CN117245228A

CN117245228A - Workpiece identification and flexible processing device and method

Info

Publication number: CN117245228A
Application number: CN202311546599.4A
Authority: CN
Inventors: 刘家云; 安辉
Original assignee: Guangdong Muqing Laser Intelligent Equipment Co ltd
Current assignee: Guangdong Muqing Laser Intelligent Equipment Co ltd
Priority date: 2023-11-20
Filing date: 2023-11-20
Publication date: 2023-12-19
Anticipated expiration: 2043-11-20
Also published as: CN117245228B

Abstract

The invention belongs to the technical field of laser marking machines, and particularly relates to a workpiece identification and flexible processing device and method in the field of laser marking, wherein the device comprises a conveying assembly, a multi-axis motion platform, a laser assembly, a visual identification assembly and an exhaust assembly; the conveying assembly corresponds to at least one processing station and conveys a tray on which a workpiece to be processed is placed; the laser component is arranged in the processing station and is positioned above the conveying component; the visual identification component is arranged at one end of the laser component, which faces the tray; the exhaust assembly is arranged between the laser assembly and the conveying assembly. According to the invention, the number, shape, placement angle and machining surface position of the to-be-machined parts are effectively identified through the visual identification component, the orientation of the laser component is flexibly adjusted through the multi-axis motion platform according to the gesture information of different to-be-machined parts, the processing requirements of various to-be-machined parts are met, and the applicability of the device is improved.

Description

Workpiece identification and flexible processing device and method

Technical Field

The invention belongs to the technical field of laser marking machines, and particularly relates to a workpiece identification and flexible processing device and method in the field of laser marking.

Background

A laser marking machine is a device that performs marking, imprinting, engraving, etc. operations using a laser technique. The method utilizes the laser beam to process the surface of the workpiece, and realizes the permanent marking of the surface of the object by controlling the position and the power of the laser beam.

The existing laser marking machine is likely to face the problem of poor adaptability in use, when facing nonstandard to-be-machined parts, operators are required to adjust the preset machining position of the laser marking head according to each to-be-machined part, and the problems of high operation labor cost and low debugging efficiency caused by complicated debugging steps occur.

Disclosure of Invention

In order to overcome the defects of the prior art, the invention provides a device and a method for workpiece identification and flexible processing, which are used for solving the problems of the prior art.

The invention provides a workpiece identification and flexible processing device, which comprises a conveying component, a multi-axis motion platform, a laser component, a visual identification component and an exhaust component, wherein the conveying component is connected with the multi-axis motion platform;

the conveying assembly corresponds to at least one processing station and is used for conveying a tray on which a workpiece to be processed is placed to the processing station;

the laser component is arranged in the processing station through the multi-axis motion platform, the laser component is positioned above the conveying component, the irradiation range of the laser component covers the tray entering the processing station, and the laser component is used for processing workpieces to be processed in the tray;

The visual recognition component is arranged at one end of the laser component, which faces the tray, and the recognition visual field of the visual recognition component covers the tray entering the processing station, and the visual recognition component is used for acquiring the position information of a workpiece to be processed in the tray and conveying the position information to the multi-axis motion platform;

the exhaust assembly is arranged between the laser assembly and the conveying assembly and is used for absorbing smoke dust or waste gas generated by the laser assembly in processing operation.

In this scheme, can discern the quantity, shape, the angle of putting of waiting to the machined part that enters into the processing station, take the place of attitude information such as machined surface position effectively, through multiaxis motion platform can be according to the attitude information of the different waiting to the machined part nimble adjustment orientation etc. of laser instrument subassembly to adapt to the processing demand of multiple waiting to the machined part, the suitability of equipment that promotes, simultaneously, through setting up exhaust subassembly laser instrument with between the conveying assembly for the absorption smoke and dust or the waste gas that laser instrument subassembly produced in the processing operation avoid smoke and dust waste gas escape, avoid environmental pollution and the influence of smoke and dust to other equipment, avoid waste gas to cause the influence to construction operation personnel health simultaneously, are more friendly to construction operation personnel.

In one preferable scheme of the invention, the visual recognition component comprises a visual processing module, a mounting bracket, a height recognition module, a first shooting module and a second shooting module, wherein the height recognition module, the first shooting module and the second shooting module are arranged on the mounting bracket;

the mounting bracket is arranged in a hollow annular or rectangular structure and is used for avoiding interference with the laser component;

the height identification module is arranged on one surface of the visual processing module facing the conveying assembly and used for acquiring the height position information of the visual identification assembly relative to the tray;

the first shooting module faces the conveying assembly, is electrically connected with the visual processing module, and is used for shooting image information of all workpieces to be processed in the tray and conveying the image information to the visual processing module; the visual processing module is used for storing a first identification model which is used for identifying the distribution condition of a workpiece to be processed and dividing a plurality of identification areas;

the second shooting module faces the conveying assembly, is electrically connected with the vision processing module, and is used for shooting the image information of the identification area and conveying the image information to the vision processing module; the visual processing module is stored with a second recognition model which is used for recognizing the gesture condition of the workpiece to be processed in the area and generating a position adjustment instruction;

The first shooting module and the second shooting module are respectively provided with at least two shooting lenses, and the shooting lenses are used for shooting the workpieces to be machined in the tray from different angles through a plurality of cameras.

In the scheme, the mounting bracket is arranged to be of a hollow annular or rectangular structure, so that interference to a laser emergent path of the laser assembly is avoided, the identification precision of a workpiece to be processed can be improved, and errors caused by large-scale identification are avoided;

furthermore, a multi-camera system is built by utilizing a plurality of cameras or cameras, objects on the tray are observed at the same time from different angles, and more comprehensive object information can be obtained by fusing or superposing images of the cameras, so that the recognition precision is further improved.

