CN108247764B

CN108247764B - Wood board cutting method and device based on machine learning, electronic equipment and medium

Info

Publication number: CN108247764B
Application number: CN201711335872.3A
Authority: CN
Inventors: 丁磊
Original assignee: Beijing Muyebang Technology Co ltd
Current assignee: Beijing Muyebang Technology Co ltd
Priority date: 2017-12-14
Filing date: 2017-12-14
Publication date: 2021-04-27
Anticipated expiration: 2037-12-14
Also published as: CN108247764A

Abstract

The invention provides a wood board cutting method, a wood board cutting device, electronic equipment and a wood board cutting medium based on machine learning, wherein the wood board cutting method comprises the following steps: acquiring an image of wood to be cut; detecting and identifying the image of the wood to be cut according to a wood board cutting model, wherein the wood board cutting model is obtained through machine learning; and obtaining a cutting route of the wood to be cut at least based on the cut texture according to the detection and identification results. The method has the advantages that planning of wood board cutting is achieved in an artificial intelligence mode, the method for machine learning is very simple, manual intervention is not needed, accuracy of the model can be improved continuously, and accordingly identification accuracy and identification effect of the cutting route are improved greatly.

Description

Wood board cutting method and device based on machine learning, electronic equipment and medium

Technical Field

The invention belongs to the field of artificial intelligence, and particularly relates to a wood board cutting method and device based on machine learning, electronic equipment and a medium.

Background

In the field of wood processing, logs are cut from end faces and then cut from the faces into rectangular boards of various sizes for splicing into flooring or for processing into furniture and various applications in industry or life. After the end face is cut, various textures are formed on the board surface, the traditional cutting mode does not take the texture shape as an important consideration factor, the cut wood board textures are uneven, and the textures on two sides of the joint part after splicing are in different shapes, so that the attractiveness is influenced.

If the texture shape is considered, the cut wood boards can show a muddy integral visual effect after being spliced, and the attractiveness and the value are improved. However, the texture shape recognition and the cutting route marking are usually completed manually, so that the higher efficiency is difficult to achieve, and the rhythm of the automatic mass production of the information society and the large-scale demand of products cannot be kept up with.

Disclosure of Invention

Aiming at the problem that the traditional cutting mode in the prior art has mismatched texture and shape, so that the attractiveness is affected, the invention provides a wood board cutting method, a wood board cutting device, electronic equipment and a wood board cutting medium based on machine learning, wherein the wood board cutting method is used for obtaining a wood board cutting route by inputting an acquired wood board image into a cutting route calculation module. The cutting route calculation module is a machine learning module adopting a machine learning method. Machine learning enables automated machines to become more reliable and flexible through massive amounts of training data. Meanwhile, the training method is very simple, manual intervention is not needed, and the accuracy of the model can be continuously improved, so that the identification accuracy and the identification effect of the cutting route are greatly improved.

In order to achieve the above object, in a first aspect of embodiments of the present invention, there is provided a machine learning-based plank cutting method, including:

acquiring an image of wood to be cut;

detecting and identifying the image of the wood to be cut according to a wood board cutting model, wherein the wood board cutting model is obtained through machine learning;

and obtaining a cutting route of the wood to be cut at least based on the cut texture according to the detection and identification results.

Optionally, the acquiring the image of the wood to be cut includes:

and irradiating the wood to be cut by using constant light, and acquiring an image of the irradiated wood to be cut.

Optionally, the acquiring the image of the wood to be cut includes:

acquiring an image of the wood to be cut from the front direction of the wood to be cut, wherein the front direction is a direction perpendicular to the largest section of the wood to be cut.

Optionally, the method further comprises:

acquiring an image of a wood sample;

receiving annotation information for the image of the wood sample;

and training machine learning according to the image of the wood sample and the labeling information to obtain the wood board cutting model.

Optionally, the marking information comprises a preferred cutting route of the wood sample.

Optionally, the preferred cutting path comprises at least one of a gauge number, a starting point, an angle, and a length.

Optionally, the tagging information further includes: and at least one of the outturn rate, the texture of the cut wood board, the size of the cut wood board and the corresponding cutting mode of the end face of the log corresponding to the optimal cutting route.

Optionally, the machine learning comprises: at least one of a neural network, a random forest, a decision tree, and a support vector machine.

Optionally, the neural network comprises: at least one of a multi-layer neural network, a deep neural network, and a convolutional neural network.

Optionally, the method further comprises: and controlling a line drawing and/or cutting device to process the wood to be cut according to the cutting route.

In another aspect of the embodiments of the present application, there is also provided a plank cutting apparatus based on machine learning, the apparatus including:

the image acquisition module is used for acquiring an image of the wood to be cut;

the model matching module is used for detecting and identifying the image of the wood to be cut according to a wood board cutting model, wherein the wood board cutting model is obtained through machine learning;

and the route planning module is used for obtaining the cutting route of the wood to be cut at least based on the cut texture according to the detection and identification result.

Optionally, the image acquisition module comprises:

and the light source control module is used for irradiating the wood to be cut by using constant light and acquiring an image of the irradiated wood to be cut.

Optionally, the image acquisition module further comprises:

the front image capturing module is used for acquiring an image of the wood to be cut from the front direction of the wood to be cut, wherein the front direction is perpendicular to the direction of the largest section of the wood to be cut.

Optionally, the apparatus further comprises:

the sample image acquisition module is used for acquiring an image of the wood sample;

the annotation input module is used for receiving annotation information of the image of the wood sample;

and the training module is used for training machine learning according to the image of the wood sample and the labeling information to obtain the wood board cutting model.

Optionally, the apparatus further comprises:

and the control module is used for controlling the line drawing and/or cutting equipment to process the wood to be cut according to the cutting route.

Further, the training data of the machine learning module includes a label of a cutting route of the acquired image of the wood board.

Further, the label of the cutting route comprises a tool distance number, a starting point, an angle and a length.

Further, the training data of the machine learning module further comprises at least one of the following labeling data: the method comprises the steps of obtaining the outturn rate corresponding to the image of the wood board, the texture of the cut wood board, the size of the cut wood board and the cutting mode of the log end face corresponding to the wood board to be cut.

Further, the generation of the training data is performed locally and/or at the network side, the training of the training data is performed locally and/or at the network side, and the storage of the training data is performed locally and/or at the network side.

Further, the machine learning method includes: neural networks, random forests, decision trees and support vector machines.

Further, the neural network includes: multilayer neural networks, deep neural networks, convolutional neural networks.

Yet another aspect of embodiments of the present invention provides an electronic device, including a memory and a processor; wherein the memory is configured to store one or more computer instructions, wherein the one or more computer instructions are executed by the processor to implement the steps of the method described above.

Yet another aspect of an embodiment of the present invention provides a computer-readable storage medium having stored thereon computer instructions which are executed to implement the steps of the method described above.

The technical scheme provided by the embodiment of the invention has the following beneficial effects:

the invention provides an artificial intelligence based automatic wood board cutting method, which is characterized in that an acquired image of a wood board is input into a cutting route calculation module, so that a cutting route of the wood board is obtained. The cutting route calculation module is a machine learning module adopting a machine learning method. Machine learning enables automated machines to become more reliable and flexible through massive amounts of training data. Meanwhile, the training method is very simple, manual intervention is not needed, and the accuracy of the model can be continuously improved, so that the identification accuracy and the identification effect of the cutting route are greatly improved.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention, as claimed.

Drawings

Other features, objects and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments thereof, when taken in conjunction with the accompanying drawings.

In the drawings:

FIG. 1 is a schematic flow chart of a method in accordance with an embodiment of the present invention;

FIG. 2 is a schematic diagram of an apparatus according to an embodiment of the present invention;

FIG. 3 is a schematic structural diagram of an electronic device according to an embodiment of the invention;

FIG. 4 is a schematic diagram of different types of texture features;

FIG. 5 is a schematic illustration of obtaining an image of a wood board in accordance with an embodiment of the present invention;

FIG. 6 is a schematic view of a knife cutting in one embodiment of the present invention;

FIG. 7 is a schematic diagram of a convolutional neural network in an embodiment of the present invention.

Detailed Description

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily implement them. Also, for the sake of clarity, parts not relevant to the description of the exemplary embodiments are omitted in the drawings.

In the present invention, it is to be understood that terms such as "including" or "having," etc., are intended to indicate the presence of the disclosed features, numbers, steps, actions, components, parts, or combinations thereof, and are not intended to preclude the possibility that one or more other features, numbers, steps, actions, components, parts, or combinations thereof may be present or added.

It should be noted that the embodiments and features of the embodiments may be combined with each other without conflict. The present invention will be described in detail below with reference to the embodiments with reference to the attached drawings.

As described in the background, in the field of wood processing, various textures are formed on a board surface after end face cutting. Because the traditional cutting mode does not take the texture shape as an important consideration factor, the cut wood board has uneven texture, and the textures on two sides of the joint show different shapes after splicing, thereby influencing the aesthetic property. If the texture shape is considered, the cut wood boards can show a muddy integral visual effect after being spliced, and the attractiveness and the value are improved. However, in the existing wood board cutting technology, the identification of the texture and the shape and the marking of the cutting route are finished manually, so that the standardization is difficult, the efficiency is extremely low, and the rhythm of the automatic mass production of the information society and the large-scale demand of products cannot be followed.

In recent years, machine learning has been rapidly developed with the improvement of computer processing technology. If machine learning is introduced to the texture shape recognition and cutting route marking, the board cutting efficiency is undoubtedly greatly improved. The cutting efficiency not only considers the processing speed, but also considers the influence of the texture shape on the aesthetic degree of the processed wood board.

Machine learning enables automated machines to become more reliable and flexible through massive amounts of training data. Meanwhile, the training method of machine learning is very simple, and only data needs to be labeled and a machine learning model needs to be trained by using a corresponding algorithm. And the machine learning model can identify the cutting route according to the texture characteristics of the wood without manual intervention. That is to say, the machine learning method abandons the dependence on specific cutting route identification models, algorithms and image characteristics, and can realize the cutting of non-standardized wood boards only by collecting enough data. In addition, the method based on machine learning can continuously improve the precision of the cutting route recognition algorithm by continuously inputting new training data, which is an advantage that the traditional method does not have. Meanwhile, the method based on machine learning can also rapidly change the recognition rule and recognition effect of the cutting route under the condition of unchanging a hardware structure by changing a labeling method and selecting and using different training data sets.

Aiming at the problem that the appearance is influenced due to the fact that textures and shapes are not matched in the traditional cutting mode in the prior art, the invention provides a board cutting method, a device, electronic equipment and a medium based on machine learning. The wood board cutting model is an artificial intelligence model obtained through machine learning. Machine learning enables automated machines to become more reliable and flexible through massive amounts of training data. Meanwhile, the training method is very simple, manual intervention is not needed, and the accuracy of the model can be continuously improved, so that the identification accuracy and the identification effect of the cutting route are greatly improved.

One embodiment of the invention also comprises a device deployed in a wood board processing factory, wherein the device comprises an image sensor, a cutting route identification module and a cutter control module. The image sensor is used for acquiring a sensing image, the cutting route recognition module is used for outputting a cutting route of the cutter through the acquired image, and the cutter control module controls the cutter to cut the wood board according to the cutting route. Further, the main body of the invention may be connected to a server via a network.

Note that the present invention is not limited to cutter cutting, and may be laser cutting or high-pressure water flow cutting, and the cutter cutting is only taken as an example here.

In a first aspect, the invention provides a wood board cutting method based on machine learning.

Fig. 1 is a schematic flow chart of a method of a wood board cutting method based on machine learning according to an embodiment of the present invention.

