CN110245459B - Laser cleaning effect previewing method and device - Google Patents
Laser cleaning effect previewing method and device Download PDFInfo
- Publication number
- CN110245459B CN110245459B CN201910575724.1A CN201910575724A CN110245459B CN 110245459 B CN110245459 B CN 110245459B CN 201910575724 A CN201910575724 A CN 201910575724A CN 110245459 B CN110245459 B CN 110245459B
- Authority
- CN
- China
- Prior art keywords
- cleaning
- picture
- network model
- cleaned
- laser
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B08—CLEANING
- B08B—CLEANING IN GENERAL; PREVENTION OF FOULING IN GENERAL
- B08B7/00—Cleaning by methods not provided for in a single other subclass or a single group in this subclass
- B08B7/0035—Cleaning by methods not provided for in a single other subclass or a single group in this subclass by radiant energy, e.g. UV, laser, light beam or the like
- B08B7/0042—Cleaning by methods not provided for in a single other subclass or a single group in this subclass by radiant energy, e.g. UV, laser, light beam or the like by laser
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/047—Probabilistic or stochastic networks
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Health & Medical Sciences (AREA)
- Optics & Photonics (AREA)
- Mathematical Physics (AREA)
- Computational Linguistics (AREA)
- Probability & Statistics with Applications (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Studio Devices (AREA)
Abstract
The embodiment of the invention provides a method and a device for previewing a laser cleaning effect, wherein the method comprises the following steps: acquiring a picture of a region to be cleaned; inputting the picture of the area to be cleaned and the cleaning parameter into a preset generation confrontation network model, and outputting a preview picture corresponding to the cleaning parameter; and the generation countermeasure network model is obtained after training according to the picture before cleaning, the corresponding cleaning parameter and the cleaning result picture corresponding to the picture before cleaning as samples. The method realizes the generation of the cleaned preview picture according to the picture of the area to be cleaned and the laser parameter, thereby being beneficial to evaluating the reasonability of the current cleaning parameter according to the cleaned preview picture and further avoiding the problem of time consumption and cost caused by unreasonable laser parameter setting.
Description
Technical Field
The invention relates to the technical field of laser application, in particular to a method and a device for previewing a laser cleaning effect.
Background
The laser cleaning technology is a technological process of irradiating the surface of a workpiece with high-energy laser beams to instantaneously evaporate or peel off dirt, rust spots, coatings and the like on the surface so as to achieve cleaning. The laser cleaning mechanism is mainly based on that pollutants on the surface of an object absorb laser energy, or are vaporized and volatilized, or are heated and expanded instantly to overcome the adsorption force of the surface to particles, so that the laser cleaning can adapt to the cleaning of various surface pollutants, the environmental pollution is very small, and the substrate can not be damaged.
The material to be cleaned can show different cleaning effects after being cleaned by different laser parameters, under the common condition, when the laser parameters are too small, the expected effect can be achieved only by needing to carry out laser cleaning for many times, and when the laser parameters are too large, the substrate of the cleaned object can be damaged after one-time cleaning, so that time and cost are inevitably consumed. If the cleaning effect can be previewed after the cleaning parameters and the picture of the surface of the object to be cleaned are given, and the setting of the parameters is further adjusted according to the previewing result, the problems can be well solved.
Therefore, whether the preview of the cleaning effect can be realized after the cleaning parameters and the picture of the surface of the object to be cleaned are given without any prior knowledge is a problem to be solved urgently at present.
Disclosure of Invention
In order to solve the above problems, embodiments of the present invention provide a method and an apparatus for previewing a laser cleaning effect.
In a first aspect, an embodiment of the present invention provides a method for previewing a laser cleaning effect, including: acquiring a picture of a region to be cleaned; inputting the picture of the area to be cleaned and the cleaning parameter into a preset generation confrontation network model, and outputting a preview picture corresponding to the cleaning parameter; and the generation countermeasure network model is obtained after training according to the picture before cleaning, the corresponding cleaning parameter and the cleaning result picture corresponding to the picture before cleaning as samples.
In a second aspect, an embodiment of the present invention provides a laser cleaning effect preview apparatus, including: the acquisition module is used for acquiring a picture of a region to be cleaned; the processing module is used for inputting the picture of the area to be cleaned and the cleaning parameter into a preset confrontation network generation model and outputting a preview picture corresponding to the cleaning parameter; and the generation countermeasure network model is obtained after training according to the picture before cleaning, the corresponding cleaning parameter and the cleaning result picture corresponding to the picture before cleaning as samples.