In one preferable scheme of the invention, the multi-axis motion platform comprises an X-axis motion module, a Z-axis motion module and a Y-axis motion module;

the Z-axis moving module is arranged on the X-axis moving module, and the X-axis moving module drives the Z-axis moving module to move along the X-axis direction;

the Y-axis moving module is arranged on the Z-axis moving module, and the Z-axis moving module drives the Y-axis moving module to move along the Z-axis direction;

The laser component is arranged on the Y-axis moving module, and the Y-axis moving module drives the laser component to move along the Y-axis direction;

the X-axis moving module comprises an X-axis mounting seat, an X-axis sliding block and a first driving motor, wherein the X-axis mounting seat is provided with a first sliding groove, the X-axis sliding block is arranged in the first sliding groove, and the first driving motor is in sliding connection with the X-axis sliding block through a first transmission piece and is used for driving the X-axis sliding block to move along the X-axis direction;

the Z-axis moving module comprises a Z-axis mounting seat, a Z-axis sliding block and a second driving motor, wherein the Z-axis mounting seat is arranged on the X-axis sliding block, the Z-axis mounting seat is provided with a second sliding groove, the Z-axis sliding block is arranged in the second sliding groove, and the second driving motor is in sliding connection with the Z-axis sliding block through a second transmission piece and is used for driving the Z-axis sliding block to move along the Z-axis direction;

the Y-axis moving module comprises a Y-axis mounting seat, a Y-axis sliding block and a third driving motor, wherein the Y-axis mounting seat is arranged on the Z-axis sliding block, the Y-axis mounting seat is provided with a third sliding groove, the Y-axis sliding block is arranged in the third sliding groove, and the third driving motor is in sliding connection with the Y-axis sliding block through a third transmission piece and is used for driving the Y-axis sliding block to move along the Y-axis direction.

In this scheme, multiaxis motion platform includes X axle removal module, Z axle removal module and Y axle removal module, is used for respectively drive the laser instrument subassembly is along X axle, Z axle, Y axle direction removal for the processing demand of waiting the machined part of arbitrary coordinate in the adaptation three-dimensional coordinate.

In one preferable scheme of the invention, the multi-axis motion platform comprises an X-axis motion module, a Z-axis motion module, a Y-axis motion module and a rotation shaft module;

the rotating shaft module is arranged between the laser component and the Y-axis moving module, and drives the laser component to rotate along the direction of a rotating central shaft;

the Y-axis moving module comprises a Y-axis mounting seat, a Y-axis sliding block and a third driving motor, wherein the Y-axis mounting seat is arranged on the Z-axis sliding block, the Y-axis mounting seat is provided with a third sliding groove, the Y-axis sliding block is arranged in the third sliding groove, and the third driving motor is in sliding connection with the Y-axis sliding block through a third transmission piece and is used for driving the Y-axis sliding block to move along the Y-axis direction;

the rotating shaft module comprises a rotating sliding table and a locking assembly, the rotating sliding table is arranged on the Y-axis sliding block, the laser assembly is arranged on the rotating sliding table, and the rotating sliding table is used for driving the laser assembly to rotate along a rotating central shaft; the locking component is in transmission connection with the rotary sliding table, and the locking component is used for locking the rotary sliding table with rotation stopped in a preset rotation angle.

In this scheme, multiaxis motion platform includes X axle removal module, Z axle removal module, Y axle removal module and rotation axis module, is used for respectively driving the laser instrument subassembly is along X axle, Z axle, Y axle direction removal, and the rotation axis module is used for driving the laser instrument subassembly is along rotation center axis direction rotation for the processing demand on waiting to add the different angle inclined planes of work piece of arbitrary coordinate in the adaptation three-dimensional coordinate.

In one preferable scheme of the invention, the exhaust assembly comprises an exhaust pipeline, a mounting bracket and a dust collection pipe, wherein the dust collection pipe is of an annular or rectangular structure, a plurality of dust collection holes are formed in the surface of the dust collection pipe, and the projection of the dust collection pipe on the processing station surrounds the irradiation range of the laser assembly; the dust collection pipe is communicated with at least one exhaust pipeline, and the exhaust pipeline is arranged on one side of the processing station through a mounting support.

During the operation of the laser marking machine, a great amount of waste gas is generated, including the evaporation of material, the gas and smoke dust generated by combustion during laser processing

In one application scenario of the scheme, the main functions of the exhaust assembly comprise waste gas treatment and smoke dust treatment, wherein the waste gas treatment part is mainly connected with an exhaust pipeline due to the fact that a fan is used for forming negative pressure at a dust collection hole of a dust collection pipe communicated with the exhaust pipeline when the fan is in operation, so that the waste gas can be effectively absorbed and discharged, the cleanness and safety of an operation environment are ensured, and meanwhile, the smoke dust is rapidly absorbed and filtered, and is prevented from being diffused into the surrounding environment.

And furthermore, the dust collection pipe is arranged in an annular or rectangular structure, so that generated waste gas and smoke dust can be collected more effectively, and are prevented from being diffused in an operation area, so that the efficiency of an exhaust system is improved, accumulation of the dust collection pipe in the operation area is avoided, and the cleanliness and the neatness of the operation area are maintained.

In one preferable scheme of the invention, the laser assembly comprises a laser generator, an optical path module and a vibrating lens, wherein the optical path module is arranged at a laser emitting end of the laser generator; the vibration lens is arranged at the laser emission end of the light path module and is used for adjusting the movement track of laser emitted from the laser emission end of the vibration lens.

In one application scenario of the scheme, the laser generator generates laser radiation, the optical path module controls the transmission path and the characteristics of the laser beam, and the vibration lens is used for adjusting the movement track of the laser beam. The components work together, so that the generation, transmission and control of laser beams can be realized, and the requirements of various application fields on laser are met.

In one preferable mode of the invention, the conveying assembly can adopt any one of a double-speed chain conveying line, a roller conveying line and a chain plate conveying line.

The conveying component is mechanical equipment commonly used in industrial production, and is mainly used for moving materials such as raw materials, finished products or semi-finished products from one position to another position, and in one application scene of the scheme, the conveying component can adopt a double-speed chain conveying line, and the double-speed chain conveying line has the following advantages: 1. the conveying speed is high: the double-speed chain conveying line has higher conveying speed, and can reach the speed of hundreds of meters per minute; 2. the bearing capacity is strong: the bearing capacity of the double-speed chain conveying line is relatively strong, and the double-speed chain conveying line is suitable for conveying heavier and larger materials; 3. the stability is good: the double-speed chain conveying line adopts sprocket drive, has the characteristic of good stability, and is difficult to fail.

In one preferable scheme of the invention, the workpiece identification and flexible processing device further comprises a marking platform cabinet and a cabinet cover, wherein the marking platform cabinet is used for bearing and arranging the conveying component, the multi-axis motion platform, the exhaust component and the laser component, and the cabinet cover is arranged on a bearing surface of the marking platform cabinet and used for protecting a plurality of module devices arranged on the marking platform cabinet;

the two sides of the cabinet cover are provided with an inlet and an outlet for avoiding the conveying assembly; the machine cabinet cover is provided with a front door of the marking machine corresponding to one side of the processing station, and the front door of the marking machine is provided with an observation window which is used for observing the processing operation state of the laser component.

In this scheme, the rack cover is used for the interval processing station and external environment prevent to avoid the laser instrument subassembly to receive the interference of external illumination or other conditions in the operation in-process, simultaneously through the marking machine qianmen sets up the observation window, can be used for observing the processing operation state of laser instrument subassembly, the operation condition of the timely observation equipment of operating personnel of being convenient for.

Furthermore, the observation window is made of any one of laser protective glass, laser filter plate or coated glass.

In one preferred aspect of the present invention, there is also provided a workpiece recognition and flexible processing method, which can be used in the workpiece recognition and flexible processing apparatus of any one of the above aspects, the method comprising:

acquiring the height position of the tray;

recognizing the distribution condition of a workpiece to be processed and dividing a plurality of recognition areas;

recognizing the posture condition of a workpiece to be machined in the region and generating a position adjustment instruction;

the multi-axis motion platform technology executes a position adjustment instruction to adjust the position of the laser component, and the laser component processes a workpiece to be processed;

and absorbing smoke dust or waste gas generated in the marking process through the exhaust component.

In the scheme, the height position information of the visual recognition component relative to the tray is recognized through the height recognition module, so that the subsequent recognition region division and gesture state recognition are more accurate; the first shooting module for large-range shooting is used for carrying out position recognition on all the workpieces to be processed in the tray, so that a plurality of recognition areas which are respectively corresponding to one workpiece to be processed are divided, and then the second small-range shooting module with high precision is moved to the position above the single recognition area to be singly recognized for the corresponding workpiece to be processed, so that the gesture condition of the workpiece to be processed in the recognition area is recognized, a position adjustment instruction is generated, the recognition precision of the workpiece to be processed can be improved, and meanwhile, errors caused by large-range recognition are avoided;

furthermore, the first shooting module and the second shooting module are both provided with at least two shooting lenses, and the shooting lenses are used for shooting the to-be-machined pieces in the tray from different angles through a plurality of cameras, a multi-camera system is built by the aid of the cameras or the cameras, objects on the tray are observed at the same time from different angles, and through fusion or superposition of images of the cameras, more comprehensive object information can be obtained, so that identification accuracy is further improved.

Simultaneously, can discern effectively that enter into the quantity, shape, angle of putting, the attitude information such as processing face position of waiting of processing station, through multiaxis motion platform can be according to the attitude information of waiting the machined part of difference and adjust in a flexible way the orientation etc. of laser component to adapt to the processing demand of multiple waiting the machined part, the suitability of equipment that promotes, further, through setting up exhaust subassembly be in the laser component with between the conveying assembly for the absorption the smoke and dust or the waste gas that the laser component produced in processing operation avoid smoke and dust waste gas loss, avoid environmental pollution and the influence of smoke and dust to other equipment, avoid waste gas to cause the influence to construction operation personnel healthy simultaneously, be more friendly to construction operation personnel.

In one preferred embodiment of the present invention, the height position of the acquisition tray may be one or more of a structured light sensor identification, a laser depth sensor identification, and an infrared depth sensor identification.

In one preferable aspect of the present invention, the identifying the distribution situation of the workpiece to be processed and dividing the plurality of identification areas includes:

acquiring first image data of all to-be-machined parts in a tray;

Preprocessing the first image data and outputting a first preprocessed image;

inputting the first preprocessing image into a target detection algorithm model to obtain distribution coordinates of a plurality of workpieces to be processed;

and inputting the first preprocessed image and the distribution coordinates into an image segmentation model, and outputting a plurality of identification areas.

In the scheme, the first image data of all the workpieces to be processed in the tray are acquired through the first shooting module, and the first image data are preprocessed, such as operations of adjusting the image size, graying, denoising and the like, so that further analysis and processing can be performed.

In one of the application scenarios of the scheme, the YOLO (You Only Look Once) and SSD (Single Shot MultiBox Detector) algorithms are preferably used for constructing the target algorithm detection model, so that a relatively high recognition speed is realized and the performance requirement is reduced.

In one of the application scenarios of the scheme, the image segmentation model is preferably constructed by adopting the U-Net and FCN algorithms, and is suitable for semantic segmentation of static images (such as images of trays in predetermined processing stations), and each pixel in the images can be divided into different categories

In one preferable aspect of the present invention, the identifying the posture condition of the workpiece to be machined in the area and generating the position adjustment instruction include:

Acquiring second image data of a single workpiece to be machined in a single identification area;

the second image data is preprocessed and a second preprocessed image is output;

performing feature recognition on the second preprocessing image based on the trained neural network model to obtain a class label and gesture data of a to-be-machined piece in a recognition area;

and processing and predicting the category labels and the gesture data based on the trained prediction model to obtain a position adjustment instruction.

In this scheme, the second image data of the single workpiece to be machined in the single identification area is obtained through the second shooting module, and the second image data is preprocessed, for example, operations such as adjusting the image size, graying, denoising and the like are performed, so that further analysis and processing can be performed.

And inputting the preprocessed second image data, and predicting through a model to obtain a class label and gesture data of the to-be-machined part, wherein the class label can be a predefined type of the to-be-machined part, such as the type or shape of the part. The gesture data may include information of a position, a rotation angle, a size, etc. of the workpiece to be processed in the three-dimensional space.

And in the scheme, the machining angle correction value of the to-be-machined part is predicted by using the trained prediction model. And inputting class labels and posture data of the to-be-machined parts, and outputting corresponding machining angle correction values, namely position adjustment instructions, by the model. According to the position adjustment instruction, the laser component can be correspondingly adjusted to correct the machining angle, so that machining operation can be accurately performed on the surface to be machined of the workpiece.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to the structures shown in these drawings without inventive effort for a person skilled in the art.

FIG. 1 is a schematic view showing a workpiece recognizing and flexible processing apparatus according to an embodiment of the present invention;

FIG. 2 is a schematic diagram of a visual recognition assembly according to one embodiment of the present invention;

FIG. 3 is a schematic diagram of a multi-axis motion platform according to an embodiment of the present invention;

FIG. 4 is a schematic structural diagram of an X-axis mobile module according to an embodiment of the invention;

FIG. 5 is a schematic diagram showing a Y-axis moving module according to an embodiment of the invention;

FIG. 6 is a schematic diagram showing a Z-axis moving module according to an embodiment of the invention;

FIG. 7 is a schematic view of a multi-axis motion platform according to another embodiment of the present invention;

FIG. 8 is a schematic view showing the structure of an exhaust assembly according to one embodiment of the present invention;

FIG. 9 is a schematic diagram of a laser assembly according to one embodiment of the present invention;

fig. 10 is a schematic structural diagram of a marking platform cabinet according to an embodiment of the invention.

The reference numerals in the drawings are as follows:

100. workpiece recognition and flexible processing device;

110. a transport assembly;

120. a multi-axis motion platform;

121. an X-axis moving module; 1211. an X-axis mounting seat; 1212. an X-axis sliding block; 1213. a first driving motor; 1214. a first sliding groove;

122. a Z-axis moving module; 1221. a Z-axis mounting seat; 1222. a Z-axis sliding block; 1223. a second driving motor; 1224. a second sliding groove;

123. a Y-axis moving module; 1231. a Y-axis mounting seat; 1232. a Y-axis sliding block; 1233. a third driving motor; 1234. a third sliding groove;

124. a rotation shaft module;

130. a laser assembly; 131. a laser generator; 132. an optical path module; 133. vibrating the lens;

140. a visual recognition component;

141. a vision processing module;

142. a first shooting module;

143. a second shooting module;

150. an exhaust assembly; 151. an exhaust duct; 152. a fixed bracket; 153. a dust collection pipe;

160. marking a platform cabinet; 161. a cabinet cover; 162. front door of marking machine.

Detailed Description

The following description of the embodiments of the present invention will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are only some, but not all embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.

It should be noted that, if a directional indication (such as up, down, left, right, front, and rear … …) is involved in the embodiment of the present invention, the directional indication is merely used to explain the relative positional relationship, movement condition, etc. between the components in a specific posture, and if the specific posture is changed, the directional indication is correspondingly changed.

In addition, if there is a description of "first", "second", etc. in the embodiments of the present invention, the description of "first", "second", etc. is for descriptive purposes only and is not to be construed as indicating or implying a relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include at least one such feature. In addition, if "and/or" and/or "are used throughout, the meaning includes three parallel schemes, for example," a and/or B "including a scheme, or B scheme, or a scheme where a and B are satisfied simultaneously. In addition, the technical solutions of the embodiments may be combined with each other, but it is necessary to base that the technical solutions can be realized by those skilled in the art, and when the technical solutions are contradictory or cannot be realized, the combination of the technical solutions should be considered to be absent and not within the scope of protection claimed in the present invention.

Referring to fig. 1, an embodiment of the present invention provides a workpiece recognition and flexible processing apparatus 100, which includes a conveying assembly 110, a multi-axis motion platform 120, a laser assembly 130, a visual recognition assembly 140, and an exhaust assembly 150;

the conveying assembly 110 corresponds to at least one processing station, and the conveying assembly 110 is used for conveying a tray on which a workpiece to be processed is placed to the processing station;

the laser assembly 130 is arranged in the processing station through the multi-axis motion platform 120, the laser assembly 130 is positioned above the conveying assembly 110, the irradiation range of the laser assembly 130 covers the tray entering the processing station, and the laser assembly 130 is used for processing workpieces to be processed in the tray;

the visual recognition component 140 is disposed at one end of the laser component 130 facing the tray, the recognition field of the visual recognition component 140 covers the tray entering the processing station, and the visual recognition component 140 is configured to obtain the position information of the workpiece to be processed in the tray and convey the position information to the multi-axis motion platform 120;

the exhaust assembly 150 is disposed between the laser assembly 130 and the conveying assembly 110, and the exhaust assembly 150 is used for absorbing smoke dust or exhaust gas generated by the laser assembly 130 during the machining operation.

In this embodiment, the visual recognition component 140 may effectively recognize the number, shape, placement angle, and position of the machining surface of the workpiece to be machined entering the machining station, and the multi-axis motion platform 120 may flexibly adjust the direction of the laser component 130 according to the gesture information of different workpieces to be machined, so as to adapt to the machining requirements of multiple workpieces to be machined, and improve the applicability of the equipment, and meanwhile, the exhaust component 150 is disposed between the laser component 130 and the conveying component 110, so as to absorb smoke dust or waste gas generated by the laser component 130 in the machining operation, thereby avoiding the dissipation of smoke dust and waste gas, avoiding environmental pollution and the influence of smoke dust on other equipment, and avoiding the influence of waste gas on the physical health of construction operators.

Referring to fig. 2, in one preferred embodiment of the present invention, the visual recognition assembly 140 includes a visual processing module 141, a mounting bracket, and a height recognition module, a first photographing module 142, and a second photographing module 143 disposed on the mounting bracket;

the mounting bracket is provided in a hollow annular or rectangular configuration for avoiding interference with the laser assembly 130;

The height recognition module is arranged on one surface of the vision processing module 141 facing the conveying assembly 110, and is used for acquiring the height position information of the vision recognition assembly 140 relative to the tray;

the first shooting module 142 faces the conveying assembly 110, and the first shooting module 142 is electrically connected with the vision processing module 141, and is configured to shoot image information of all workpieces to be processed in the tray and convey the image information to the vision processing module 141; the vision processing module 141 stores a first recognition model, which is used for recognizing the distribution situation of the workpiece to be processed and dividing a plurality of recognition areas;

the second shooting module 143 faces the conveying assembly 110, and the second shooting module 143 is electrically connected with the vision processing module 141, and is configured to shoot the image information of the identification area and convey the image information to the vision processing module 141; the vision processing module 141 stores a second recognition model, where the second recognition model is used to recognize the posture condition of the workpiece to be processed in the region and generate a position adjustment instruction;

the first photographing module 142 and the second photographing module 143 are respectively provided with at least two photographing lenses for photographing the workpiece to be processed in the tray from different angles through a plurality of cameras.

In this embodiment, the height recognition module may be integrally disposed in the first photographing module 142, and the mounting bracket is configured to be a hollow ring or rectangular structure, so as to avoid interference with the laser exit path of the laser assembly 130, and at the same time, the height position information of the visual recognition assembly 140 relative to the tray is recognized by the height recognition module, so that the subsequent recognition area division and gesture state recognition are more accurate; and, the first shooting module 142 shooting in a large range is used for carrying out position recognition on all the workpieces to be processed in the tray, so that a plurality of recognition areas which individually correspond to one workpiece to be processed are divided, and then the second shooting module 143 shooting in a second small range with high precision is moved to the position above a single recognition area to individually recognize the corresponding workpiece to be processed, so that the recognition precision of the workpiece to be processed can be improved, and meanwhile, errors caused by large-range recognition are avoided;

further, the first photographing module 142 and the second photographing module 143 are both provided with at least two photographing lenses, and are configured to photograph the workpiece to be processed in the tray from different angles through a plurality of cameras, and construct a multi-camera system by using the plurality of cameras or the cameras, so that objects on the tray are observed from different angles at the same time, and more comprehensive object information can be obtained by fusing or overlapping the images of the plurality of cameras, thereby further improving the recognition accuracy.

Referring to fig. 3-6, in one preferred embodiment of the present invention, the multi-axis motion platform 120 includes an X-axis motion module 121, a Z-axis motion module 122, and a Y-axis motion module 123;

the Z-axis moving module 122 is disposed on the X-axis moving module 121, and the X-axis moving module 121 drives the Z-axis moving module 122 to move along the X-axis direction;

the Y-axis moving module 123 is disposed on the Z-axis moving module 122, and the Z-axis moving module 122 drives the Y-axis moving module 123 to move along the Z-axis direction;

the laser assembly 130 is disposed on the Y-axis moving module 123, and the Y-axis moving module 123 drives the laser assembly 130 to move along the Y-axis direction;

the X-axis moving module 121 includes an X-axis mounting seat 1211, an X-axis sliding block 1212, and a first driving motor 1213, where the X-axis mounting seat 1211 is provided with a first sliding groove 1214, the X-axis sliding block 1212 is disposed in the first sliding groove 1214, and the first driving motor 1213 is slidably connected with the X-axis sliding block 1212 through a first transmission member, and is used for driving the X-axis sliding block 1212 to move along the X-axis direction;

the Z-axis moving module 122 includes a Z-axis mounting base 1221, a Z-axis sliding block 1222, and a second driving motor 1223, where the Z-axis mounting base 1221 is disposed on the X-axis sliding block 1212, the Z-axis mounting base 1221 is provided with a second sliding groove 1224, the Z-axis sliding block 1222 is disposed in the second sliding groove 1224, and the second driving motor 1223 is slidably connected with the Z-axis sliding block 1222 through a second transmission member, and is used for driving the Z-axis sliding block 1222 to move along the Z-axis direction;

The Y-axis moving module 123 includes a Y-axis mounting base 1231, a Y-axis sliding block 1232, and a third driving motor 1233, where the Y-axis mounting base 1231 is disposed on the Z-axis sliding block 1222, the Y-axis mounting base 1231 is provided with a third sliding groove 1234, the Y-axis sliding block 1232 is disposed in the third sliding groove 1234, and the third driving motor 1233 is slidably connected with the Y-axis sliding block 1232 through a third transmission member, and is used for driving the Y-axis sliding block 1232 to move along the Y-axis direction.

In this embodiment, the multi-axis motion platform 120 includes an X-axis moving module 121, a Z-axis moving module 122, and a Y-axis moving module 123, which are respectively configured to drive the laser assembly 130 to move along the X-axis, Z-axis, and Y-axis directions, so as to adapt to the processing requirements of the workpiece to be processed in any coordinate in the three-dimensional coordinates.

Referring to fig. 7, in one preferred embodiment of the present invention, the multi-axis motion platform 120 includes an X-axis motion module 121, a Z-axis motion module 122, a Y-axis motion module 123, and a rotation axis module 124;

the rotation shaft module 124 is disposed between the laser assembly 130 and the Y-axis moving module 123, and the rotation shaft module 124 drives the laser assembly 130 to rotate along the rotation central axis direction;

The Y-axis moving module 123 includes a Y-axis mounting base 1231, a Y-axis sliding block 1232, and a third driving motor 1233, where the Y-axis mounting base 1231 is disposed on the Z-axis sliding block 1222, the Y-axis mounting base 1231 is provided with a third sliding groove 1234, the Y-axis sliding block 1232 is disposed in the third sliding groove 1234, and the third driving motor 1233 is slidably connected with the Y-axis sliding block 1232 through a third transmission member, and is used for driving the Y-axis sliding block 1232 to move along the Y-axis direction;

the rotation shaft module 124 includes a rotation sliding table and a locking assembly, the rotation sliding table is disposed on the Y-axis sliding block 1232, the laser assembly 130 is disposed on the rotation sliding table, and the rotation sliding table is used to drive the laser assembly 130 to rotate along a rotation central axis; the locking component is in transmission connection with the rotary sliding table, and the locking component is used for locking the rotary sliding table with rotation stopped in a preset rotation angle.

In this embodiment, the multi-axis motion platform 120 includes an X-axis moving module 121, a Z-axis moving module 122, a Y-axis moving module 123, and a rotation axis module 124, which are respectively configured to drive the laser assembly 130 to move along the X-axis, Z-axis, and Y-axis directions, and the rotation axis module 124 is configured to drive the laser assembly 130 to rotate along the rotation center axis direction, so as to adapt to the processing requirements of inclined surfaces of different angles of the workpiece to be processed in any coordinate in the three-dimensional coordinates.

Referring to fig. 8, in one preferred embodiment of the present invention, the exhaust assembly 150 includes an exhaust pipe 151, a fixing bracket 152, and a dust suction pipe 153, the dust suction pipe 153 is configured in a ring or rectangular structure, and a plurality of dust suction holes are formed on a surface of the dust suction pipe 153, and a projection of the dust suction pipe 153 on the processing station surrounds an irradiation range surrounding the laser assembly 130; the dust suction pipe 153 is communicated with at least one exhaust pipeline 151, and the exhaust pipeline 151 is arranged on one side of the processing station through a mounting support.

In one application scenario of the present embodiment, the main functions of the exhaust assembly 150 include exhaust gas treatment and smoke dust treatment, where the exhaust gas treatment portion is mainly connected to the exhaust pipe 151, so that when the fan is operated, a negative pressure is formed at the dust suction hole of the dust suction pipe 153 that is in communication with the exhaust pipe 151, so that the exhaust gas can be effectively absorbed and discharged, thereby ensuring the cleanness and safety of the operation environment, and simultaneously, the smoke dust can be quickly absorbed and filtered, so as to avoid the diffusion of the smoke dust into the surrounding environment.

And further, the dust collection pipe 153 is arranged in a ring-shaped or rectangular structure, so that generated exhaust gas and smoke dust can be collected more effectively, and are prevented from being diffused in an operation area, thereby improving the efficiency of the exhaust system, avoiding accumulation of the exhaust system in the operation area, and keeping the operation area clean and tidy.

Referring to fig. 9, in one preferred embodiment of the present invention, the laser assembly 130 includes a laser generator 131, an optical path module 132, and a vibration lens 133, wherein the optical path module 132 is disposed at a laser emitting end of the laser generator 131; the vibration lens 133 is disposed at the laser emitting end of the optical path module 132, and the vibration lens 133 is used for adjusting the movement track of the laser emitted from the laser emitting end.

In one application scenario of the present embodiment, a Laser Source 131 is a core component in a Laser, and is capable of generating excitation energy required by Laser light, and the Laser Source 131 may be classified into various types according to different operation principles, including a gas Laser, a solid state Laser, a semiconductor Laser, etc., and generates Laser radiation by exciting a medium (e.g., an activating gas, a solid state material, or a semiconductor structure).

The optical path module 132 (Optical Path Module) is an important component in the laser, and is located at the laser emitting end of the laser generator 131, where the optical path module 132 is mainly used to control and shape the transmission path of the laser beam, so as to ensure that the laser beam maintains high quality as much as possible during the transmission process, and the optical path module 132 generally includes optical elements such as lenses, polarizers, diffraction elements, etc., and by adjusting the positions and angles of these elements, the transmission direction, focusing degree and spatial characteristics of the laser beam can be changed.

The vibration lens 133 (Scanning Mirror) is located at the laser emitting end of the optical path module 132, and is mainly used for adjusting the movement track of the laser emitted from the laser, and the vibration lens 133 is generally composed of two mirrors, and by controlling the angles and the vibration frequencies of the two mirrors, the Scanning, the deflection, the focusing and other operations of the laser beam can be realized. The vibration lens 133 can precisely control the direction and position of the laser beam so that it can be accurately irradiated to the target area.

The laser generator 131 generates laser radiation, the optical path module 132 controls the transmission path and characteristics of the laser beam, and the vibration lens 133 is used to adjust the movement track of the laser beam. The components work together, so that the generation, transmission and control of laser beams can be realized, and the requirements of various application fields on laser are met.

In one preferred embodiment of the present invention, the conveying assembly 110 may employ any one of a double speed chain conveyor line, a roller conveyor line, and a link plate conveyor line.

The conveying assembly 110 is a mechanical device commonly used in industrial production, and is mainly used for moving materials such as raw materials, finished products or semi-finished products from one position to another, and in one application scenario of the embodiment, the conveying assembly 110 can adopt a double-speed chain conveying line, and the double-speed chain conveying line has the following advantages: 1. the conveying speed is high: the double-speed chain conveying line has higher conveying speed, and can reach the speed of hundreds of meters per minute; 2. the bearing capacity is strong: the bearing capacity of the double-speed chain conveying line is relatively strong, and the double-speed chain conveying line is suitable for conveying heavier and larger materials; 3. the stability is good: the double-speed chain conveying line adopts sprocket drive, has the characteristic of good stability, and is difficult to fail.

Referring to fig. 10, in one preferred embodiment of the present invention, the workpiece recognition and flexible processing apparatus 100 further includes a marking platform cabinet 160 and a cabinet cover 161, wherein the marking platform cabinet 160 is used for carrying and disposing the conveying assembly 110, the multi-axis motion platform 120, the air exhaust assembly 150 and the laser assembly 130, and the cabinet cover 161 is disposed on a carrying surface of the marking platform cabinet 160 and is used for protecting a plurality of module devices disposed on the marking platform cabinet 160;

The two sides of the cabinet cover 161 are provided with an inlet and an outlet for avoiding the conveying assembly 110; the cabinet cover 161 is provided with a front door 162 of the marking machine on one side corresponding to the processing station, and the front door 162 of the marking machine is provided with an observation window for observing the processing operation state of the laser assembly 130.

In this embodiment, the cabinet cover 161 is configured to space the processing station from the external environment, prevent the laser assembly 130 from being interfered by external illumination or other conditions during the operation process, and simultaneously set an observation window through the front door 162 of the marking machine, so that the machine cabinet cover can be used for observing the processing operation state of the laser assembly 130, so that an operator can observe the operation condition of the equipment in time.

There is also indicated in one of the preferred embodiments of the present invention a workpiece identification and flexible processing method, usable with the workpiece identification and flexible processing apparatus 100 of any of the above-described embodiments, the method comprising:

s10, acquiring the height position of a tray;

s20, recognizing the distribution condition of the to-be-machined parts and dividing a plurality of recognition areas;

S30, recognizing the posture condition of the workpiece to be machined in the region and generating a position adjustment instruction;

s40, executing a position adjustment instruction by using a multi-axis motion platform technology to adjust the position of the laser assembly, and processing a workpiece to be processed by the laser assembly;

s50, absorbing smoke dust or waste gas generated in the marking process through the exhaust assembly.

In this embodiment, the height position information of the visual recognition component 140 relative to the tray is recognized by the height recognition module, so that the subsequent recognition region division and gesture state recognition are more accurate; and, the first shooting module 142 shooting in a large range is used for carrying out position recognition on all the workpieces to be processed in the tray, so that a plurality of recognition areas which individually correspond to one workpiece to be processed are divided, and then the second small-range shooting module with high precision is moved to the position above the single recognition area to individually recognize the corresponding workpiece to be processed, so that the gesture condition of the workpiece to be processed in the recognition area is recognized, a position adjustment instruction is generated, the recognition precision of the workpiece to be processed can be improved, and meanwhile, errors caused by large-range recognition are avoided;

Simultaneously, can discern effectively that enter into the attitude information such as quantity, shape, angle of putting, substitution processing face position of waiting to add into the processing station, through multiaxis motion platform 120 can be according to the attitude information of waiting to add the work piece of difference and adjust in a flexible way orientation etc. of laser subassembly 130 to adapt to the processing demand of multiple waiting to add the work piece, the suitability of equipment that promotes, further, through setting up exhaust subassembly 150 be in between laser subassembly 130 with transport subassembly 110 is used for absorbing the smoke and dust or the waste gas that laser subassembly 130 produced in the processing operation, avoid smoke and dust waste gas escape, avoid environmental pollution and the influence of smoke and dust to other equipment, avoid waste gas to cause the influence to construction operation personnel health simultaneously, be more friendly to construction operation personnel.

In one preferred embodiment of the present invention, the height position of the acquisition tray may be one or more of structured light sensor identification, laser depth sensor identification, and infrared depth sensor identification.

In this embodiment, the acquisition of the height position of the tray may be performed by the height identification module integrally disposed in the first photographing module 142;

Specifically, the structured light sensor identification includes the steps of: (1) projecting structured light onto a tray; (2) The structured light is reflected by the tray surface to form a series of patterns; (3) The sensor measures the information of the shape, the size, the position and the like of the pattern; (4) calculating the tray height position based on the measurement result. The three-dimensional measurement with high precision can be realized by the identification of the structured light sensor; the measuring range is wider, and the method is suitable for objects with different shapes and materials; the data processing is simple and the speed is high.

The laser depth sensor identification includes the steps of: (1) generating a laser beam using a laser emitter; (2) The laser beam irradiates the tray and returns to the sensor after being reflected; (3) The sensor measures the time difference of the laser beam and calculates the height position of the tray. The laser depth sensor has high recognition measurement precision and can reach the sub-millimeter level; the measuring speed is high, and the measuring device is suitable for measuring high-speed objects; the long-distance measurement can be realized, and the method is suitable for measuring large objects.

The infrared depth sensor identification includes the steps of: (1) Using an infrared light emitter and a receiver to form a sensor; (2) The infrared light irradiates the tray and returns to the receiver after being reflected; (3) The sensor calculates the tray height position based on the intensity of the infrared light.

The infrared depth sensor is low in identification cost and suitable for large-scale application; meanwhile, the volume is small, and the device is easy to integrate into other equipment; and will not interfere with the object being measured.

In one preferred embodiment of the present invention, the identifying the distribution situation of the to-be-machined piece and dividing the plurality of identification areas includes:

s201, acquiring first image data of all to-be-machined parts in a tray;

s202, preprocessing the first image data and outputting a first preprocessed image;

s203, inputting the first preprocessing image into a target detection algorithm model to obtain distribution coordinates of a plurality of workpieces to be processed;

s204, inputting the first preprocessed image and the distribution coordinates into an image segmentation model, and outputting a plurality of identification areas.

In this embodiment, the first image data of all the workpieces to be processed in the tray is acquired by the first shooting module 142, and the first image data is preprocessed, for example, operations such as adjusting the image size, graying, denoising, etc., so as to further analyze and process.

In one application scenario of the present embodiment, the target detection algorithm model may be created based on any one of Faster R-CNN, yolo (You Only Look Once), SSD (Single Shot MultiBox Detector), where:

Faster R-CNN is an object detection algorithm based on a regional advice network (Region Proposal Network, RPN) that divides the object detection problem into two phases: first, candidate frames are generated using RPN, and then classified and position regressed using a classifier, fast R-CNN is excellent in accuracy and position accuracy, but relatively slow.

YOLO is a fast and efficient object detection algorithm that uses the concept of "You Only Look Once" to transform the object detection problem into a regression problem, which breaks the input image into multiple grids and predicts whether each grid contains an object and the bounding box and class of the object, which has real-time performance but may be relatively weak in terms of small object detection and positioning accuracy.

SSD is another object detection algorithm based on single multi-box detection, and predicts bounding boxes of different sizes and corresponding categories on feature maps of different scales simultaneously, and SSD can detect objects on multiple scales, so that the detection effect on small targets is good, and meanwhile the speed is high.

In one of the application scenarios of the present embodiment, the YOLO (You Only Look Once) and SSD (Single Shot MultiBox Detector) algorithms are preferably used to construct the target algorithm detection model, so as to achieve a faster recognition speed and reduce performance requirements.

In one of the application scenarios of this embodiment, the image segmentation model may be constructed using a semantic segmentation algorithm (e.g., U-Net, FCN, etc.) or an instance segmentation algorithm (e.g., mask R-CNN), where:

U-Net is a common semantic segmentation algorithm that consists of an encoder and a decoder. The encoder is responsible for extracting features from the input image, while the decoder gradually upsamples the extracted features and fuses the extracted features with features of the corresponding level of the encoder, and finally generates a semantic segmentation result at the pixel level. U-Net generally achieves good results in the fields of medical image segmentation and the like.

The full convolution network (Fully Convolutional Network, FCN) is a classical semantic segmentation algorithm. FCNs transform traditional classification networks into a network structure that can output pixel level predictions by removing the fully connected layers. It can effectively map an input image to an output feature map, where each pixel corresponds to a class label.

Mask R-CNN is an example segmentation algorithm that extends on the basis of a target detection algorithm, and can detect not only the position and bounding box of an object, but also generate a pixel-level segmentation Mask for each object. Mask R-CNN generates segmentation Mask by adding a branch network, so that the algorithm has better performance in detection and segmentation tasks.

In one of the application scenarios of this embodiment, the image segmentation model is preferably constructed by using U-Net and FCN algorithms, and is suitable for semantic segmentation of static images (such as images of trays in predetermined processing stations), and each pixel in the images can be classified into different categories

In one preferred embodiment of the present invention, the identifying the posture of the workpiece to be machined in the area and generating the position adjustment command include:

s301, acquiring second image data of a single workpiece to be machined in a single identification area;

s302, preprocessing the second image data and outputting a second preprocessed image;

s303, carrying out feature recognition on the second preprocessing image based on the trained neural network model to obtain a class label and gesture data of a to-be-machined piece in a recognition area;

s304, processing and predicting the category labels and the gesture data based on the trained prediction model to obtain a position adjustment instruction.

In this embodiment, the second image data of the single workpiece to be machined in the single identification area is acquired by the second shooting module 143, and the second image data is preprocessed, for example, operations of adjusting the image size, graying, denoising, etc., so as to further analyze and process.

The neural network model is a Convolutional Neural Network (CNN) model, and any one of the neural network model and the convolutional neural network model can be selected as VGG, resNet, inception, and a self-defined model structure can be designed according to specific requirements. CNN models are typically composed of a convolution layer for extracting image features, a pooling layer for reducing feature dimensions, a fully connected layer for classification, and so on. The trained model can be used for real-time identification application of the to-be-machined piece. And inputting the preprocessed second image data, and predicting through a model to obtain a class label and gesture data of the to-be-machined part, wherein the class label can be a predefined type of the to-be-machined part, such as the type or shape of the part. The gesture data may include information of a position, a rotation angle, a size, etc. of the workpiece to be processed in the three-dimensional space.

And in this embodiment, the prediction model may use a machine learning algorithm, such as a regression model, to establish the relationship between the class labels and pose data of the workpieces to be processed and the processing angle. The training data of the model may be known class labels and pose data and corresponding machining angles. And predicting the machining angle correction value of the to-be-machined workpiece by using the trained prediction model. And inputting class labels and posture data of the to-be-machined parts, and outputting corresponding machining angle correction values, namely position adjustment instructions, by the model. According to the position adjustment command, the laser assembly 130 can be correspondingly adjusted to correct the machining angle, so as to correctly perform machining operation on the surface to be machined of the workpiece.

The foregoing description is only of the preferred embodiments of the present invention and is not intended to limit the scope of the invention, and all equivalent structural changes made by the description of the present invention and the accompanying drawings or direct/indirect application in other related technical fields are included in the scope of the invention.

Claims

1. The workpiece identification and flexible processing device is characterized by comprising a conveying assembly, a multi-axis motion platform, a laser assembly, a visual identification assembly and an exhaust assembly;

2. The workpiece recognition and flexible processing device according to claim 1, wherein the visual recognition component comprises a visual processing module, a mounting bracket, and a height recognition module, a first shooting module and a second shooting module which are arranged on the mounting bracket;

3. The workpiece recognition and flexible processing apparatus of claim 1, wherein the multi-axis motion stage comprises an X-axis motion module, a Z-axis motion module, and a Y-axis motion module;

4. The workpiece recognition and flexible processing apparatus of claim 1, wherein the multi-axis motion stage comprises an X-axis motion module, a Z-axis motion module, a Y-axis motion module, a rotation axis module;

5. The workpiece recognition and flexible processing device according to claim 1, wherein the exhaust assembly comprises an exhaust pipeline, a mounting bracket and a dust collection pipe, the dust collection pipe is of an annular or rectangular structure, a plurality of dust collection holes are formed in the surface of the dust collection pipe, and the projection of the dust collection pipe on the processing station surrounds an irradiation range surrounding the laser assembly; the dust collection pipe is communicated with at least one exhaust pipeline, and the exhaust pipeline is arranged on one side of the processing station through a mounting support;

and/or the laser component comprises a laser generator, an optical path module and a vibrating lens, wherein the optical path module is arranged at a laser emitting end of the laser generator; the vibration lens is arranged at the laser emission end of the light path module and is used for adjusting the movement track of laser emitted from the laser emission end of the vibration lens;

and/or, the conveying assembly can adopt any one of a double-speed chain conveying line, a roller conveying line and a chain plate conveying line.

6. The workpiece recognition and flexible processing apparatus of any of claims 1-5, further comprising a marking platform cabinet for carrying the transport assembly, the multi-axis motion platform, the exhaust assembly, and the laser assembly, and a cabinet cover disposed on a carrying surface of the marking platform cabinet for protecting a plurality of module devices disposed on the marking platform cabinet;

7. A workpiece identification and flexible processing method, applicable to the workpiece identification and flexible processing device of any one of claims 1 to 6, characterized in that the method comprises:

acquiring the height position of the tray;

8. The method of claim 7, wherein the height position of the acquisition tray is one or more of structured light sensor identification, laser depth sensor identification, and infrared depth sensor identification.

9. The method of workpiece recognition and flexible processing according to claim 7, wherein the recognizing the distribution of the workpieces to be processed and dividing a plurality of recognition areas includes:

acquiring first image data of all to-be-machined parts in a tray;

preprocessing the first image data and outputting a first preprocessed image;

10. The method of workpiece recognition and flexible processing according to claim 9, wherein recognizing the posture condition of the workpiece to be processed in the region and generating the position adjustment instruction includes: