGB color space into the HSV color space and the Lab color space, and then evenly divide the color space into 255 small color intervals, each of which becomes a bin of the histogram; then, calculate the number of pixels whose colors fall in each small interval to obtain the color histogram; then, extract the first-order color moment and second-order color moment of the image according to the following formula for calculating the color moment; where p i，j is the pixel value of the image at coordinate (i, j), and N represents the number of pixels in the image; μ i is the first-order moment of the image; σ i is the second-order moment.

3. The wood board detection system for furniture production according to claim 1, characterized in that: Image preprocessing includes filtering and noise reduction.

4. A method for detecting wooden boards for furniture production, characterized in that: The following steps are involved: S1: transport the raw material board to the painting line; the painting line is equipped with a dot matrix nozzle for painting the raw material board; when painting, the movement direction of the raw material board on the conveyor belt is perpendicular to the dot matrix nozzle; the painting process is continuous, S2: transport the painted wooden board to the electronic saw, turn on the industrial computer, input the cutting data, and cut the painted wooden board; transport the cut target board to the assembly line of the next process; send the cut waste board to the color detection assembly line; the waste board cut from the same target board and the corresponding target board are all marked with the same number x; the waste board at least includes the head and tail edge boards of the painted wooden board; S3: Performing color detection on the cut waste boards, and judging whether the paint color of the whole wooden board is consistent by color detection of the waste boards of multiple raw boards; Color detection in S3 includes the following specific steps: S31: collecting an image of the waste plate through an image sensor; S32: preprocessing the collected images; S33: extracting color features from the preprocessed image; S34: Compare the color features of different waste boards to determine whether the paint colors are consistent; if all the colors are consistent, the paint is determined to be qualified; the cut target board is usable; if the colors are inconsistent, further determine whether the numbers of the waste boards with inconsistent colors are the same; if the numbers of the waste boards with inconsistent colors are the same, the corresponding target board is unqualified for paint; S35: After judging that the target board x fails to be painted for two consecutive times, the count value A is increased by 1. When A is greater than a preset value, it is judged that the corresponding nozzle is faulty, the corresponding nozzle is stopped from being used, and the relevant personnel are notified for maintenance. The preset value is 2.

5. The method for detecting wooden boards for furniture production according to claim 4, characterized in that: S33 includes the following specific steps: converting the preprocessed image into a color space, and performing experimental operations in different color spaces to extract color features; Convert the original color coordinates in the RGB color space into the HSV color space and the Lab color space, and then evenly divide the color space into 255 small color intervals, each of which becomes a bin of the histogram; then, calculate the number of pixels whose colors fall in each small interval to obtain the color histogram; then, extract the first-order color moment and second-order color moment of the image according to the following formula for calculating the color moment; where p i，j is the pixel value of the image at coordinate (i, j), and N represents the number of pixels in the image; μ i is the first-order moment of the image; σ i is the second-order moment.

6. The method for detecting wooden boards for furniture production according to claim 4, characterized in that: The preprocessing in S32 includes filtering and noise reduction.

Images