In a third aspect, an embodiment of the present invention provides a laser cleaning system, including: the laser cleaning device also comprises a camera and a computer; the camera is connected with a computer, and the computer is connected with the laser cleaning device; the camera is used for acquiring pictures of an area to be cleaned; the computer comprises a laser cleaning effect previewing device of the second aspect of the invention, and is used for outputting a previewed picture corresponding to the cleaning parameter and receiving the cleaning parameter selected according to the previewed picture; the laser cleaning device is used for receiving cleaning parameters which are sent by a computer and selected according to the preview picture, and performing laser cleaning.
In a fourth aspect, an embodiment of the present invention provides an electronic device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, where the processor executes the computer program to implement the steps of the method for previewing the laser cleaning effect according to the first aspect of the present invention.
In a fifth aspect, an embodiment of the present invention provides a non-transitory computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the steps of the laser cleaning effect previewing method according to the first aspect of the present invention.
According to the laser cleaning effect previewing method and device provided by the embodiment of the invention, the picture of the area to be cleaned and the cleaning parameter are input into the preset generation confrontation network model, the preview picture corresponding to the cleaning parameter is output, and the generated confrontation network model is obtained after training according to the picture before cleaning, the corresponding cleaning parameter and the cleaning result picture corresponding to the picture before cleaning as a sample, so that the preview picture after cleaning is generated according to the picture of the area to be cleaned and the laser parameter, the rationality of the current cleaning parameter can be evaluated according to the preview picture after cleaning, and the problem of time consumption and cost caused by unreasonable laser parameter setting is avoided.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.
Fig. 1 is a flowchart of a laser cleaning effect preview method according to an embodiment of the present invention;
fig. 2 is a structural diagram of a laser cleaning effect preview device according to an embodiment of the present invention;
fig. 3 is a schematic physical structure diagram of an electronic device according to an embodiment of the present invention;
fig. 4 is a structural diagram of a laser cleaning system according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The laser cleaning technology is a supplement and an extension of the traditional cleaning method at present, and has wide application prospect due to the inherent advantages. The laser cleaning has the following characteristics:
1. no contact, no damage to the substrate. Because there is no direct physical contact during cleaning, no physical damage is caused to the workpiece.
2. The applicability is wide, and the cleaning effect is good. The laser has good directionality, can accurately position the workpiece part, thereby performing selective operation, being not limited by the surface appearance of the workpiece and being suitable for various cleaning objects; the cleaning process can be accurately controlled, and the required cleaning effect is achieved.
3. Convenient and flexible use and simple operation. The laser cleaning handheld series equipment is movable, very convenient to carry, light in working head and capable of being operated for a long time; the automatic cleaning equipment can be arranged in motion mechanisms such as a motion platform and a robot, stable automatic cleaning is realized, and the operation is simple.
4. Green and environment-friendly, and high cost performance. Only electric energy is consumed in the cleaning process, and the long-term cleaning cost is low.
In order to solve the problem that unreasonable setting of laser parameters in the existing laser cleaning technology may cause time and cost consumption, the embodiment of the invention provides a laser cleaning effect previewing method. And generating a cleaned effect picture according to the area to be cleaned and the laser parameters. It should be noted that the "plural groups" in the following embodiments of the invention are two or more groups, unless otherwise specified.
Fig. 1 is a flowchart of a method for previewing a laser cleaning effect according to an embodiment of the present invention, and as shown in fig. 1, the embodiment of the present invention provides a method for previewing a laser cleaning effect, including:
101, obtaining a picture of a region to be cleaned.
In 101, a picture of an area to be cleaned, i.e. a picture of the area to be cleaned, may be captured by a camera. A camera is arranged beside the laser head, the camera is used for collecting pictures of an area to be cleaned, when people want to clean a certain area, the area to be cleaned is shot by the camera, and the collected pictures are transmitted to a computer. In the embodiment of the invention, the picture of the area to be cleaned, which is shot by a camera, is obtained.
And 102, inputting the picture of the area to be cleaned and the cleaning parameter into a preset generation confrontation network model, and outputting a preview picture corresponding to the cleaning parameter.