In one embodiment of the present invention, as shown in fig. 1, there is provided a machine learning-based plank cutting method, the method comprising:

s101, acquiring an image of wood to be cut;

s102, detecting and identifying the image of the wood to be cut according to a wood board cutting model, wherein the wood board cutting model is obtained through machine learning;

s103, obtaining a cutting route of the wood to be cut at least based on the cut texture according to the detection and identification result.

In one embodiment, the acquiring the image of the wood to be cut includes:

In the embodiment of the application, the cutting route is obtained at least based on the texture features of the cut wood board, and the texture features can be obtained by acquiring texture images through a camera.

The surface textures of the wood are different, some of the wood are like waves, and some of the wood are straight. And the shape of the grain is related to the cutting mode of the log processing plank. According to the difference of the included angle between the annual ring on the end surface of the log and the board surface, the log processing board has four cutting modes: chordal (planar) cutting, quarter (tangential) cutting, diametral (tangential) cutting, and straight (live) cutting, whereby the panel surface exhibits different types of texture, as shown in fig. 4.

After a predetermined number of wood board image samples and annotation data corresponding to the image samples are collected in advance, a training data generation and model training stage is started.

The texture of the board surface to be cut is set, and the texture and the outturn rate are comprehensively considered for how the wood board is cut. The existing cutting method only considers the outturn percentage but neglects the texture, so that the texture of the cut wood board presents irregular randomness, and each wood board cannot be guaranteed to have higher commercial value. The cut texture may be a horizontal texture, a vertical texture, an inclined texture, or an arc-shaped envelope texture, considering the effect of the texture. The design of the cutting scheme needs to comprehensively consider various characteristics of the texture, such as the envelope texture, and the cutting needs to maintain the integrity of the texture as much as possible so as to avoid damaging the envelope texture. However, unilateral consideration of texture can result in significant loss in lumber recovery, resulting in substantial wastage of lumber. The less timber remaining at the end of a cutting plan means higher outturn rate, and if only a small number of boards with good grain appearance can be cut, it means that the cost is increased due to the reduction of outturn rate, so that a better cutting plan is capable of maintaining high outturn rate and simultaneously ensuring the good appearance of the grain of each board. Due to the myriad cutting schemes available for a single board to be processed, selection of a cutting scheme that has a high outturn rate and good grain aesthetics requires a very elaborate planning of the cutting scheme. The cut size is also required, the values of which are enumerated and can only be selected from a limited number of sizes. The wood boards with different sizes can be cut by different wood boards to be cut, and the wood boards with different sizes can also be cut on the same wood board to be cut. In the traditional wood processing field, trained workers judge the shape and texture of each wood board and plan a cutting scheme.

The machine learning method is introduced into the field of wood processing, so that the cutting efficiency of the wood board can be effectively improved.

In one embodiment, the method further comprises:

acquiring an image of a wood sample;

receiving annotation information for the image of the wood sample;

In one embodiment, the marking information includes a preferred cutting route of the wood sample.

In one embodiment, the preferred cutting path includes at least one of a gauge number, a starting point, an angle, and a length.

The method comprises the following specific steps:

the method comprises the steps of firstly, manually designing and marking a cutting route of a wood board sample, manually designing a cutting route for each wood board according to the planning based on the aspects of texture, outturn percentage and the like, and then marking the cutting route in a segmented mode according to straight lines, wherein each straight line is used as a marking unit. For example, the labeling method using two tools may be as follows:

{ tool 1, starting point 1, angle 1, length 1}, { tool 2, starting point 2, angle 2, length 2 }.

By way of the above notation, a specific implementation of a cutting scheme can be described. For the case of different numbers of knives and different numbers of cutting lines, reference is made to the above example and so on.

In one embodiment, the annotation information further comprises: and at least one of the outturn rate, the texture of the cut wood board, the size of the cut wood board and the corresponding cutting mode of the end face of the log corresponding to the optimal cutting route.

The method comprises the following specific steps:

in one embodiment, the marking includes not only marking the cutting path scheme, but also marking the texture features of the cut wood board, such as straight lines, horizontal lines, diagonal lines, and enveloping lines. In this way, the specific textural characteristics of each cut plank may be included in the marking information.

In one embodiment, the outturn rate may also be labeled such that the final outturn rate of a scenario can be included in the labeling information.

In one embodiment, the size of each wood board generated by cutting under the cutting route can be labeled, so that the size of the wood board generated by cutting is included in the labeling information.

In one embodiment, the cutting mode of the end face of the log corresponding to the plank to be cut can be marked.

In one embodiment, the machine learning comprises: at least one of a neural network, a random forest, a decision tree, and a support vector machine.

After the training data is obtained, a cutting route automatic identification model can be trained through a machine learning method. Because the image data contains texture information and the labeling information also contains a cutting scheme and related information designed according to the texture information, the cutting scheme optimization design method based on the double aspects of texture and outturn rate, which can be realized only manually at present, can be realized by the machine learning method under the premise of acquiring a sufficient amount of training data. Here we use a neural network approach as an example, and similarly, the training data and methods can be applied to random forests, decision trees, support vector machines, and other machine learning methods.

In one embodiment, the neural network comprises: at least one of a multi-layer neural network, a deep neural network, and a convolutional neural network.

The neural network comprises a plurality of layers, each layer comprises a plurality of nodes, and the neural network with trainable weight exists between the plurality of nodes of two adjacent layers. Fig. 7 shows a schematic diagram of a convolutional neural network. The figure includes a plurality of convolutional and downsampling layers and a fully connected layer. The convolutional layer is a core module of a convolutional neural network, and connects a plurality of nodes of a previous layer with nodes of a next layer through a convolution operation with a filter (filter). Generally, each node of a convolutional layer is connected to only a portion of the nodes of the previous layer. Through training, the filter using the initial value can continuously change the weight of the filter, and then a final filter value is generated. The down-sampling layer may use a max-pooling (max-pooling) method to reduce a group of nodes into one node, and preferably, the down-sampling layer may use a non-linear maximum-value method to reduce a group of nodes into one node. After passing through the plurality of convolutional layers and downsampling layers, a fully-connected layer is ultimately used to produce the identified data output.

In the machine learning process, namely the training process, image sample data of the wood board is used as input, attributes such as cutting route marking and the like where the image sample data are located are used as output, the weight value of a filter in a neural network is changed through a training algorithm, for example, a gradient descent algorithm, and therefore the recognition difference between the output and the sample data is minimum. Along with the continuous increase of the used training data quantity, the weight value of the filter in the neural network is continuously changed and converged, and the recognition capability of the neural network is correspondingly improved. When training is completed, a trained neural network includes the designed network architecture, such as the hierarchical design and the connection method between the hierarchies in fig. 7, and the filter weight values that are changed through training. These weight values are recorded and reused in later applications.

Neural networks are commonly used for classification, and here cut-line identification uses neural networks for regression, the only difference being whether the final layer of the network converts continuous data into discrete data. The core technology of classification or regression is to make the network train and learn through back propagation, compare the predicted value with the actual reference value, and continuously adjust the weight and deviation of the network filter until the prediction error is minimized. Therefore, other machine learning methods that can be used for regression are also applicable to the solution of the present invention.

In one embodiment, the generation of the training data is performed locally and/or at the network side, the training of the training data is performed locally and/or at the network side, and the storage of the training data is performed locally and/or at the network side.

Specifically, the method comprises the following steps:

the machine learning process can be completed locally or in the cloud. In one embodiment, as shown in fig. 5, the collector collects image data of the wood board sample and the labeled data set and transmits the image data and the labeled data set to the cloud server for training, and the server transmits the trained machine learning model to the local processor and sets the model.

In one embodiment, the cloud server may use training data from multiple sources. Such as data from multiple local image acquisitions and annotations. The labeling standard of the data set used for training the model needs to be matched with the image acquisition equipment, otherwise the accuracy of later cutting route identification is influenced. Meanwhile, the labeling standards of multiple data sets need to be consistent, so that the multiple data sets can be used for training the same neural network.

After training is completed, the neural network can output cutting routes, and one cutting route can have similar results with the labeled information, for example: { cutter 1, starting point 1, angle 1, length 1 }.

In one embodiment, the machine learning method adopted by the machine learning module comprises the following steps: inputting the acquired image into the trained machine learning module, so as to obtain the cutting route of the wood board.

In one embodiment, the method further comprises: and controlling a line drawing and/or cutting device to process the wood to be cut according to the cutting route.

The method comprises the following specific steps:

after the plank cutting device completes training, the collected plank sample image is input to the cutting route calculation module, and the output cutting route is transmitted to the cutter control module, as shown in fig. 6. In one embodiment, the output cutting path is not directly transmitted to the tool control module, but is simply a path drawn or printed directly on the sample of wood board, after which the particular cut is identified by a path identification module (e.g., an automatic vision sensor-based identification device) which in turn controls the tool control module to complete the cut. The cutter control module controls a plurality of motor-driven cutters to cut the wood board, for example, one caliper fixes the wood board, the motor 1 drives the caliper to rotate the wood board, so that the relative angle between the cutter and the caliper reaches an angle 1, the motor 2 drives the conveying device, the cutting position of the cutter reaches a starting point 1, and the motor 3 drives the conveying device to stop cutting after the cutting length 1 is finished. The above method is a cutting method in which the cutter is fixed and the caliper responsible for fixing the wood board moves. Similarly, the cutter can be driven to complete the cutting under the condition that the caliper is not moved.

In one embodiment, the neural network can simultaneously evaluate and output the cut-out rate of the cutting route, for example, the specific cut-out rate of the wood board sample is sent to a display terminal.

In one embodiment, the neural network is simultaneously capable of outputting texture information for each cut plank, which information may be used for subsequent plank classification.

In one embodiment, the neural network is simultaneously capable of outputting dimensional information for each cut plank, which information may be used for subsequent plank classification.

In one embodiment, the neural network is capable of outputting the log end face cutting mode corresponding to the plank to be cut at the same time, and this information can be used for subsequent plank classification.

In an implementation mode, the neural network can simultaneously output the log end face cutting mode, the texture of the wood board to be cut, the outturn rate, the size of each wood board to be cut and other related information, and the related information can be used for carrying out statistical analysis on subsequent products. For example, the marking related information can find out which end face cutting mode can generate more high-quality products, and then the end face cutting mode is fed back to the cutting mode to optimize the cutting mode.

In a second aspect, the present invention provides a board cutting apparatus based on machine learning. The device embodiment can be used for executing the method embodiment of the invention.

Fig. 2 shows a block diagram of a machine learning-based plank cutting apparatus according to an embodiment of the present invention, which may be implemented as part or all of an electronic device by software, hardware, or a combination of both. As shown in fig. 2, a plank cutting apparatus based on machine learning, comprising:

the image acquisition module 201 is used for acquiring an image of wood to be cut;

the model matching module 202 is used for detecting and identifying the image of the wood to be cut according to a wood board cutting model, wherein the wood board cutting model is obtained through machine learning;

and the route planning module 203 is used for obtaining the cutting route of the wood to be cut at least based on the cut texture according to the detection and identification result.

In one embodiment, the image acquisition module comprises:

In one embodiment, the image acquisition module further comprises:

In one embodiment, the apparatus further comprises:

In this embodiment, the wood board cutting device based on machine learning corresponds to the method described above, and specific details can be referred to the description of the method section above, which is not described herein again.

In a third aspect, an embodiment of the present invention provides an electronic device, including a memory and a processor; wherein the memory is configured to store one or more computer instructions, wherein the one or more computer instructions are executed by the processor to implement the method steps of the first aspect.

FIG. 3 is a schematic block diagram of a machine learning based plank cutting electronic device suitable for use in implementing embodiments of the present invention.

As shown in fig. 3, the electronic apparatus 300 includes a Central Processing Unit (CPU)301 that can execute various processes in the embodiment shown in fig. 1 described above according to a program stored in a Read Only Memory (ROM)302 or a program loaded from a storage section 308 into a Random Access Memory (RAM) 303. In the RAM303, various programs and data necessary for the operation of the electronic apparatus 300 are also stored. The CPU301, ROM302, and RAM303 are connected to each other via a bus 304. An input/output (I/O) interface 305 is also connected to bus 304.

The following components are connected to the I/O interface 305: an input portion 306 including a keyboard, a mouse, and the like; an output section 307 including a display such as a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and the like, and a speaker; a storage section 308 including a hard disk and the like; and a communication section 309 including a network interface card such as a LAN card, a modem, or the like. The communication section 309 performs communication processing via a network such as the internet. A drive 310 is also connected to the I/O interface 305 as needed. A removable medium 311 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is mounted on the drive 310 as necessary, so that a computer program read out therefrom is mounted into the storage section 308 as necessary.

In a fourth aspect, embodiments of the present invention provide a computer-readable storage medium having stored thereon computer instructions which, when executed by a processor, implement the method steps recited in the first aspect of the invention. The computer-readable storage medium may be the computer-readable storage medium included in the apparatus described in the above embodiment; or it may be a separate computer readable storage medium not incorporated into the device. The computer readable storage medium stores one or more programs which are used by one or more processors to perform the methods described in the present invention. In such an embodiment, the computer program may be downloaded and installed from a network through the communication section 309, and/or installed from the removable medium 311.

The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowcharts or block diagrams may represent a module, a program segment, or a portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.

The units or modules described in the embodiments of the present disclosure may be implemented by software or hardware. The units or modules described may also be provided in a processor, and the names of the units or modules do not in some cases constitute a limitation of the units or modules themselves.

The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.

Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.

Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.

Claims

1. A method for machine learning based plank cutting, the method comprising:

irradiating wood to be cut by using constant light, and acquiring an image of the irradiated wood to be cut from the front direction of the wood to be cut; wherein the front direction is a direction perpendicular to the largest section of the wood to be cut;

obtaining a cutting route of the wood to be cut at least based on the cut texture according to the detection and identification result; the texture features are obtained by acquiring texture images through a camera.

2. The method of claim 1, wherein the method further comprises:

acquiring an image of a wood sample;

receiving annotation information for the image of the wood sample;

3. The method of claim 2, wherein the marking information comprises a preferred cutting route of the wood sample.

4. The method of claim 3, wherein the preferred cutting path comprises at least one of a gauge number, a starting point, an angle, and a length.

5. The method of claim 3, wherein the annotation information further comprises: and at least one of the outturn rate corresponding to the preferred cutting route, the texture of the cut wood board, the size of the cut wood board and the corresponding cutting mode of the end face of the log.

6. The method of claim 2, wherein the machine learning comprises: at least one of a neural network, a random forest, a decision tree, and a support vector machine.

7. The method of claim 6, wherein the neural network comprises: at least one of a multi-layer neural network, a deep neural network, and a convolutional neural network.

8. The method of any one of claims 1-7, further comprising: and controlling a line drawing and/or cutting device to process the wood to be cut according to the cutting route.

9. A plank cutting apparatus based on machine learning, the apparatus comprising:

the light source control module is used for irradiating the wood to be cut by using constant light;

the image acquisition module is used for acquiring the irradiated image of the wood to be cut from the front direction of the wood to be cut; wherein the front direction is a direction perpendicular to the largest section of the wood to be cut;

the route planning module is used for obtaining a cutting route of the wood to be cut at least based on the cut texture according to the detection and identification result; the texture features are obtained by acquiring texture images through a camera.

10. The apparatus of claim 9, wherein the apparatus further comprises:

11. The apparatus of claim 9 or 10, wherein the apparatus further comprises:

12. An electronic device comprising a memory and a processor; wherein the content of the first and second substances,

the memory is for storing one or more computer instructions, wherein the one or more computer instructions are executed by the processor to implement the method steps of any of claims 1-8.

13. A computer-readable storage medium having stored thereon computer instructions, characterized in that the computer instructions are executed to implement the method steps according to any one of claims 1-8.