Furniture production control method, terminal equipment and storage medium

Abstract

The invention is suitable for the technical field of furniture production, and provides a furniture production control method, terminal equipment and a storage medium, and the method comprises the following steps: determining a plurality of processing equipment on a furniture production line; acquiring original image data from the part, and acquiring original point cloud data from the part; screening target image data representing the parts from the original image data, and screening target point cloud data representing the parts from the original point cloud data; reading a configuration file of the processing equipment; loading a quality inspection program corresponding to the quality inspection label according to the similarity of the processing equipment on the processing part; starting a quality inspection program, and executing quality inspection operation on the part according to the target image data and / or the target point cloud data to obtain quality information; and if a quality inspection instruction is received when the quality information is displayed on the part model, processing the part according to the quality inspection instruction, and processing the quality inspection program according to the quality information and the quality inspection instruction. By adopting the method, the yield of furniture production can be effectively improved.

Description

Technical Field

[0001] The present application belongs to the field of furniture production technology, and in particular, relates to a furniture production control method, terminal equipment and storage medium. Background Art

[0002] With the development of the economy, furniture has gradually become a fast-moving consumer product, and furniture has shown changing trends such as small scale and rapid model replacement.

[0003] Therefore, the production equipment of furniture is constantly undergoing fine-tuning to adapt to the changing trends of furniture. Since fine-tuning will have a certain impact on the stability of the production equipment, it is easy to cause certain defects in the furniture during production.

[0004] At present, during furniture production, technical personnel are mainly responsible for quality inspection of furniture and checking whether the furniture has any defects.

[0005] However, some defects are relatively minor, which may lead to some defects being missed. Moreover, when the furniture models are changed, the defects also change, and the corresponding quality inspection methods also change, which makes the quality inspection have a certain learning cost, aggravates the situation of defects being missed, reduces the efficiency of furniture production, and leads to a low yield of furniture production. Summary of the invention

[0006] In view of this, the present application provides a furniture production control method, a terminal device and a storage medium to improve the yield of furniture production.

[0007] A first aspect of an embodiment of the present application provides a furniture production control method, comprising:

[0008] Determine a plurality of processing equipment on the furniture production line; each of the processing equipment is used to process parts of the furniture, and each of the processing equipment is equipped with a trigger, a camera and a radar;

[0009] When the processing equipment conveys the processed component to the crawler and the component activates the trigger, the camera is called to collect original image data from the component and the radar is called to collect original point cloud data from the component;

[0010] Comparing the original image data with the original point cloud data to select target image data representing the component from the original image data and to select target point cloud data representing the component from the original point cloud data;

[0011] Reading a configuration file of the processing equipment; the configuration file records one or more quality inspection labels;

[0012] Loading the quality inspection program corresponding to the quality inspection label according to the similarity of the processing equipment in processing the parts;

[0013] Starting the quality inspection program, performing quality inspection operations on the component according to the target image data and / or the target point cloud data, and obtaining quality information;

[0014] Constructing a three-dimensional part model for the component according to the target image data and the target point cloud data;

[0015] If a quality inspection instruction is received when the quality information is displayed on the part model, the component is processed according to the quality inspection instruction, and the quality inspection procedure is processed according to the quality information and the quality inspection instruction.

[0016] A second aspect of an embodiment of the present application provides a furniture production control device, comprising:

[0017] A processing equipment determination module is used to determine multiple processing equipment on the furniture production line; each of the processing equipment is used to process the parts of the furniture, and each of the processing equipment is equipped with a trigger, a camera and a radar;

[0018] A raw data acquisition module, used for simultaneously calling the camera to collect raw image data from the component and calling the radar to collect raw point cloud data from the component when the processing equipment conveys the processed component to the crawler and the component activates the trigger;

[0019] A target data screening module, used for comparing the original image data with the original point cloud data, so as to screen the target image data representing the component from the original image data and screen the target point cloud data representing the component from the original point cloud data;

[0020] A configuration file reading module, used to read a configuration file of the processing equipment; the configuration file records one or more quality inspection labels;

[0021] A quality inspection program loading module, used to load the quality inspection program corresponding to the quality inspection label according to the similarity of the processing equipment in processing the parts;

[0022] A component quality inspection module, used to start the quality inspection program, perform quality inspection operations on the component according to the target image data and / or the target point cloud data, and obtain quality information;

[0023] A part model building module, used to build a three-dimensional part model for the component according to the target image data and the target point cloud data;

[0024] The quality inspection processing module is used to process the component according to the quality inspection instruction if a quality inspection instruction is received when the quality information is displayed on the part model, and to process the quality inspection procedure according to the quality information and the quality inspection instruction.

[0025] A third aspect of an embodiment of the present application provides a terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the furniture production control method as described in the first aspect above is implemented.

[0026] A fourth aspect of an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the furniture production control method as described in the first aspect above is implemented.

[0027] A fifth aspect of the embodiments of the present application provides a computer program product. When the computer program product is run on a computer, the computer executes the furniture production control method described in the first aspect.

[0028] Compared with the prior art, the embodiments of the present application have the following beneficial effects:

[0029] In this embodiment, a plurality of processing equipment are determined on a furniture production line; each processing equipment is used to process furniture parts, and each processing equipment is equipped with a trigger, a camera and a radar; when the processing equipment transports the processed parts to the track and the parts activate the trigger, the camera is called to collect original image data from the parts, and the radar is called to collect original point cloud data from the parts; the original image data and the original point cloud data are compared with each other to filter target image data representing the parts from the original image data, and to filter target point cloud data representing the parts from the original point cloud data; the configuration file of the processing equipment is read; one or more quality inspection labels are recorded in the configuration file; the quality inspection program corresponding to the quality inspection label is loaded according to the similarity of the processing equipment on the processed parts; the quality inspection program is started, and quality inspection operations are performed on the parts according to the target image data and / or target point cloud data to obtain quality information; a three-dimensional part model is constructed for the parts according to the target image data and the target point cloud data; if a quality inspection instruction is received when the quality information is displayed on the part model, the parts are processed according to the quality inspection instruction, and the quality inspection program is processed according to the quality information and the quality inspection instruction. This embodiment combines visual image data and tactile point cloud data to construct a furniture quality inspection framework, implements a general furniture quality inspection solution, decouples furniture processing operations from quality inspection logic, and flexibly adjusts programs and labels within the furniture quality inspection framework. It can quickly follow up on quality inspection solutions when production equipment is slightly adjusted, thereby reducing the learning cost of quality inspection. Furthermore, visual image data and tactile point cloud data can detect tiny defects in furniture, and the quality inspection has high accuracy, greatly reducing the possibility of missed inspections, thereby effectively improving the efficiency of furniture production as a whole and improving the yield of furniture production. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the embodiments or prior art descriptions. Obviously, the drawings described below are only some embodiments of the present application, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0031] Figure 1 is a schematic diagram of a furniture production control method provided in an embodiment of the present application;

[0032] Figure 2 is a schematic diagram of a quality inspection network provided in an embodiment of the present application;

[0033] Figure 3 is a schematic diagram of a furniture production control device provided in an embodiment of the present application;

[0034] Figure 4 It is a schematic diagram of a terminal device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0035] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0036] The technical solution of the present application is described below through specific embodiments.

[0037] Reference Figure 1 , shows a schematic diagram of a furniture production control method provided by an embodiment of the present application, which may specifically include the following steps:

[0038] Step 101: determine a plurality of processing equipment on a furniture production line.

[0039] In this embodiment, one or more production lines are arranged in a production workshop, one of which can produce one or more models of furniture. A plurality of fully automatic or semi-automatic processing equipment are arranged on the production line, and each processing equipment is used to process furniture parts, such as pressing, cutting, punching, polishing, painting, etc.

[0040] Among them, furniture parts refer to the components of furniture. For example, the parts of a bed include the headboard, bed board, bed beams, bed feet, etc.

[0041] Each processing device is equipped with a trigger, a camera and a radar. The trigger is electrically connected to a controller, and the controller is electrically connected to the camera and the radar respectively.

[0042] Among them, the controller includes PLC (Programmable Logic Controller), the trigger can be activated by infrared, pressure, etc., the camera includes RGB (red, green, and blue) camera, etc., and the radar includes low-linearity laser radar, etc.

[0043] Cameras can provide visual image data and radars can provide tactile point cloud data. Both can perceive furniture components from different dimensions and provide a data basis for general quality inspection.

[0044] Step 102: When the processing equipment delivers the processed parts to the crawler and the parts activate the trigger, the camera is called to collect original image data from the parts and the radar is called to collect original point cloud data from the parts.

[0045] The status of each processing equipment can be monitored in systems such as EAP (Equipment Automation Programming). If a processing equipment completes processing a certain component of furniture, the processing equipment will transport the component to the track for quality inspection. At this time, when the component passes through the trigger of the processing equipment, the trigger is activated, indicating that the component has entered the perception range of the camera and radar. The trigger sends a trigger signal to the controller, and the controller sends an enable signal to the camera and radar at the same time. The camera collects image data from its perception range (including components) based on the enable signal, which is recorded as raw image data. The radar collects point cloud data from its perception range (including components) based on the enable signal, which is recorded as raw point cloud data.

[0046] Step 103 : Compare the original image data with the original point cloud data to select target image data representing the component from the original image data, and select target point cloud data representing the component from the original point cloud data.

[0047] In practical applications, the original image data and the original point cloud data are data that perceive the same component from different dimensions. The content of the two is the same, but there are certain differences in the data. Therefore, the original image data and the original point cloud data can be compared with each other to find the similarities and differences between the two. On the one hand, the pixel points that represent the components are filtered out from the original image data as the target image data. On the other hand, the (radar) points that represent the components are filtered out from the original point cloud data as the target point cloud data.

[0048] Then, the target image data and the target point cloud data focus more on the parts themselves, eliminating the influence of the background, and provide the object of attention for general quality inspection.

[0049] In one embodiment of the present application, step 103 may include the following steps:

[0050] Step 1031: read the conversion matrix calibrated between the camera and the radar.

[0051] In this embodiment, when installing the camera and the radar, the camera and the radar can be jointly calibrated to obtain a transformation matrix between the two. The transformation matrix represents the transformation relationship between the coordinate system of the camera and the coordinate system of the radar (such as translation relationship, rotation relationship, etc.).

[0052] Step 1032: Project the original point cloud data into the original image data according to the transformation matrix to obtain reconstructed image data.

[0053] In this embodiment, each (radar) point of the original point cloud data is multiplied by the transfer matrix to obtain each (radar) point located in the coordinate system of the camera, so as to project each (radar) point of the original point cloud data into the original image data. At this time, the original image data contains both the information of the pixel point (such as RGB) and the information of the (radar) point (such as reflection intensity, spatial coordinates, etc.). For the convenience of distinction, the original image data is recorded as reconstructed image data.

[0054] Step 1033: When querying the processing equipment for unprocessed parts, call the reference image data collected by the camera.

[0055] Generally speaking, the positions of processing equipment and its tracks are relatively fixed. Therefore, when the processing equipment is not processing parts, there are no parts on the tracks. At this time, the camera can be called to collect reference image data. At this time, the reference image data is a pure background.

[0056] Step 1034: Calculate differential image data between the original image data and the reference image data as first candidate image data.

[0057] In this embodiment, the differential image data between the original image data and the reference image data can be calculated and recorded as the first candidate image data. At this time, the background can be effectively eliminated, and the main content of the first candidate image data is the parts.

[0058] Since the crawler of the processing equipment is continuously running and its operation is usually periodic, there may be some background data in the differential image data. Therefore, when the processing equipment is not processing parts, the camera can be called to collect multiple frames of reference image data to cover the operation cycle of the crawler of the processing equipment.

[0059] If there are multiple frames of reference image data, the difference image data between the original image data and each frame of reference image data can be calculated;

[0060] The area of ​​the foreground data in each frame of differential image data is counted, and the area of ​​the foreground data in each frame of differential image data is compared.

[0061] Selecting the differential image data with the smallest area of ​​the foreground data to obtain the first candidate image data can effectively reduce the impact of the periodic operation of the crawler of the processing equipment, and the calculation is simple and fast.

[0062] Step 1035: construct a reference surface representing the track based on the original point cloud data.

[0063] In practical applications, parts are located on the tracks. Therefore, clustering algorithms such as DBSCAN (Density Based Spatial Clustering of Applications with Noise) can be used to cluster the original point cloud data to construct a reference surface to represent the tracks.

[0064] Step 1036: Filter the data located above the reference plane in the reconstructed image data to obtain second candidate image data.

[0065] In this embodiment, each data (pixel point, (radar) point) in the reconstructed image data is compared with the reference plane. If the data is above the reference plane, the data can be screened out to form the second candidate image data.

[0066] At this time, the periodic operation of the crawler of the processing equipment has a certain impact on the reference surface, so that the second candidate image data may contain some background data.

[0067] Step 1037: Take the intersection of the first candidate image data and the second candidate image data to obtain feature image data.

[0068] The periodic operation of the crawler of the processing equipment has different effects on the first candidate image data and the second candidate image data, that is, there may be a large difference in the background data included in the first candidate image data and the second candidate image data.

[0069] Since there are pixel points in the first candidate image data and there are pixel points in the second candidate image data, the intersection of the first candidate image data and the second candidate image data can be directly taken to obtain feature image data (including pixel points and (radar) points), which can effectively exclude background data and obtain relatively pure foreground data.

[0070] Step 1038: Determine that the pixel points in the feature image data constitute the target image data representing the component, and the points in the feature image data constitute the target point cloud data representing the component.

[0071] When the feature image data is relatively pure foreground data, on the one hand, it is determined that the pixel points in the feature image data can constitute the target image data representing the component, and on the other hand, the (radar) points in the feature image data can constitute the target point cloud data representing the component.

[0072] Step 104: Read the configuration file of the processing equipment.

[0073] In this embodiment, a configuration file can be set for the processing equipment based on the need for quality inspection of parts processed by the processing equipment, wherein one or more quality inspection labels are recorded in the configuration file, and the quality inspection label is information identifying the same type of quality inspection procedure, and the quality inspection procedure is a procedure for quality inspection of parts processed by the processing equipment.

[0074] When making minor adjustments to processing equipment or replacing new parts for processing, technicians can choose appropriate quality inspection procedures based on actual conditions, write their quality inspection labels into the configuration files, implement hot updates of the configuration files, and decouple the business (i.e., processing equipment processing parts) from the underlying logic (i.e., quality inspection procedures), allowing technicians to focus more on developing general quality inspection procedures.

[0075] Step 105 , loading the quality inspection program corresponding to the quality inspection label according to the similarity of the processing equipment on the processed parts.

[0076] In actual applications, the production workshop is a weak network environment. There is a relatively strict access mechanism between its internal LAN and the external Internet to ensure the network security of the production workshop. The stability of downloading data from the external Internet is poor, and download interruptions are prone to occur.

[0077] Taking into account the similarities in the operations of processing parts and components among processing equipment, their quality inspection procedures have a certain degree of universality. Therefore, the quality inspection procedures corresponding to the quality inspection labels used by other processing equipment can be loaded based on the similarities in the processing parts and components, thereby improving the efficiency of loading the quality inspection procedures corresponding to the quality inspection labels.

[0078] In one embodiment of the present application, step 105 may include the following steps:

[0079] Step 1051: read the process information set for the processing equipment for processing parts.

[0080] Read the process information of processing equipment settings and processing parts from the furniture production plan, processing equipment manual and other materials.

[0081] Step 1052: Filter multiple keywords from the process information.

[0082] In this embodiment, a natural language processing algorithm such as TextRank (text ranking) may be used to filter multiple keywords from the process information.

[0083] Step 1053: Convert multiple keywords into process vectors of processing equipment.

[0084] In this embodiment, one-hot, word2vec and other algorithms may be used to vectorize multiple keywords and thus convert them into process vectors of processing equipment.

[0085] Step 1054: Calculate the similarity between the process vector of the current processing equipment and the process vectors of other processing equipment.

[0086] In this embodiment, each processing equipment can be traversed, and algorithms such as cosine angle can be used to calculate the similarity between the process vector of the current processing equipment and the process vectors of other processing equipment (that is, all processing equipment except the current processing equipment).

[0087] Step 1055: If the similarity is greater than or equal to the preset first threshold, determine that other processing equipment is a candidate equipment.

[0088] The similarity is compared with a preset first threshold. If the similarity is greater than or equal to the first threshold, it means that the similarity between the process vector of the current processing equipment and the process vector of other processing equipment is high, and the other processing equipment is determined as a candidate equipment.

[0089] Step 1056: If the candidate device has been configured with a quality inspection program corresponding to the quality inspection tag, and the quality inspection program corresponding to the quality inspection tag has been running on the candidate device for a time period exceeding a preset second threshold, then the quality inspection program corresponding to the quality inspection tag is loaded from the candidate device.

[0090] If the candidate device has been configured with a quality inspection program corresponding to the quality inspection tag, the running time of the quality inspection program corresponding to the quality inspection tag on the candidate device can be counted, and the running time of the quality inspection program corresponding to the quality inspection tag on the candidate device can be compared with a preset second threshold.

[0091] If the length of time that the quality inspection program corresponding to the quality inspection label runs on the candidate device exceeds the second threshold, it means that the quality inspection program corresponding to the quality inspection label runs on the candidate device for a long time and the quality inspection program corresponding to the quality inspection label is a stable quality inspection program. Then, the quality inspection program corresponding to the quality inspection label is loaded from the candidate device, and the quality inspection program corresponding to the quality inspection label is configured to the current processing equipment.

[0092] In a specific implementation, if the current processing equipment is not configured with a quality inspection program corresponding to a quality inspection tag, the quality inspection program corresponding to the quality inspection tag with the highest similarity can be directly loaded from the candidate equipment.

[0093] If the current processing equipment has been configured with a quality inspection program corresponding to the quality inspection tag, the first version information of the quality inspection program corresponding to the quality inspection tag in the current processing equipment and the second version information of the quality inspection program corresponding to the quality inspection tag in the candidate equipment are queried respectively.

[0094] If the first version information is lower than the second version information and the quality inspection program corresponding to the quality inspection label in the current processing equipment is not prohibited from being upgraded to the second version information, one or more differential upgrade packages between the first version information and the second version information are downloaded from the candidate device, and one differential upgrade package is used to upgrade one version.

[0095] The quality inspection program corresponding to the quality inspection tag in the current processing equipment is upgraded using the differential upgrade package, that is, the quality inspection program corresponding to the quality inspection tag in the current processing equipment is upgraded from the first version information to the second version information.

[0096] If the first version information is higher than the second version information, and / or the quality inspection program corresponding to the quality inspection label in the current processing equipment is prohibited from being upgraded to the second version information (that is, the quality inspection program corresponding to the quality inspection label in the current processing equipment does not meet the preset performance conditions when upgraded to the second version information, and is not applicable to the quality inspection of the current parts), then the quality inspection program corresponding to the quality inspection label configured for the candidate equipment is ignored.

[0097] Step 106: Start the quality inspection program, perform quality inspection operations on the parts according to the target image data and / or the target point cloud data, and obtain quality information.

[0098] In practical applications, threads can be used to load quality inspection programs, perform quality inspection operations on parts based on target image data and / or target point cloud data, and obtain quality information of parts.

[0099] On the one hand, the quality inspection operations that can be supported by the single target image data include inspection of paint coverage, uniformity, foreign matter detection, etc.

[0100] On the other hand, the quality inspection operations that can be supported by the single target point cloud data include contour size detection, contour curvature detection, and so on.

[0101] On the other hand, the quality inspection operations supported by the combined target image data and target point cloud data include crack detection, defect detection, and the like.

[0102] In one embodiment of the present application, step 106 may include the following steps:

[0103] Step 10611, start the quality inspection procedure.

[0104] In this embodiment, the quality inspection program includes a quality inspection network constructed and trained based on deep learning and some other quality inspection logics. When the quality inspection program is started, the quality inspection network is started.

[0105] Among them, Figure 2As shown, the quality inspection network includes a first feature structure Feature_1, a second feature structure Feature_2, a first head structure Head_1 and a second head structure Head_2.

[0106] Step 10612: Convert the feature image data consisting of the target image data and the target point cloud data into grayscale image data.

[0107] In this embodiment, the characteristic image data composed of the target image data and the target point cloud data is read, and the color values ​​such as RGB of the pixel points in the characteristic image data are converted into grayscale values ​​under the condition of retaining the (radar) points to obtain grayscale image data.

[0108] Step 10613: Input the grayscale image data into the first feature structure to extract the common domain features.

[0109] In this embodiment, the grayscale image data is input into the first feature structure Feature_1 to extract the common domain features applicable to both the first head structure Head_1 and the second head structure Head_2.

[0110] In one design, Figure 2 As shown, the first feature structure Feature_1 includes a first convolution block ConvBlock_1 and a second convolution block ConvBlock_2. Both the first convolution block ConvBlock_1 and the second convolution block ConvBlock_2 encapsulate the layer structure related to convolution to realize feature downsampling, for example, convolution layer, BN (Batch Normalization) layer, ReLU (Rectified Linear Unit), pooling layer, and so on.

[0111] Correspondingly, the public domain features include the first basic features and the second basic features.

[0112] Then, the grayscale image data is input into the first convolution block ConvBlock_1 and a convolution operation is performed according to its structure to obtain the first basic feature.

[0113] The first basic feature is input into the second convolution block ConvBlock_2, and a convolution operation is performed according to its structure to obtain the second basic feature.

[0114] Step 10614: Input the public domain features into the first head structure to segment the crack data on the parts.

[0115] In this embodiment, the public domain features are input into the first head structure Head_1 to complete the semantic segmentation operation, thereby segmenting the data with the semantics of cracks on the parts and recording them as crack data, that is, the crack data represents the cracks appearing on the parts.

[0116] During the furniture making process, due to factors such as materials (especially wood) and craftsmanship (excessive temperature, excessive pressure, etc.), cracks may occur in various processes such as pressing, cutting, punching, and polishing. Cracks are a relatively common quality problem.

[0117] In one design, Figure 2 As shown, the first head structure Head_1 includes a first deconvolution block DeConvBlock_1 and a second deconvolution block DeConvBlock_2. Both the first deconvolution block DeConvBlock_1 and the second deconvolution block DeConvBlock_2 encapsulate layer structures related to deconvolution to achieve feature upsampling, such as deconvolution layer, BN layer, ReLU, pooling layer, and so on.

[0118] Furthermore, the structures of the first deconvolution block DeConvBlock_1 and the second deconvolution block DeConvBlock_2 may be the same or different, which is not limited in this embodiment.

[0119] Then, the second basic feature is input into the first deconvolution block DeConvBlock_1, and a deconvolution operation is performed according to its structure to obtain a first reconstructed feature.

[0120] The first reconstructed feature and the first basic feature are concatenated into a second reconstructed feature to realize skip connection, so as to increase the information content of the feature.

[0121] The second reconstructed feature is input into the second deconvolution block DeConvBlock_2, and a deconvolution operation is performed according to its structure to obtain the crack data on the component.

[0122] Step 10615: Input the public domain features into the second feature structure and convert them into private domain features.

[0123] In this embodiment, the public domain features are input into the second feature structure Feature_2 for dimension reduction or adjustment, and are converted into private domain features to better adapt to the task of target detection.

[0124] In one design, Figure 2As shown, the second feature structure Feature_2 includes a residual network ResNet, a pooling layer Pooling and a multilayer perceptron (MLP). Generally, the residual network ResNet is a shallow residual structure, such as 18 layers, 34 layers, etc. The pooling layer Pooling is ASPP (Atrous Spatial Pyramid Pooling). When faced with public domain features with a large amount of information, it can obtain a larger receptive field and extract more contextual information.

[0125] Then, concatenate the first basic feature and the second basic feature into the third basic feature.

[0126] The third basic feature is input into the residual network ResNet to extract the first intermediate feature according to its structure.

[0127] The first intermediate feature is input into the pooling layer and the pooling operation is performed according to its structure to obtain the second intermediate feature.

[0128] The second intermediate feature is input into the multi-layer perceptron MLP and mapped into the private domain feature.

[0129] Step 10616: Input the private domain features into the second head structure to detect defect data on the parts.

[0130] In this embodiment, the private domain features are input into the second head structure Head_2 to complete the target detection operation, thereby detecting data with the semantics of defects on the parts, which are recorded as defect data, that is, the defect data represents the defects on the parts.

[0131] During the furniture making process, due to factors such as materials (especially wood) and craftsmanship (such as blade shaking), defects such as dents and warping may occur in various processes such as pressing, cutting, punching and polishing. Defects are a relatively common quality problem.

[0132] This embodiment packages the quality inspection of two common quality problems, cracks and defects, into a task for a quality inspection network. Compared with using independent deep learning models to detect cracks and defects separately, it has higher versatility, lower resource occupancy, and lower probability of errors during operation. It is suitable for production workshops with limited LAN resources. In addition, managing a single quality inspection network in a production workshop can improve the ease of management.

[0133] In one design, Figure 2As shown, the second head structure Head_2 includes a third convolution block ConvBlock_3 and multiple fully connected layers (FC). The third convolution block ConvBlock_3 encapsulates the layer structure related to convolution to realize feature downsampling, for example, convolution layer, BN (Batch Normalization) layer, ReLU (Rectified Linear Unit), pooling layer, and so on.

[0134] Furthermore, the structures of the first convolution block ConvBlock_1, the second convolution block ConvBlock_2 and the third convolution block ConvBlock_3 may be the same or different, and this embodiment does not limit this.

[0135] Then, the private domain features are input into the third convolution block ConvBlock_3 and the convolution operation is performed according to its structure to obtain the target features.

[0136] The target features are sequentially input into multiple fully connected layers FC to map the defect data on the parts according to their structures.

[0137] In this embodiment, the quality inspection network implements a modular design. Each module is responsible for completing a specific task. Each module can be designed, trained and optimized individually. It can reuse third-party pre-trained structures and can also be fine-tuned and reconstructed according to the characteristics of the furniture. This improves the flexibility and interpretability of the quality inspection network, and also facilitates the debugging and optimization of the quality inspection network based on fine-tuning of furniture production.

[0138] Step 10617: Calculate the first abnormal index for the crack data and calculate the second abnormal index for the defect data.

[0139] In this embodiment, a first abnormality indicator is counted from multiple dimensions for the crack data, for example, the length and width of the minimum circumscribed rectangle of the crack data, the actual length of the crack data, the actual maximum width of the crack data, etc., and a second abnormality indicator is counted from multiple dimensions for the defect data, for example, the length and width of the defect data, the actual area of ​​the defect data, the shortest distance between the defect data and the edge, etc.

[0140] Step 10618: Query the first quality inspection requirement information and the second quality inspection requirement information set for the parts.

[0141] In this embodiment, according to the production design requirements of the furniture, first quality inspection requirement information can be set in multiple dimensions for cracks appearing on each component, and second quality inspection requirement information can be set in multiple dimensions for defects appearing on each component.

[0142] Step 10619: If the first abnormal indicator meets the first quality inspection requirement information, and the second abnormal indicator meets the second quality inspection requirement information, then the quality information of the component is determined to be qualified.

[0143] In this embodiment, the first abnormality indicator is compared with the corresponding first quality inspection requirement information, and the second abnormality indicator is compared with the corresponding second quality inspection requirement information.

[0144] If the first abnormal indicators all meet the corresponding first quality inspection requirement information, and the second abnormal indicators all meet the corresponding second quality inspection requirement information, it means that the cracks on the parts meet the quality inspection requirements and the defects on the parts meet the quality inspection requirements. Then the quality information of the parts is determined to be qualified and allowed to enter the next process.

[0145] Step 10620: If the first abnormal indicator does not meet the first quality inspection requirement information, and / or the second abnormal indicator does not meet the second quality inspection requirement information, the crack data, defect data and unqualified quality are set as the quality information of the component.

[0146] If the first abnormal indicators do not meet the corresponding first quality inspection requirement information, and / or the second abnormal indicators do not meet the corresponding second quality inspection requirement information, indicating that the cracks on the parts do not meet the quality inspection requirements and the defects on the parts do not meet the quality inspection requirements, the quality information of the parts is determined to be qualified and the next process is suspended.

[0147] Step 107: construct a three-dimensional part model for the component based on the target image data and the target point cloud data.

[0148] In this embodiment, the target image data and the target point cloud data can be combined to construct a high-precision three-dimensional part model for the component.

[0149] In the specific implementation, on the one hand, grid, voxelization and other algorithms can be called to use the target point cloud data to build a three-dimensional model backbone structure for the parts, and optimize the model backbone structure by smoothing, filling holes and other methods.

[0150] On the other hand, texture maps can be extracted from the target image data, for example, diffuse map, specular map, normal map, roughness map, and so on.

[0151] Input the texture map model backbone structure into the renderer under the framework such as Pytorch3D or Redner, render the texture map into the model backbone structure, and obtain a three-dimensional part model to improve the details and realism of the model.

[0152] Step 108: If a quality inspection instruction is received when the quality information is displayed on the part model, the part is processed according to the quality inspection instruction, and the quality inspection procedure is processed according to the quality information and the quality inspection instruction.

[0153] In this embodiment, the part model can be displayed, and the quality information can be displayed on the part model using visual elements. The technician can use zooming, translation, rotation and other methods to adjust the vision so as to verify the part and its quality information.

[0154] For example, when the quality information is qualified, a green quality inspection mark is displayed on the part model; when the quality information is crack data, defect data and unqualified quality, the crack data and defect data are highlighted in red on the part model, and an infrared quality inspection mark is displayed, and so on.

[0155] When the technician completes the verification, a quality inspection instruction can be triggered. At this time, the parts can be processed according to the instructions of the quality inspection instruction.

[0156] If the quality inspection instruction indicates that the quality inspection is successful, the parts are allowed to be transported to the next processing equipment on the production line for processing, wherein the sequence between the processing equipment can be set when arranging the plan for furniture production.

[0157] If the quality inspection instruction indicates that the quality inspection has failed, the component will be removed from the production line and will be prohibited from being transported to the next processing equipment for processing. At this time, the production engineer can repair the component. If the repair fails, the component will be scrapped.

[0158] In addition, the quality information and the quality inspection instructions may have the same quality inspection conclusions or opposite quality inspection conclusions. In this case, the performance of the quality inspection procedure can be evaluated based on the quality information and the quality inspection instructions, thereby processing the quality inspection procedure.

[0159] In a specific implementation, if the quality inspection program is upgraded from the first version information to the second version information, the quality information is compared with the quality inspection instruction to calculate multiple statistical indicators, such as recall rate, accuracy, precision, F1 value, etc.

[0160] Compare multiple statistical indicators to pre-set performance conditions.

[0161] If multiple statistical indicators meet the preset performance conditions, it means that the performance of the quality inspection program under the second version information meets the requirements, then it is determined that the quality inspection program under the second version information is valid for the quality inspection operation of the processing equipment, and the quality inspection program under the second version information is maintained to perform the quality inspection operation of the processing equipment.

[0162] If multiple statistical indicators fail to meet the preset performance conditions, indicating that the performance of the quality inspection program under the second version information does not meet the requirements, it is determined that the quality inspection program under the second version information is invalid for the quality inspection operation of the processing equipment, and the quality inspection program is rolled back from the second version information to the first version information, and the quality inspection operation of the processing equipment is performed using the quality inspection program under the first version information with more stable performance.

[0163] In this embodiment, a plurality of processing equipment are determined on a furniture production line; each processing equipment is used to process furniture parts, and each processing equipment is equipped with a trigger, a camera and a radar; when the processing equipment transports the processed parts to the track and the parts activate the trigger, the camera is called to collect original image data from the parts, and the radar is called to collect original point cloud data from the parts; the original image data and the original point cloud data are compared with each other to filter target image data representing the parts from the original image data, and to filter target point cloud data representing the parts from the original point cloud data; the configuration file of the processing equipment is read; one or more quality inspection labels are recorded in the configuration file; the quality inspection program corresponding to the quality inspection label is loaded according to the similarity of the processing equipment on the processed parts; the quality inspection program is started, and quality inspection operations are performed on the parts according to the target image data and / or target point cloud data to obtain quality information; a three-dimensional part model is constructed for the parts according to the target image data and the target point cloud data; if a quality inspection instruction is received when the quality information is displayed on the part model, the parts are processed according to the quality inspection instruction, and the quality inspection program is processed according to the quality information and the quality inspection instruction. This embodiment combines visual image data and tactile point cloud data to construct a furniture quality inspection framework, implements a general furniture quality inspection solution, decouples furniture processing operations from quality inspection logic, and flexibly adjusts programs and labels within the furniture quality inspection framework. It can quickly follow up on quality inspection solutions when production equipment is slightly adjusted, thereby reducing the learning cost of quality inspection. Furthermore, visual image data and tactile point cloud data can detect tiny defects in furniture, and the quality inspection has high accuracy, greatly reducing the possibility of missed inspections, thereby effectively improving the efficiency of furniture production as a whole and improving the yield of furniture production.

[0164] It should be noted that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0165] Reference Figure 3 , shows a schematic diagram of a furniture production control device provided in an embodiment of the present application, which may specifically include the following modules:

[0166] The processing equipment determination module 301 is used to determine a plurality of processing equipment on the furniture production line; each of the processing equipment is used to process the parts of the furniture, and each of the processing equipment is equipped with a trigger, a camera and a radar;

[0167] The raw data acquisition module 302 is used to call the camera to collect raw image data from the component and call the radar to collect raw point cloud data from the component when the processing equipment conveys the processed component to the crawler and the component activates the trigger;

[0168] A target data screening module 303 is used to compare the original image data with the original point cloud data to screen the target image data representing the component from the original image data and screen the target point cloud data representing the component from the original point cloud data;

[0169] A configuration file reading module 304 is used to read a configuration file of the processing equipment; the configuration file records one or more quality inspection labels;

[0170] A quality inspection program loading module 305 is used to load the quality inspection program corresponding to the quality inspection label according to the similarity of the processing equipment in processing the parts;

[0171] A component quality inspection module 306 is used to start the quality inspection program, perform quality inspection operations on the component according to the target image data and / or the target point cloud data, and obtain quality information;

[0172] A part model building module 307 is used to build a three-dimensional part model for the component according to the target image data and the target point cloud data;

[0173] The quality inspection processing module 308 is used to process the component according to the quality inspection instruction if a quality inspection instruction is received when the quality information is displayed on the part model, and to process the quality inspection procedure according to the quality information and the quality inspection instruction.

[0174] In one embodiment of the present application, the target data screening module 303 includes:

[0175] A conversion matrix reading module, used to read the conversion matrix calibrated between the camera and the radar;

[0176] An original point cloud data projection module, used for projecting the original point cloud data into the original image data according to the transformation matrix to obtain reconstructed image data;

[0177] A reference image data query module, used to query the reference image data collected by the camera when the processing equipment is not processing the component;

[0178] A first candidate image data calculation module, used for calculating differential image data between the original image data and the reference image data as first candidate image data;

[0179] A reference plane construction module, used to construct a reference plane representing the crawler according to the original point cloud data;

[0180] A second candidate image data screening module, used for screening data located above the reference plane in the reconstructed image data to obtain second candidate image data;

[0181] A feature image data generating module, used for taking the intersection of the first candidate image data and the second candidate image data to obtain feature image data;

[0182] The target data determination module is used to determine that the pixel points in the feature image data constitute the target image data representing the component, and the points in the feature image data constitute the target point cloud data representing the component.

[0183] In one embodiment of the present application, the first candidate image data calculation module includes:

[0184] A differential image data calculation module, configured to calculate differential image data between the original image data and the reference image data of each frame if there are multiple frames of the reference image data;

[0185] An area statistics module, used for counting the area of ​​foreground data in the differential image data;

[0186] The area selection module is used to select the differential image data with the smallest area of ​​the foreground data to obtain the first candidate image data.

[0187] In one embodiment of the present application, the quality inspection program loading module 305 includes:

[0188] A process information reading module, used to read the process information set for the processing equipment for processing the parts;

[0189] A keyword screening module, used for screening a plurality of keywords from the process information;

[0190] A process vector conversion module, used for converting the plurality of keywords into a process vector of the processing equipment;

[0191] A similarity calculation module, used for calculating the similarity between the process vector of the current processing equipment and the process vectors of other processing equipment;

[0192] A candidate device determination module, configured to determine the other processing devices as candidate devices if the similarity is greater than or equal to a preset first threshold;

[0193] The candidate device loading module is used to load the quality inspection program corresponding to the quality inspection tag from the candidate device if the candidate device has been configured with the quality inspection program corresponding to the quality inspection tag and the running time of the quality inspection program corresponding to the quality inspection tag on the candidate device exceeds a preset second threshold.

[0194] In one embodiment of the present application, the candidate device loading module includes:

[0195] A version information query module, for respectively querying the first version information of the quality inspection program corresponding to the quality inspection tag in the current processing equipment and the second version information of the quality inspection program corresponding to the quality inspection tag in the candidate equipment if the current processing equipment has been configured with the quality inspection program corresponding to the quality inspection tag;

[0196] A differential upgrade package downloading module, configured to download the differential upgrade package between the first version information and the second version information from the candidate device if the first version information is lower than the second version information and the quality inspection program corresponding to the quality inspection label in the current processing device is not prohibited from being upgraded to the second version information;

[0197] The quality inspection program upgrade module is used to use the differential upgrade package to upgrade the quality inspection program corresponding to the quality inspection label in the current processing equipment.

[0198] In one embodiment of the present application, the component quality inspection module 306 includes:

[0199] A quality inspection program startup module, used to start the quality inspection program; the quality inspection program includes a quality inspection network; the quality inspection network includes a first characteristic structure, a second characteristic structure, a first head structure and a second head structure;

[0200] A grayscale image data conversion module, used for converting the feature image data composed of the target image data and the target point cloud data into grayscale image data;

[0201] A common domain feature extraction module, used for inputting the grayscale image data into the first feature structure to extract common domain features;

[0202] A crack data segmentation module, used for inputting the public domain features into the first head structure to segment the crack data on the component;

[0203] A private domain feature conversion module, used for converting the public domain feature into the second feature structure and into a private domain feature;

[0204] A defective data detection module, used for inputting the private domain feature into the second header structure to detect defective data on the component;

[0205] An abnormality index statistical module, used for calculating a first abnormality index for the crack data and a second abnormality index for the defect data;

[0206] A quality inspection requirement information query module, used to query the first quality inspection requirement information and the second quality inspection requirement information set for the component;

[0207] A quality qualified determination module, configured to determine that the quality information of the component is qualified if the first abnormality indicator meets the first quality inspection requirement information and the second abnormality indicator meets the second quality inspection requirement information;

[0208] A quality failure determination module is used to set the crack data, the defect data and the quality failure as the quality information of the component if the first abnormal indicator does not meet the first quality inspection requirement information and / or the second abnormal indicator does not meet the second quality inspection requirement information.

[0209] In one embodiment of the present application, the first feature structure includes a first convolution block and a second convolution block, the second feature structure includes a residual network, a pooling layer and a multi-layer perceptron, the first head structure includes a first deconvolution block and a second deconvolution block, and the second head structure includes a third convolution block and a plurality of fully connected layers; the public domain features include a first basic feature and a second basic feature;

[0210] The public domain feature extraction module is also used for:

[0211] Inputting the grayscale image data into the first convolution block to perform a convolution operation to obtain a first basic feature;

[0212] Inputting the first basic feature into the second convolution block to perform a convolution operation to obtain a second basic feature;

[0213] The crack data segmentation module is also used for:

[0214] Inputting the second basic feature into the first deconvolution block to perform a deconvolution operation to obtain a first reconstructed feature;

[0215] splicing the first reconstructed feature and the first basic feature into a second reconstructed feature;

[0216] Inputting the second reconstructed feature into the second deconvolution block to perform a deconvolution operation to obtain crack data on the component;

[0217] The private domain feature conversion module is also used for:

[0218] splicing the first basic feature and the second basic feature into a third basic feature;

[0219] Inputting the third basic feature into the residual network to extract the first intermediate feature;

[0220] Inputting the first intermediate feature into the pooling layer to perform a pooling operation to obtain a second intermediate feature;

[0221] Inputting the second intermediate feature into the multi-layer perceptron and mapping it into a private domain feature;

[0222] The defective data detection module is also used for:

[0223] Inputting the private domain feature into the third convolution block to perform a convolution operation to obtain a target feature;

[0224] The target features are sequentially input into a plurality of the fully connected layers to map the defect data on the component.

[0225] In one embodiment of the present application, the part model building module 307 includes:

[0226] A model backbone structure construction module, used to construct a three-dimensional model backbone structure for the component using the target point cloud data;

[0227] A texture map extraction module, used to extract a texture map from the target image data;

[0228] A texture map rendering module, used for rendering the texture map into the model backbone structure to obtain a part model;

[0229] The quality inspection processing module 308 includes:

[0230] A processing permission module, used for allowing the component to be transported to the next processing equipment for processing if the quality inspection instruction indicates that the quality inspection is successful;

[0231] A processing prohibition module, used for removing the component from the production line and prohibiting the component from being transported to the next processing equipment for processing if the quality inspection instruction indicates that the quality inspection has failed;

[0232] The quality inspection processing module 308 includes:

[0233] a statistical indicator calculation module, configured to compare the quality information with the quality inspection instruction to calculate a plurality of statistical indicators if the quality inspection program is upgraded from the first version information to the second version information;

[0234] A validity determination module, configured to determine that the quality inspection procedure under the second version information is valid for the quality inspection operation of the processing equipment if the plurality of statistical indicators meet a preset performance condition;

[0235] An invalidity determination module is used to determine that the quality inspection program under the second version information is invalid for the quality inspection operation of the processing equipment if multiple statistical indicators do not meet the preset performance conditions, and to roll back the quality inspection program from the second version information to the first version information.

[0236] An embodiment of the present application provides a furniture production control device, and by using the device, each step in the aforementioned method embodiments can be implemented.

[0237] As for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the description of the method embodiment part.

[0238] Reference Figure 4 , shows a schematic diagram of a terminal device provided in an embodiment of the present application. Figure 4 As shown, the terminal device 400 in the embodiment of the present application includes: a processor 410, a memory 420, and a computer program 421 stored in the memory 420 and executable on the processor 410. When the processor 410 executes the computer program 421, the steps in each embodiment of the above-mentioned furniture production control method are implemented, such as Figure 1 Alternatively, when the processor 410 executes the computer program 421, the functions of each module / unit in the above-mentioned device embodiments are realized, for example Figure 3 The functions of the processing equipment determination module 301, the original data acquisition module 302, the target data screening module 303, the configuration file reading module 304, the quality inspection program loading module 305, the parts quality inspection module 306, the part model construction module 307 and the quality inspection processing module 308 are shown.

[0239] Exemplarily, the computer program 421 can be divided into one or more modules / units, and the one or more modules / units are stored in the memory 420 and executed by the processor 410 to complete the present application. The one or more modules / units can be a series of computer program instruction segments that can complete specific functions, and the instruction segments can be used to describe the execution process of the computer program 421 in the terminal device 400. For example, the computer program 421 can be divided into a processing equipment determination module, a raw data acquisition module, a target data screening module, a configuration file reading module, a quality inspection program loading module, a parts quality inspection module, a parts model construction module and a quality inspection processing module. The specific functions of each module are as follows:

[0240] A processing equipment determination module is used to determine multiple processing equipment on the furniture production line; each of the processing equipment is used to process the parts of the furniture, and each of the processing equipment is equipped with a trigger, a camera and a radar;

[0241] A raw data acquisition module, used for simultaneously calling the camera to collect raw image data from the component and calling the radar to collect raw point cloud data from the component when the processing equipment conveys the processed component to the crawler and the component activates the trigger;

[0242] A target data screening module, used for comparing the original image data with the original point cloud data, so as to screen the target image data representing the component from the original image data and screen the target point cloud data representing the component from the original point cloud data;

[0243] A configuration file reading module, used to read a configuration file of the processing equipment; the configuration file records one or more quality inspection labels;

[0244] A quality inspection program loading module, used to load the quality inspection program corresponding to the quality inspection label according to the similarity of the processing equipment in processing the parts;

[0245] A component quality inspection module, used to start the quality inspection program, perform quality inspection operations on the component according to the target image data and / or the target point cloud data, and obtain quality information;

[0246] A part model building module, used to build a three-dimensional part model for the component according to the target image data and the target point cloud data;

[0247] The quality inspection processing module is used to process the component according to the quality inspection instruction if a quality inspection instruction is received when the quality information is displayed on the part model, and to process the quality inspection procedure according to the quality information and the quality inspection instruction.

[0248] The terminal device 400 may be the furniture production control device in the above-mentioned embodiments, and the furniture production control device may be a computing device such as a desktop computer or a cloud server. The terminal device 400 may include, but is not limited to, a processor 410 and a memory 420. Those skilled in the art will understand that Figure 4 It is only an example of the terminal device 400 and does not constitute a limitation on the terminal device 400. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the terminal device 400 may also include input and output devices, network access devices, buses, etc.

[0249] The processor 410 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0250] The memory 420 may be an internal storage unit of the terminal device 400, such as a hard disk or memory of the terminal device 400. The memory 420 may also be an external storage device of the terminal device 400, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the terminal device 400. Further, the memory 420 may also include both an internal storage unit of the terminal device 400 and an external storage device. The memory 420 is used to store the computer program 421 and other programs and data required by the terminal device 400. The memory 420 may also be used to temporarily store data that has been output or is to be output.

[0251] An embodiment of the present application also discloses a terminal device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the furniture production control method as described in the aforementioned embodiments is implemented.

[0252] An embodiment of the present application further discloses a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the furniture production control method as described in the above-mentioned embodiments is implemented.

[0253] An embodiment of the present application further discloses a computer program product. When the computer program product is run on a computer, the computer is enabled to execute the furniture production control method described in the aforementioned embodiments.

[0254] The above-described embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application is described in detail with reference to the above-mentioned embodiments, a person skilled in the art should understand that the technical solutions described in the above-mentioned embodiments can still be modified, or some of the technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A furniture production control method, characterized in that: include: Determine a plurality of processing equipment on the furniture production line; each of the processing equipment is used to process parts of the furniture, and each of the processing equipment is equipped with a trigger, a camera and a radar; When the processing equipment conveys the processed component to the crawler and the component activates the trigger, the camera is called to collect original image data from the component and the radar is called to collect original point cloud data from the component; Comparing the original image data with the original point cloud data to select target image data representing the component from the original image data and to select target point cloud data representing the component from the original point cloud data; Reading a configuration file of the processing equipment; the configuration file records one or more quality inspection labels; Loading the quality inspection program corresponding to the quality inspection label according to the similarity of the processing equipment in processing the parts; Starting the quality inspection program, performing quality inspection operations on the component according to the target image data and / or the target point cloud data, and obtaining quality information; Constructing a three-dimensional part model for the component according to the target image data and the target point cloud data; If a quality inspection instruction is received when the quality information is displayed on the part model, the component is processed according to the quality inspection instruction, and the quality inspection procedure is processed according to the quality information and the quality inspection instruction.

2. The method according to claim 1, characterized in that The step of comparing the original image data with the original point cloud data to select target image data representing the component from the original image data and to select target point cloud data representing the component from the original point cloud data comprises: Reading a conversion matrix calibrated between the camera and the radar; Projecting the original point cloud data onto the original image data according to the transformation matrix to obtain reconstructed image data; querying the reference image data collected by the camera when the processing equipment has not processed the component; Calculating differential image data between the original image data and the reference image data as first candidate image data; Constructing a reference surface representing the crawler track according to the original point cloud data; Screening the data located above the reference plane in the recombined image data to obtain second candidate image data; Taking the intersection of the first candidate image data and the second candidate image data to obtain feature image data; It is determined that the pixel points in the feature image data constitute target image data representing the component, and the points in the feature image data constitute target point cloud data representing the component.

3. The method according to claim 2, characterized in that The calculating the difference image data between the original image data and the reference image data as the first candidate image data comprises: If there are multiple frames of reference image data, calculating the difference image data between the original image data and the reference image data of each frame; Counting the area of ​​foreground data in the differential image data; The differential image data with the smallest area of ​​the foreground data is selected to obtain the first candidate image data.

4. The method according to claim 1, characterized in that: The step of loading the quality inspection program corresponding to the quality inspection label according to the similarity of the processing equipment in processing the parts includes: Reading process information set for the processing equipment for processing the component; screening a plurality of keywords from the process information; converting a plurality of the keywords into process vectors of the processing equipment; Calculate the similarity between the process vector of the current processing equipment and the process vectors of other processing equipment; If the similarity is greater than or equal to a preset first threshold, determining the other processing equipment as candidate equipment; If the candidate device has been configured with a quality inspection program corresponding to the quality inspection tag, and the running time of the quality inspection program corresponding to the quality inspection tag on the candidate device exceeds a preset second threshold, the quality inspection program corresponding to the quality inspection tag is loaded from the candidate device.

5. The method according to claim 4, characterized in that The step of loading the quality inspection program corresponding to the quality inspection tag from the candidate device includes: If the current processing equipment has been configured with the quality inspection program corresponding to the quality inspection label, query the first version information of the quality inspection program corresponding to the quality inspection label in the current processing equipment and the second version information of the quality inspection program corresponding to the quality inspection label in the candidate equipment respectively; If the first version information is lower than the second version information and the quality inspection program corresponding to the quality inspection label in the current processing equipment is not prohibited from being upgraded to the second version information, then downloading the differential upgrade package between the first version information and the second version information from the candidate equipment; The differential upgrade package is used to upgrade the quality inspection program corresponding to the quality inspection label in the current processing equipment.

6. The method according to any one of claims 1 to 5, characterized in that The starting of the quality inspection program, performing a quality inspection operation on the component according to the target image data and / or the target point cloud data, and obtaining quality information, includes: Starting the quality inspection program; the quality inspection program includes a quality inspection network; the quality inspection network includes a first characteristic structure, a second characteristic structure, a first head structure and a second head structure; Converting the feature image data composed of the target image data and the target point cloud data into grayscale image data; Inputting the grayscale image data into the first feature structure to extract common domain features; Inputting the public domain features into the first head structure to segment the crack data on the component; Inputting the public domain feature into the second feature structure to convert it into a private domain feature; Inputting the private domain feature into the second head structure to detect defect data on the component; Counting a first abnormality index for the crack data and a second abnormality index for the defect data; Querying the first quality inspection requirement information and the second quality inspection requirement information set for the component; If the first abnormality indicator meets the first quality inspection requirement information, and the second abnormality indicator meets the second quality inspection requirement information, then determining that the quality information of the component is qualified; If the first abnormal indicator does not meet the first quality inspection requirement information, and / or the second abnormal indicator does not meet the second quality inspection requirement information, the crack data, the defect data and the unqualified quality are set as the quality information of the component.

7. The method according to claim 6, characterized in that The first feature structure includes a first convolution block and a second convolution block, the second feature structure includes a residual network, a pooling layer and a multi-layer perceptron, the first head structure includes a first deconvolution block and a second deconvolution block, and the second head structure includes a third convolution block and a plurality of fully connected layers; the public domain features include a first basic feature and a second basic feature; The step of inputting the grayscale image data into the first feature structure to extract the common domain features comprises: Inputting the grayscale image data into the first convolution block to perform a convolution operation to obtain a first basic feature; Inputting the first basic feature into the second convolution block to perform a convolution operation to obtain a second basic feature; The step of inputting the public domain feature into the first head structure to segment the crack data on the component comprises: Inputting the second basic feature into the first deconvolution block to perform a deconvolution operation to obtain a first reconstructed feature; splicing the first reconstructed feature and the first basic feature into a second reconstructed feature; Inputting the second reconstructed feature into the second deconvolution block to perform a deconvolution operation to obtain crack data on the component; The step of inputting the public domain feature into the second feature structure and converting it into a private domain feature includes: splicing the first basic feature and the second basic feature into a third basic feature; Inputting the third basic feature into the residual network to extract the first intermediate feature; Inputting the first intermediate feature into the pooling layer to perform a pooling operation to obtain a second intermediate feature; Inputting the second intermediate feature into the multi-layer perceptron and mapping it into a private domain feature; The step of inputting the private domain feature into the second head structure to detect defect data on the component comprises: Inputting the private domain feature into the third convolution block to perform a convolution operation to obtain a target feature; The target features are sequentially input into a plurality of the fully connected layers to map the defect data on the component.

8. The method according to any one of claims 1 to 5, characterized in that The constructing of a three-dimensional part model for the component according to the target image data and the target point cloud data comprises: Using the target point cloud data to construct a three-dimensional model backbone structure for the component; Extracting a texture map from the target image data; Rendering the texture map into the model backbone structure to obtain a part model; Processing the parts according to the quality inspection instruction includes: If the quality inspection instruction indicates that the quality inspection is successful, the component is allowed to be transported to the next processing equipment for processing; If the quality inspection instruction indicates that the quality inspection has failed, the component is removed from the production line, and the component is prohibited from being transported to the next processing equipment for processing; Processing the quality inspection procedure according to the quality information and the quality inspection instruction includes: If the quality inspection program is upgraded from the first version information to the second version information, the quality information is compared with the quality inspection instruction to calculate a plurality of statistical indicators; If a plurality of the statistical indicators meet the preset performance conditions, determining that the quality inspection procedure under the second version information is valid for the quality inspection operation of the processing equipment; If a plurality of the statistical indicators do not meet the preset performance conditions, it is determined that the quality inspection program under the second version information is invalid for the quality inspection operation of the processing equipment, and the quality inspection program is rolled back from the second version information to the first version information.

9. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the furniture production control method as described in any one of claims 1-8 is implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the furniture production control method as described in any one of claims 1-8 is implemented.

Images