In 102, the preset generated confrontation network model is obtained by training the picture before cleaning, the corresponding cleaning parameter and the cleaning result picture corresponding to the picture before cleaning as samples. And cleaning the picture before cleaning in the sample according to the corresponding cleaning parameters to obtain a cleaning result picture. And taking the combination of the picture before cleaning and the corresponding cleaning parameter of the known cleaning result picture as a sample. After the generated confrontation network model is established, training is carried out through a large number of samples, and therefore the preset generated confrontation network model is obtained. Wherein, the cleaning parameters can be set by a user or randomly generated by a computer.
Subsequently acquired pictures of the area to be cleaned and corresponding cleaning parameters are input into a preset generation confrontation network model, and preview pictures corresponding to the cleaning parameters can be quickly and accurately obtained. By generating the confrontation network model, the preview picture of the to-be-cleaned area after being cleaned by using the corresponding cleaning parameter is obtained according to the picture and the cleaning parameter of the to-be-cleaned area, so that whether the cleaning parameter is reasonable or not can be determined according to the preview picture.
According to the laser cleaning effect previewing method provided by the embodiment of the invention, the picture of the area to be cleaned and the cleaning parameter are input into the preset generation confrontation network model, the preview picture corresponding to the cleaning parameter is output, and the generated confrontation network model is obtained after training according to the picture before cleaning, the corresponding cleaning parameter and the cleaning result picture corresponding to the picture before cleaning as a sample, so that the preview picture after cleaning is generated according to the picture of the area to be cleaned and the laser parameter, the rationality of the current cleaning parameter can be evaluated according to the preview picture after cleaning, and the problem of time consumption and cost caused by unreasonable laser parameter setting is avoided.
Based on the content of the foregoing embodiment, as an optional embodiment, there are multiple sets of cleaning parameters of the pictures in the region to be cleaned, and after outputting the preview picture corresponding to the cleaning parameters, the method further includes: and selecting cleaning parameters meeting preset conditions according to the preview pictures corresponding to the multiple groups of cleaning parameters for laser cleaning.
In the embodiment of the invention, the picture of the area to be cleaned and a plurality of groups of laser cleaning parameters are used as the input of the network model, and the plurality of groups of laser parameters can be randomly generated by a computer. And respectively inputting the combination of the picture to be cleaned and each laser cleaning parameter into a preset generation confrontation network model, thereby obtaining a plurality of preview pictures. And selecting the cleaning parameter with the best preview effect from the multiple groups of cleaning parameters according to the multiple preview pictures and the preset conditions required to be met by the preview effect. And inputting the cleaning parameters with the best cleaning effect into a laser cleaning device to realize laser cleaning.
According to the laser cleaning effect previewing method provided by the embodiment of the invention, the cleaning parameters meeting the preset conditions are selected according to the previewing pictures corresponding to the multiple groups of cleaning parameters for laser cleaning, so that the better cleaning parameters are quickly determined, and the accuracy of the cleaning result is ensured.
Based on the content of the foregoing embodiment, as an optional embodiment, the generating the countermeasure network model specifically includes: multi-conditional generation confrontation network model (MC-GAN). The network model takes the additional auxiliary information as multiple conditions, and can convert the area image before cleaning into a corresponding image after cleaning according to specific parameters.
Based on the content of the foregoing embodiment, as an optional embodiment, inputting the picture of the area to be cleaned and the cleaning parameter into a preset generation confrontation network model, and outputting a preview picture corresponding to the cleaning parameter includes: inputting the picture of the area to be cleaned and the cleaning parameters into a down-sampling layer for generating the confrontation network model, utilizing the down-sampling layer to perform feature extraction on the picture of the area to be cleaned and the cleaning parameters, and outputting two-dimensional feature vectors corresponding to the picture of the area to be cleaned and the cleaning parameters; inputting the two-dimensional feature vector to a residual network layer of the generated countermeasure network model; and inputting the output result of the residual error network layer into an upper sampling layer, and outputting a preview picture corresponding to the cleaning parameter.
Specifically, the processing flow of inputting the picture of the area to be cleaned and the cleaning parameters into the generation of the confrontation network model can be realized by the following method:
firstly, inputting and inputting the combination of the picture of the area to be cleaned and the parameter set of the cleaning parameter from an input layer of the network, and extracting the characteristics of the combination of the picture of the area to be cleaned and the parameter set of the cleaning parameter through a down-sampling layer. Secondly, inputting the extracted features into a residual error network layer of the countermeasure network model, wherein the residual error network layer can be composed of a plurality of residual error network modules. And then, inputting the output result of the residual error network layer into a down-sampling layer, thereby obtaining a preview picture corresponding to the cleaning parameter.
According to the laser cleaning effect preview method provided by the embodiment of the invention, the features are extracted through the preset down-sampling layer for generating the confrontation network model, and after the residual error network layer is input, the preview picture is generated through the upper adoption layer, so that the accuracy of the generated preview picture is ensured.
Based on the content of the foregoing embodiment, as an optional embodiment, before inputting the picture of the area to be cleaned and the cleaning parameter into a preset generation countermeasure network model, the method further includes: and taking the combination of each picture before cleaning, the cleaning parameter and the cleaning result picture as a training sample, thereby obtaining a plurality of training samples, and training the antagonistic network model by utilizing the plurality of training samples.
Inputting the picture of the area to be cleaned and the cleaning parameters into a preset countermeasure network model before generating the countermeasure network model, and training the neural network, so as to obtain the preset neural network model capable of obtaining a preview picture according to the picture of the cleaning area and the corresponding parameters, wherein the specific steps are as follows:
firstly, obtaining a plurality of samples with determined cleaning results, obtaining a picture before cleaning and used cleaning parameters corresponding to each sample in the plurality of samples, and taking the picture of the cleaning results as a discrimination standard for generating the training of the confrontation network model.
Secondly, each picture before cleaning, a cleaning parameter corresponding to the picture before cleaning and a cleaning result picture corresponding to the cleaning parameter are combined to be used as a training sample, and therefore a plurality of training samples are obtained. Inputting the picture before cleaning in each sample and the cleaning parameters corresponding to the picture before cleaning into the constructed generation confrontation network model, adjusting the relevant parameters of the generation confrontation network model according to the output result and the cleaning result picture of the sample as the discrimination standard for training the generation confrontation network model, realizing the training process of the generation confrontation network model, and thus obtaining the preset generation confrontation network model.
According to the laser cleaning effect previewing method provided by the embodiment of the invention, each picture sample before cleaning, the cleaning parameter corresponding to each picture before cleaning and the cleaning result picture corresponding to the cleaning parameter are obtained, so that a plurality of training samples are obtained, the countermeasure network model is trained by utilizing the plurality of training samples, and therefore, the accurate previewing picture can be obtained for the picture of the area to be cleaned and the cleaning parameter which are input into the countermeasure network model.
Based on the content of the foregoing embodiments, as an alternative embodiment, training the reactive network model by using a plurality of training samples includes: inputting the combination of the picture before cleaning and the cleaning parameter of any sample into the generation confrontation network model, and generating a preview picture through the generator for generating the confrontation network model; judging the probability that the generated preview picture is a cleaning result picture by using the discriminator for generating the confrontation network model; and if the probability that the preview picture generated by the generator is judged to be the cleaning result picture by the discriminator is 1/2, finishing the training of the generated confrontation network model.
When the generation of the antagonistic network model is trained, the picture before cleaning in the N samples is recorded as:
the N groups of laser parameters are as follows: p1,…,PNAnd cleaning the picture:
The training objective is to establish a mapping function for the regions before and after washing:
generator G at given XpreAnd P, generating a preview picture Xpost. Inputting preview picture X by discriminator DpostOr G (X)preP, z), it is discriminated whether the input picture is the real input or generated by the generator.
When the discriminator is used for discrimination, the combination of the picture before cleaning and the picture after cleaning is input into the discriminator for generating the confrontation network model, the discriminator comprises a convolution layer, the convolution layer comprises a plurality of convolution networks and is used for carrying out feature extraction on the combination of the picture before cleaning and the picture after cleaning, and finally the discriminator outputs the probability that the picture after cleaning is generated by the generator.
Through multiple training, parameters are adjusted to approach the mapping function f, so that the discriminator D cannot distinguish whether the input washed picture is a real input picture or a picture generated by the generator, namely the probability that the preview picture generated by the generator is a washing result picture is judged to be 1/2, and the network training is finished.
After the network training is completed, the generator can generate a preview picture from the input parameters and the picture before cleaning, as follows:
Xv＝G(Xpre,P,z)。
fig. 2 is a structural diagram of a laser cleaning effect preview apparatus according to an embodiment of the present invention, and as shown in fig. 2, the laser cleaning effect preview apparatus includes: an acquisition module 201 and a processing module 202. The acquiring module 201 is configured to acquire a picture of an area to be cleaned; the processing module 202 is configured to input the picture of the area to be cleaned and the cleaning parameter into a preset confrontation network generation model, and output a preview picture corresponding to the cleaning parameter; and the generation countermeasure network model is obtained after training according to the picture before cleaning, the corresponding cleaning parameter and the cleaning result picture corresponding to the picture before cleaning as samples.
The picture of the area to be cleaned, namely the picture of the area to be cleaned, can be acquired through the camera. A camera is arranged beside the laser head, the camera is used for collecting pictures of an area to be cleaned, when people want to clean a certain area, the area to be cleaned is shot by the camera, and the collected pictures are transmitted to a computer. In the embodiment of the present invention, the obtaining module 201 obtains a picture of an area to be cleaned, which is obtained by shooting with a camera.
The preset generated confrontation network model is obtained by training a picture before cleaning, a corresponding cleaning parameter and a cleaning result picture corresponding to the picture before cleaning as samples. And cleaning the picture before cleaning in the sample according to the corresponding cleaning parameters to obtain a cleaning result picture. And taking the combination of the picture before cleaning and the corresponding cleaning parameter of the known cleaning result picture as a sample. After the generated confrontation network model is established, training is carried out through a large number of samples, and therefore the preset generated confrontation network model is obtained. Wherein, the cleaning parameters can be set by a user or randomly generated by a computer.
Subsequently acquired pictures of the area to be cleaned and corresponding cleaning parameters are input into the processing module 202, a preset generation confrontation network model is arranged in the processing module 202, and preview pictures corresponding to the cleaning parameters can be quickly and accurately obtained. By generating the confrontation network model, the preview picture of the to-be-cleaned area after being cleaned by using the corresponding cleaning parameter is obtained according to the picture and the cleaning parameter of the to-be-cleaned area, and whether the cleaning parameter is reasonable or not is determined according to the preview picture.
The device embodiment provided in the embodiments of the present invention is for implementing the above method embodiments, and for details of the process and the details, reference is made to the above method embodiments, which are not described herein again.
According to the laser cleaning effect previewing device provided by the embodiment of the invention, the picture of the area to be cleaned and the cleaning parameter are input into the preset generation confrontation network model, the preview picture corresponding to the cleaning parameter is output, and the generated confrontation network model is obtained after training according to the picture before cleaning, the corresponding cleaning parameter and the cleaning result picture corresponding to the picture before cleaning, so that the effect picture after cleaning is generated according to the area to be cleaned and the laser parameter, and the problem of time consumption and cost caused by unreasonable laser parameter setting is solved.
Fig. 3 is a schematic entity structure diagram of an electronic device according to an embodiment of the present invention, and as shown in fig. 3, the electronic device may include: a processor (processor)301, a communication Interface (communication Interface)302, a memory (memory)303 and a bus 304, wherein the processor 301, the communication Interface 302 and the memory 303 complete communication with each other through the bus 304. The communication interface 302 may be used for information transfer of an electronic device. Processor 301 may call logic instructions in memory 303 to perform a method comprising: acquiring a picture of a region to be cleaned; inputting the picture of the area to be cleaned and the cleaning parameter into a preset generation confrontation network model, and outputting a preview picture corresponding to the cleaning parameter; and the generation countermeasure network model is obtained after training according to the picture before cleaning, the corresponding cleaning parameter and the cleaning result picture corresponding to the picture before cleaning as samples.
In addition, the logic instructions in the memory 303 may be implemented in the form of software functional units and stored in a computer readable storage medium when the logic instructions are sold or used as independent products. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the above-described method embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
An embodiment of the present invention further provides a laser cleaning system, and fig. 4 is a structural diagram of the laser cleaning system provided in the embodiment of the present invention, and as shown in fig. 4, the system includes: a laser cleaning device 403, a camera 401, and a computer 402; the camera 401 is connected with a computer 402, and the computer 402 is connected with a laser cleaning device 403; the camera 401 is used for acquiring pictures of an area to be cleaned; the computer 402 comprises the laser cleaning effect previewing device of the embodiment of the device, and is used for outputting the preview picture corresponding to the cleaning parameter and receiving the cleaning parameter selected according to the preview picture; the laser cleaning device 403 is configured to receive the cleaning parameters selected according to the preview picture sent by the computer 402, and perform laser cleaning.
The system comprises three parts: the first part is a camera 401 for image acquisition, which may be a camera module integrating camera functions; the second part is a computer 402 which can provide laser parameter generation, generate different cleaning effect previews through input images and different parameters, receive selection input, and select and input an optimal cleaning effect graph and corresponding laser parameters; the third part is a laser cleaning device 403. The picture of the area to be cleaned is acquired by a camera on the laser head and sent to the computer 402, after the preview picture is calculated, the parameter selection of a user can be received, and the laser head is controlled to clean the area to be cleaned according to the corresponding laser parameter.
The camera 401: the position of the camera 401 can be set according to the requirement, for example, a camera is arranged beside the laser head, and the function of the camera is to collect pictures of the area to be cleaned.
The computer 402: multiple groups of laser cleaning parameters can be generated within a reasonable range. The computer 402 includes the laser cleaning effect preview apparatus of the above embodiment, and can calculate and obtain a corresponding cleaned effect map and corresponding laser parameters by using a trained deep network model. If the camera 401 is arranged near the laser head, before cleaning, a picture of the area to be cleaned is acquired by the camera 401 on the laser head. And then, the acquired picture before cleaning and the expected cleaning parameters are used as the input of the deep network, and the output of the deep network model is a corresponding cleaned effect picture corresponding to the cleaning parameters.
Laser cleaning apparatus 403: the computer 401 transmits the cleaning parameters corresponding to the selected preview image with a good effect to the laser cleaning device 403, and the laser cleaning device 403 applies the laser cleaning parameters transmitted by the computer 401 to the laser head to clean the region.
The laser cleaning system provided by the embodiment of the invention can obtain the laser cleaning effect in a priori manner, adjust the parameters in real time and obtain the optimal cleaning effect and the corresponding laser parameters. The calculated optimal parameters can be cleaned at one time to obtain the expected cleaning effect, the cleaning times of the same area are effectively reduced, and the substrate in the cleaning area is not damaged, so that time and cost are saved in the true sense.
In another aspect, an embodiment of the present invention further provides a non-transitory computer-readable storage medium, on which a computer program is stored, where the computer program is implemented by a processor to perform the method provided by the foregoing embodiments, for example, including: acquiring a picture of a region to be cleaned; inputting the picture of the area to be cleaned and the cleaning parameter into a preset generation confrontation network model, and outputting a preview picture corresponding to the cleaning parameter; and the generation countermeasure network model is obtained after training according to the picture before cleaning, the corresponding cleaning parameter and the cleaning result picture corresponding to the picture before cleaning as samples.
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 (8)
1. A laser cleaning effect previewing method is characterized by comprising the following steps:
acquiring a picture of a region to be cleaned;
inputting the picture of the area to be cleaned and the cleaning parameter into a preset generation confrontation network model, and outputting a preview picture corresponding to the cleaning parameter; the generation of the confrontation network model is a multi-condition generation confrontation network model;
the generation countermeasure network model is obtained after training according to the picture before cleaning, the corresponding cleaning parameter and the cleaning result picture as samples; the laser parameters include: outputting power W, frequency F and speed S;
inputting the picture of the area to be cleaned and the cleaning parameter into a preset generation confrontation network model, and outputting a preview picture corresponding to the cleaning parameter, wherein the preview picture comprises the following steps:
inputting the picture of the area to be cleaned and the cleaning parameters into a down-sampling layer for generating the confrontation network model, and performing feature extraction on the picture of the area to be cleaned and the cleaning parameters by using the down-sampling layer;
inputting the extracted features into a residual network layer of the generated countermeasure network model;
and inputting the output result of the residual error network layer into an upper sampling layer, and outputting a preview picture corresponding to the cleaning parameter.
2. The method according to claim 1, wherein there are multiple sets of cleaning parameters for the pictures of the region to be cleaned, and after the outputting of the preview picture corresponding to the cleaning parameters, the method further comprises:
and selecting cleaning parameters meeting preset conditions according to the preview pictures corresponding to the multiple groups of cleaning parameters for laser cleaning.
3. The method of claim 1, wherein before inputting the picture of the area to be cleaned and the cleaning parameters into a preset generation countermeasure network model, the method further comprises:
acquiring a plurality of picture samples before cleaning, and cleaning parameters corresponding to each picture sample before cleaning and a cleaning result picture corresponding to the cleaning parameters;
and taking the combination of each picture before cleaning, the cleaning parameter and the cleaning result picture as a training sample so as to obtain a plurality of training samples, and training the generated confrontation network model by using the plurality of training samples.
4. The method of claim 3, wherein training the generative confrontation network model with the plurality of training samples comprises:
inputting the combination of the picture before cleaning and the cleaning parameter of any sample into the generation confrontation network model, and generating a preview picture through the generator for generating the confrontation network model;
judging the probability that the generated preview picture is a cleaning result picture by using the discriminator for generating the confrontation network model;
and if the probability that the preview picture generated by the generator is judged to be the cleaning result picture by the discriminator is 1/2, finishing the training of the generated confrontation network model.
5. A laser cleaning effect preview device, comprising:
the acquisition module is used for acquiring a picture of a region to be cleaned;
the processing module is used for inputting the picture of the area to be cleaned and the cleaning parameter into a preset confrontation network generation model and outputting a preview picture corresponding to the cleaning parameter; the generation of the confrontation network model is a multi-condition generation confrontation network model;
the processing module is specifically configured to:
inputting the picture of the area to be cleaned and the cleaning parameters into a down-sampling layer for generating the confrontation network model, and performing feature extraction on the picture of the area to be cleaned and the cleaning parameters by using the down-sampling layer;
inputting the extracted features into a residual network layer of the generated countermeasure network model;
inputting the output result of the residual error network layer into an upper sampling layer, and outputting a preview picture corresponding to the cleaning parameter;
the generation countermeasure network model is obtained after training according to the picture before cleaning, the corresponding cleaning parameters and the cleaning result picture corresponding to the picture before cleaning as samples; the laser parameters include: output power W, frequency F and rate S.
6. A laser cleaning system comprises a laser cleaning device and is characterized by further comprising a camera and a computer;
the camera is connected with a computer, and the computer is connected with the laser cleaning device;
the camera is used for acquiring pictures of an area to be cleaned;
the computer comprises the laser cleaning effect previewing device of claim 5, and is used for outputting a preview picture corresponding to the cleaning parameters and receiving the cleaning parameters selected according to the preview picture;
the laser cleaning device is used for receiving cleaning parameters which are sent by a computer and selected according to the preview picture, and performing laser cleaning.
7. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor executes the program to implement the steps of the laser cleaning effect preview method according to any one of claims 1 to 4.
8. A non-transitory computer readable storage medium, on which a computer program is stored, wherein the computer program, when executed by a processor, implements the steps of the laser cleaning effect preview method according to any one of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910575724.1A CN110245459B (en) | 2019-06-28 | 2019-06-28 | Laser cleaning effect previewing method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910575724.1A CN110245459B (en) | 2019-06-28 | 2019-06-28 | Laser cleaning effect previewing method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110245459A CN110245459A (en) | 2019-09-17 |
| CN110245459B true CN110245459B (en) | 2021-06-01 |
Family
ID=67890193
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910575724.1A Active CN110245459B (en) | 2019-06-28 | 2019-06-28 | Laser cleaning effect previewing method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110245459B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111103301B (en) * | 2019-12-30 | 2023-04-07 | 上海锐戎激光焊接技术有限公司 | Laser cleaning quality detection method |
| CN111420938B (en) * | 2020-04-28 | 2022-03-15 | 株洲国创轨道科技有限公司 | Intelligent laser cleaning method and device for multiple laser heads |
| DE102020207578A1 (en) * | 2020-06-18 | 2021-12-23 | BSH Hausgeräte GmbH | System with a dishwasher, method for operating a dishwasher and computer program product |
| CN113409279B (en) * | 2021-06-24 | 2024-07-05 | 北京车和家信息技术有限公司 | Method, device, equipment and medium for evaluating effect of laser radar cleaning system |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102245340A (en) * | 2008-10-10 | 2011-11-16 | Jp赛席尔联合股份有限公司 | Laser machining systems and methods with vision correction and/or tracking |
| CN108492258A (en) * | 2018-01-17 | 2018-09-04 | 天津大学 | A kind of radar image denoising method based on generation confrontation network |
| CN109272055A (en) * | 2018-10-22 | 2019-01-25 | 西安云鑫电子科技有限公司 | Away rust by laser method and device based on machine learning |
| CN109345441A (en) * | 2018-10-19 | 2019-02-15 | 上海唯识律简信息科技有限公司 | A kind of de-watermarked method and system of image based on generation confrontation network |
| CN109447141A (en) * | 2018-10-22 | 2019-03-08 | 西安云鑫电子科技有限公司 | Away rust by laser method and device based on machine learning |
| CN109492764A (en) * | 2018-10-24 | 2019-03-19 | 平安科技（深圳）有限公司 | Training method, relevant device and the medium of production confrontation network |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| NL2018518B1 (en) * | 2017-03-15 | 2018-09-24 | P Laser N V | Pulsed laser device for cleaning or treating a surface |
| CN107392255B (en) * | 2017-07-31 | 2020-06-12 | 深圳先进技术研究院 | Generation method and device of minority picture sample, computing equipment and storage medium |
-
2019
- 2019-06-28 CN CN201910575724.1A patent/CN110245459B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102245340A (en) * | 2008-10-10 | 2011-11-16 | Jp赛席尔联合股份有限公司 | Laser machining systems and methods with vision correction and/or tracking |
| CN108492258A (en) * | 2018-01-17 | 2018-09-04 | 天津大学 | A kind of radar image denoising method based on generation confrontation network |
| CN109345441A (en) * | 2018-10-19 | 2019-02-15 | 上海唯识律简信息科技有限公司 | A kind of de-watermarked method and system of image based on generation confrontation network |
| CN109272055A (en) * | 2018-10-22 | 2019-01-25 | 西安云鑫电子科技有限公司 | Away rust by laser method and device based on machine learning |
| CN109447141A (en) * | 2018-10-22 | 2019-03-08 | 西安云鑫电子科技有限公司 | Away rust by laser method and device based on machine learning |
| CN109492764A (en) * | 2018-10-24 | 2019-03-19 | 平安科技（深圳）有限公司 | Training method, relevant device and the medium of production confrontation network |
Non-Patent Citations (3)
| Title |
|---|
| Ben Mills,et al.Predictive capabilities for laser machining via a neural network.《Optics Express》.2018,第26卷(第13期), * |
| Predictive capabilities for laser machining via a neural network;Ben Mills,et al;《Optics Express》;20180625;第26卷(第13期);第17245-17253页 * |
| 激光加工在线监测技术研究进展;丁烨等;《中国科学:物理学 力学 天文学》;20190327;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110245459A (en) | 2019-09-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110245459B (en) | Laser cleaning effect previewing method and device | |
| CN110238131B (en) | Laser cleaning parameter generation method and device | |
| Shi et al. | Deep networks for compressed image sensing | |
| Xiang et al. | Single-image de-raining with feature-supervised generative adversarial network | |
| US9521391B2 (en) | Settings of a digital camera for depth map refinement | |
| CN110689500B (en) | Face image processing method and device, electronic equipment and storage medium | |
| Ma et al. | Learning to jointly generate and separate reflections | |
| CN108427961B (en) | Synthetic aperture focusing imaging depth evaluation method based on convolutional neural network | |
| CN111951313B (en) | Image registration method, device, equipment and medium | |
| CN107610108A (en) | Image processing method and device | |
| CN105825228B (en) | Image-recognizing method and device | |
| CN108668069A (en) | A kind of image background weakening method and device | |
| CN111091093A (en) | Method, system and related device for estimating number of high-density crowds | |
| Rasouli et al. | Visual saliency improves autonomous visual search | |
| Liao et al. | Weakly supervised instance segmentation using hybrid networks | |
| CN106101831B (en) | Video vectorization method and device | |
| CN110336948A (en) | A kind of method, apparatus, storage medium and electronic device that presumptive area focuses | |
| Wang et al. | A parametric model for describing the correlation between single color images and depth maps | |
| Chan et al. | Improving the reliability of phase detection autofocus | |
| CN116862779A (en) | Real image self-supervision denoising method and system | |
| CN112102149A (en) | Figure hair style replacing method, device, equipment and medium based on neural network | |
| CN113888501A (en) | Non-reference image quality evaluation method based on attention positioning network | |
| CN115953821A (en) | Virtual face image generation method and device and electronic equipment | |
| CN111369612A (en) | Three-dimensional point cloud image generation method and equipment | |
| KR102042078B1 (en) | Operating method of a system for reconstucting 3-d shapes using neural network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB03 | Change of inventor or designer information |
Inventor after: He Jun Inventor after: Zhang Yinghui Inventor after: Sun Bo Inventor after: Yu Lejun Inventor before: He Jun Inventor before: Sun Bo Inventor before: Yu Lejun Inventor before: Zhang Yinghui |
|
| CB03 | Change of inventor or designer information | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |