Mould dimension detection method, device, equipment and medium

Abstract

The present application discloses a mold size detection method, device, equipment and medium, belonging to the industrial field, the method comprising: when receiving a detection instruction for mold size detection, determining the corresponding mold to be detected; determining a preset object detection model; inputting information collected by a preset camera into the object detection model, processing the information based on the object detection model, and obtaining the mold type corresponding to the mold to be detected; based on the mold type, searching in a mold information library that associates the mold type with the mold three-dimensional size, and obtaining the mold three-dimensional size corresponding to the mold type. In the present application, after the object detection model automatically identifies the mold type, the mold size detection device automatically searches directly for the mold three-dimensional size corresponding to the mold type in the preset mold information library, which simplifies the detection steps when manually detecting the three-dimensional size of the mold, that is, improves the detection efficiency when detecting the three-dimensional size of the mold.

Description

Technical Field

[0001] The present application relates to the industrial field, and in particular to a mold size detection method, device, equipment and medium. Background Art

[0002] At present, when detecting the three-dimensional size of a mold, a mold detection tool is used manually to measure each mold to be detected. When measuring manually, the measurement procedure is cumbersome, and the measurement result is easily affected by the actual operation of the detection personnel. In addition, different mold types correspond to differentiated mold structures, and a variety of mold types to be detected further increase the measurement workload of the measurement personnel. That is, there is a problem in the prior art that the detection efficiency is low when manually detecting the three-dimensional size of the mold.

[0003] The above contents are only used to assist in understanding the technical solution of the present application and do not constitute an admission that the above contents are prior art. Summary of the invention

[0004] The main purpose of the present invention is to provide a mold size detection method, device, equipment and medium, aiming to solve the problem of low detection efficiency when manually detecting the three-dimensional size of the mold in the prior art.

[0005] To achieve the above object, the present application provides a mold size detection method, the method comprising:

[0006] When receiving a detection instruction for mold size detection, determining a corresponding mold to be detected;

[0007] Determine a preset object detection model;

[0008] Inputting information collected by a preset camera into the object detection model, and processing the information based on the object detection model to obtain a mold type corresponding to the mold to be detected;

[0009] Based on the mold type, the mold information library that associates the mold type with the mold three-dimensional size is searched to obtain the mold three-dimensional size corresponding to the mold type.

[0010] Optionally, the preset camera is an RGB-D camera, and the RGB-D camera includes an infrared projector, an RGB color camera, and an infrared camera. When a detection instruction for mold size detection is received, before the step of determining the corresponding mold to be detected, the following steps are included:

[0011] determining a first infrared ray reflected after being projected onto the target mold by the infrared projector;

[0012] determining a second infrared ray reflected after being projected by the infrared projector onto the target reference plane;

[0013] The three-dimensional size of the target mold is determined based on a parallax distance between a first projection point of the first infrared ray and a second projection point of the second infrared ray.

[0014] Optionally, after the step of determining the three-dimensional size of the target mold based on the parallax distance between the first projection point of the first infrared ray and the second projection point of the second infrared ray, the method further comprises:

[0015] Determine the Yolov5 model to be trained;

[0016] Get sample images to be trained;

[0017] Preprocessing the sample images to be trained, wherein the area of ​​the black borders filled around the images when the mold size detection device scales the sample images to be trained with an aspect ratio other than 1:1 is reduced;

[0018] Based on the preprocessed sample images to be trained, the Yolov5 model to be trained is trained to obtain an object detection model having a feature map of the mold type, wherein the feature map of the mold type after feature fusion is obtained by transferring high-level mold type feature information during top-down upsampling and fusing multi-scale mold type features during bottom-up downsampling;

[0019] The characteristic graph of the mold type is associated with the three-dimensional size of the mold to obtain the mold information library.

[0020] Optionally, when the number of molds to be inspected is more than one, the number of the information collected by the preset camera is more than one, and the number of the three-dimensional dimensions of the mold corresponding to the mold type is more than one.

[0021] Optionally, the step of inputting the information collected by the preset camera into the object detection model, and processing the information based on the object detection model to obtain the mold type corresponding to the mold to be detected includes:

[0022] Determine that the photo of the mold to be inspected collected by the RGB color camera in the preset camera is the information;

[0023] The information is input into the object detection model, and the information is processed based on the object detection model to obtain the mold type corresponding to the mold to be detected.

[0024] Optionally, when a detection instruction for mold size detection is received, after the step of determining the corresponding mold to be detected, the method further comprises:

[0025] If it is necessary to expand the type of molds to be inspected, determine the preset updated object inspection model;

[0026] Inputting information collected by a preset camera into the updated object detection model, and processing the information based on the updated object detection model to obtain a mold type corresponding to the mold to be detected;

[0027] Based on the mold type, the mold information library that has been expanded to associate the mold type with the mold three-dimensional size is searched to obtain the mold three-dimensional size corresponding to the mold type.

[0028] Optionally, if the three-dimensional size of the mold corresponding to the mold type in the mold information library needs to be updated, the mold information library is updated based on the corresponding update requirements.

[0029] Optionally, after the step of searching in a mold information library that associates the mold type with the mold three-dimensional size based on the mold type to obtain the mold three-dimensional size corresponding to the mold type, the method further comprises:

[0030] The three-dimensional size of the mold corresponding to the mold type is displayed in the form of a three-dimensional diagram.

[0031] In addition, to achieve the above-mentioned purpose, the present application also provides a mold size detection device, the mold size detection device comprising:

[0032] A first determination module, configured to determine a corresponding mold to be detected when receiving a detection instruction for mold size detection;

[0033] A second determination module, used to determine a preset object detection model;

[0034] A processing module, used for inputting the information collected by the preset camera into the object detection model, and processing the information based on the object detection model to obtain the mold type corresponding to the mold to be detected;

[0035] The search module is used to search in a mold information library that associates the mold type with the mold three-dimensional size based on the mold type, so as to obtain the mold three-dimensional size corresponding to the mold type.

[0036] In addition, to achieve the above-mentioned purpose, the present application also provides a mold size detection device, which is a physical node device, and the mold size detection device includes: a memory, a processor, and a mold size detection program stored in the memory and executable on the processor, and the processor executes the mold size detection program to implement the steps of the mold size detection method.

[0037] In addition, to achieve the above-mentioned purpose, the present application also provides a medium, on which is stored a program for implementing the mold size detection method, and when the mold size detection program is executed by a processor, the steps of the mold size detection method described above are implemented.

[0038] The present application provides a mold size detection method, device, equipment and medium. Compared with the problem of low detection efficiency when manually detecting the three-dimensional size of the mold in the prior art, in the present application, when a detection instruction for mold size detection is received, the corresponding mold to be detected is determined; a preset object detection model is determined; information collected by a preset camera is input into the object detection model, and the information is processed based on the object detection model to obtain the mold type corresponding to the mold to be detected; based on the mold type, the mold information library that associates the mold type with the three-dimensional size of the mold is searched to obtain the mold three-dimensional size corresponding to the mold type. In the present application, after processing the information of the mold to be detected based on the preset object detection model, the mold type obtained automatically searches for the mold three-dimensional size corresponding to the mold type in the preset mold information library, automatically identifies the mold type, and automatically finds the corresponding mold three-dimensional size based on the mold type, which simplifies the detection steps when manually detecting the three-dimensional size of the mold, that is, improves the detection efficiency when detecting the three-dimensional size of the mold. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is a schematic diagram of the process involved in the embodiment of the present application;

[0040] Figure 2 is a schematic diagram of calculating the parallax distance involved in the embodiment of the present application;

[0041] Figure 3 This is a logical architecture diagram of a mold size detection device according to an embodiment of the present application;

[0042] Figure 4 This is a schematic diagram of the equipment structure of the hardware operating environment involved in the mold size detection method of this application.

[0043] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0044] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work belong to the scope of protection of the present invention.

[0045] Embodiment 1

[0046] The present application embodiment provides a mold size detection method. In the common embodiment of the mold size detection method of the present application, refer to Figure 1 , applied to a mold size detection device, the mold size detection method includes:

[0047] Step S10, when a detection instruction for mold size detection is received, determining the corresponding mold to be detected;

[0048] Step S20, determining a preset object detection model;

[0049] Step S30, inputting the information collected by the preset camera into the object detection model, and processing the information based on the object detection model to obtain the mold type corresponding to the mold to be detected;

[0050] Step S40: Based on the mold type, searching in a mold information library that associates the mold type with the three-dimensional size of the mold, to obtain the three-dimensional size of the mold corresponding to the mold type.

[0051] In this embodiment, there may be various types of molds for the mold to be inspected, such as a shrinkage mold, or a bearing, etc. In this embodiment, the mold to be inspected is a bearing to be inspected.

[0052] In this embodiment, the object detection model is a model for detecting the mold type of the mold to be detected. As an example, the model to be detected is a trained Yolov5 model, wherein the Yolov5 model is a neural network model, and the Yolov5 model can be used for object detection.

[0053] In this embodiment, the information collected by the preset camera is a picture of the mold to be inspected taken by the preset camera.

[0054] In this embodiment, the mold information library is an information library that associates mold types with mold three-dimensional dimensions. As an example, the mold information library stores three-dimensional information 1 of X1 type shrinkage molds, three-dimensional information 2 of X2 type shrinkage molds, three-dimensional information 3 of Y1 type bearings, and three-dimensional information 4 of Y2 type bearings.

[0055] In this embodiment, when a detection instruction for mold size detection is received, the corresponding mold to be detected is determined, a preset object detection model is determined, information collected by a preset camera is input into the object detection model, and the information is processed based on the object detection model to obtain the mold type corresponding to the mold to be detected. Based on the mold type, the mold information library that associates the mold type with the three-dimensional size of the mold is searched to obtain the three-dimensional size of the mold corresponding to the mold type.

[0056] The specific steps are as follows:

[0057] Step S10, when a detection instruction for mold size detection is received, determining the corresponding mold to be detected;

[0058] In this embodiment, when the mold size detection device receives a detection instruction for mold size detection, the mold size detection device determines a mold to be detected based on the detection instruction.

[0059] In this embodiment, the preset camera is an RGB-D camera, and the RGB-D camera includes an infrared projector, an RGB color camera, and an infrared camera. Step S10, when receiving a detection instruction for mold size detection, before determining the corresponding step of the mold to be detected, includes:

[0060] Step S01, determining the first infrared ray reflected after being projected onto the target mold by the infrared projector;

[0061] In this embodiment, if Figure 2 As shown, it includes: a camera, a baseline, a target mold, and a target reference plane. The camera is the RGB-D camera in this embodiment, and the RGB-D camera includes an infrared projector, an RGB color camera, and an infrared camera. The infrared projector is at P, and projects infrared rays to point O on the surface of the target mold. The infrared rays are reflected at point O on the surface of the target mold. The reflected infrared rays are first infrared rays, and the path of the first infrared rays is OK1.

[0062] Step S02, determining a second infrared ray reflected after being projected by the infrared projector onto the target reference plane;

[0063] In this embodiment, the infrared projector is at point P and projects infrared rays to point R of the target reference plane. The infrared rays are reflected at point R of the target reference plane. The reflected infrared rays are second infrared rays. The path of the second infrared rays is RR1. OK1 and RR1 intersect at point C.

[0064] Step S03 , determining the three-dimensional size of the target mold based on the parallax distance between the first projection point of the first infrared ray and the second projection point of the second infrared ray.

[0065] In this embodiment, the first projection point of the first infrared ray is K1, and the second projection point of the second infrared ray is R1. The distance d between K1 and R1 is the parallax distance between the first projection point of the first infrared ray and the second projection point of the second infrared ray.

[0066] In this embodiment, f is the focal length of the RGB-D camera, b is the baseline length, Z0 is the distance between the target reference plane and the RGB-D camera, and Z is the distance between point O on the target mold surface and the RGB-D camera.

[0067] According to the similarity between ΔRKC and ΔR1K1C1, we can obtain According to the similarity between ΔOKR and ΔOCP, we can get Then the distance between point O on the target mold surface and the RGB-D camera (current depth) Based on the calculation principle of the current depth, the three-dimensional size of the target mold is determined.

[0068] Step S03, after the step of determining the three-dimensional size of the target mold based on the parallax distance between the first projection point of the first infrared ray and the second projection point of the second infrared ray, comprises:

[0069] Step S04, determining the Yolov5 model to be trained;

[0070] Step S05, obtaining sample images to be trained;

[0071] In this embodiment, an RGB color camera is used to photograph the target mold, and the photographed target mold image is the sample image to be trained.

[0072] Step S06, scaling the sample image to be trained, wherein when scaling the sample image to be trained with an aspect ratio other than 1:1, reducing the area of ​​the black border filled around the sample image;

[0073] In this embodiment, for the sample images to be trained whose aspect ratio is not 1:1, when the mold size detection device is scaled, black borders will be filled around the sample images to be trained, and the area of ​​the black borders is relatively large. The filled black borders affect the training effect of the model, so it is necessary to reduce the area of ​​the filled black borders. In this embodiment, the sample images to be trained are scaled, wherein when scaling the sample images to be trained whose aspect ratio is not 1:1, the area of ​​the black borders filled around them is reduced;.

[0074] Step S07, training the Yolov5 model to be trained based on the scaled sample image to be trained to obtain an object detection model for detecting mold types, wherein during the training process, upsampling is performed from top to bottom to transfer high-level mold type feature information, and downsampling is performed from bottom to top to fuse multi-scale mold type features to obtain a feature map of the mold type after feature fusion;

[0075] In this embodiment, based on the scaled sample image to be trained, the Yolov5 model to be trained is trained, wherein, during the training process, upsampling is performed from top to bottom to transfer high-level mold type feature information, and downsampling is performed from bottom to top to fuse multi-scale mold type features to obtain a feature map of the mold type after feature fusion. If the verification error rate is within the preset allowable error range, the training ends, and an object detection model for detecting mold types is obtained.

[0076] Step S08, associating the characteristic graph of the mold type with the three-dimensional size of the mold to obtain the mold information library.

[0077] In this embodiment, the object detection model obtained by training the Yolov5 model to be trained has a feature map of the mold type. Since each mold type corresponds to a three-dimensional size of the mold, the feature map of the mold type corresponds to the three-dimensional size of the mold. The feature map of the mold type is associated with the three-dimensional size of the mold to obtain the mold information library.

[0078] As an example, the mold information library is obtained by associating the feature map of mold type X1 with the mold three-dimensional size x1, the feature map of mold type X2 with the mold three-dimensional size x2, and the feature map of mold type X3 with the mold three-dimensional size x3.

[0079] Step S20, determining a preset object detection model;

[0080] In this embodiment, a preset object detection model is determined. The preset object detection model is a model obtained by performing model training on the sample images to be trained.

[0081] As an example, when the sample image to be trained is an X1 type shrinkage mold, the object detection model X1 corresponding to the X1 type shrinkage mold is used; when the sample image to be trained is a Y1 type bearing, the object detection model Y1 corresponding to the Y1 type bearing is used.

[0082] Step S30, inputting the information collected by the preset camera into the object detection model, and processing the information based on the object detection model to obtain the mold type corresponding to the mold to be detected;

[0083] As an example, the information collected by the preset camera is a picture of the mold to be inspected taken by an RGB color camera in a preset RGB-D camera.

[0084] In this embodiment, the information collected by the preset camera is input into the object detection model, and the information is processed based on the object detection model to obtain the mold type corresponding to the mold to be detected.

[0085] As an example, a bearing image taken by an RGB color camera in a preset RGB-D camera is input into the object detection model, and the information is processed based on the object detection model to obtain that the mold type corresponding to the mold to be detected is a Y1 type bearing.

[0086] Step S40: Based on the mold type, searching in a mold information library that associates the mold type with the three-dimensional size of the mold, to obtain the three-dimensional size of the mold corresponding to the mold type.

[0087] In this embodiment, the mold type is searched in the mold information library to obtain the mold three-dimensional size corresponding to the mold type.

[0088] The present application provides a mold size detection method, device, equipment and medium. Compared with the problem of low detection efficiency when manually detecting the three-dimensional size of the mold in the prior art, in the present application, when a detection instruction for mold size detection is received, the corresponding mold to be detected is determined; a preset object detection model is determined; information collected by a preset camera is input into the object detection model, and the information is processed based on the object detection model to obtain the mold type corresponding to the mold to be detected; based on the mold type, the mold information library that associates the mold type with the three-dimensional size of the mold is searched to obtain the mold three-dimensional size corresponding to the mold type. In the present application, after processing the information of the mold to be detected based on the preset object detection model, the mold type obtained automatically searches for the mold three-dimensional size corresponding to the mold type in the preset mold information library, automatically identifies the mold type, and automatically finds the corresponding mold three-dimensional size based on the mold type, which simplifies the detection steps when manually detecting the three-dimensional size of the mold, that is, improves the detection efficiency when detecting the three-dimensional size of the mold.

[0089] Embodiment 2

[0090] Further, based on the first embodiment of the present application, another embodiment of the present application is provided. In this embodiment, in step S10, when a detection instruction for mold size detection is received, after determining the corresponding step of the mold to be detected

[0091] Step A1: if it is necessary to expand the type of mold to be inspected, determine the preset updated object detection model;

[0092] In this embodiment, if the mold type of the mold to be inspected cannot be found in the mold information library that associates the mold type with the three-dimensional size of the mold, that is, if the mold type of the current mold to be inspected is not added in the mold information library, it is necessary to expand the mold type of the current mold to be inspected before receiving the detection instruction for mold size detection.

[0093] Step A2: inputting the information collected by the preset camera into the updated object detection model, and processing the information based on the updated object detection model to obtain the mold type corresponding to the mold to be detected;

[0094] Step A3: Based on the mold type, search in the expanded mold information library that associates the mold type with the mold three-dimensional size to obtain the mold three-dimensional size corresponding to the mold type.

[0095] In this embodiment, if it is necessary to expand the type of mold to be detected, first, update the preset object detection model. Based on the sample images of the mold type to be detected to be expanded, train the Yolov5 model to be trained to obtain an updated object detection model. Secondly, input the information collected by the preset camera into the updated object detection model, and then, based on the updated object detection model, process the information to obtain the mold type corresponding to the mold to be detected. Finally, based on the mold type, search in the mold information library that has been expanded to associate the mold type with the mold three-dimensional size to obtain the mold three-dimensional size corresponding to the mold type.

[0096] In this embodiment, if the mold type of the mold to be inspected cannot be found in the mold information library that associates the mold type with the three-dimensional size of the mold, the corresponding mold information library and the object detection model are updated according to the expansion requirements of the mold type, thereby avoiding the problem that the mold size detection device cannot detect the mold to be inspected corresponding to the current detection instruction, that is, further improving the detection efficiency when detecting the three-dimensional size of the mold.

[0097] Embodiment 3

[0098] Furthermore, based on Example 1 and Example 2 of the present application, another embodiment of the present application is provided. In this embodiment, if it is necessary to update the three-dimensional size of the mold corresponding to the mold type in the mold information library, the mold information library is updated based on the corresponding update requirements.

[0099] In this embodiment, if the three-dimensional dimensions of the mold recorded in the mold information library in the mold size detection device need to be modified, as an example, after the three-dimensional dimensions corresponding to the Y1 type bearing are stored in the mold information library, the manufacturer improves the structure of the Y1 type bearing, then the three-dimensional dimensions of the Y1 type bearing after the structural improvement will change accordingly. If the three-dimensional dimensions of the mold recorded in the mold information library continue to be used as the actual dimensions, it is unreasonable. Therefore, the three-dimensional dimensions of the mold corresponding to the mold type in the mold information library need to be updated.

[0100] In this embodiment, if the three-dimensional size of the mold corresponding to the mold type in the mold information library needs to be updated, the mold information library is updated based on the corresponding update requirements.

[0101] As an example, after the three-dimensional dimensions corresponding to the Y1-type bearings are stored in the mold information library, the manufacturer improves the structure of the Y1-type bearings. The mold dimension detection device determines the first infrared ray reflected after being projected by the infrared projector onto the improved Y1-type bearings, determines the second infrared ray reflected after being projected by the infrared projector onto the target reference plane, determines the mold three-dimensional dimensions of the improved Y1-type bearings based on the parallax distance between the first projection point of the first infrared ray and the second projection point of the second infrared ray, uses the mold three-dimensional dimensions as the new mold three-dimensional dimensions, and updates the mold three-dimensional dimensions of the Y1-type bearings in the mold information library.

[0102] In this embodiment, the step of updating the mold information library is added if the three-dimensional size of the mold corresponding to the mold type in the mold information library needs to be updated. This enhances the real-time nature of the mold information recorded in the mold information library. That is, the dynamic mold information library further improves the detection efficiency when detecting the three-dimensional size of the mold.

[0103] Embodiment 4

[0104] Furthermore, based on all the above embodiments in the present application, another embodiment of the present application is provided. In this embodiment, when the number of molds to be inspected is more than one, the number of information collected by the preset camera is more than one, and the number of three-dimensional dimensions of the mold corresponding to the mold type is more than one.

[0105] In this embodiment, if there are more than one molds to be detected, the preset camera needs to collect information for each mold to be detected, and the number of collected information is more than one. Based on multiple molds to be detected, the object detection model outputs multiple mold types, and multiple corresponding mold three-dimensional sizes are obtained based on multiple mold types.

[0106] In this embodiment, when the number of molds to be inspected is more than one, the mold size detection device outputs multiple corresponding mold three-dimensional dimensions based on the multiple molds to be inspected, and simultaneously processes the mold size detection tasks of the multiple molds to be inspected, thereby speeding up the mold size detection speed, that is, further improving the detection efficiency when detecting the three-dimensional dimensions of the mold.

[0107] Embodiment 5

[0108] Further, based on all the above embodiments, another embodiment of the present application is provided. In this embodiment, after the step S40 of searching in a mold information library that associates the mold type with the mold three-dimensional size based on the mold type to obtain the mold three-dimensional size corresponding to the mold type, the following steps are included:

[0109] The three-dimensional size of the mold corresponding to the mold type is displayed in the form of a three-dimensional diagram.

[0110] In this embodiment, after obtaining the three-dimensional dimensions of the mold corresponding to the mold type, modeling is performed based on the three-dimensional dimensions of the mold. After obtaining the corresponding three-dimensional simulation diagram, the mold size detection device displays the three-dimensional simulation diagram so that the user can intuitively view the structure of the mold.

[0111] In this embodiment, the three-dimensional dimensions of the mold are displayed in the form of a three-dimensional simulation diagram, which increases the intuitiveness of the mold size after detection and reduces the abstractness of the three-dimensional dimensions of the mold, that is, further improves the detection efficiency when detecting the three-dimensional dimensions of the mold.

[0112] Embodiment 6

[0113] Further, based on all the above embodiments, another embodiment of the present application is provided. In this embodiment, Figure 3 , a mold size detection device is provided, the device comprising:

[0114] A first determination module, configured to determine a corresponding mold to be detected when receiving a detection instruction for mold size detection;

[0115] A second determination module, used to determine a preset object detection model;

[0116] A processing module, used for inputting the information collected by the preset camera into the object detection model, and processing the information based on the object detection model to obtain the mold type corresponding to the mold to be detected;

[0117] The search module is used to search in a mold information library that associates the mold type with the mold three-dimensional size based on the mold type, so as to obtain the mold three-dimensional size corresponding to the mold type.

[0118] Optionally, in a possible implementation manner of the present application, the preset camera is an RGB-D camera, and the RGB-D camera includes an infrared projector, an RGB color camera, and an infrared camera. When a detection instruction for mold size detection is received, before the step of determining the corresponding mold to be detected, the device includes:

[0119] A third determination module is used to determine the first infrared ray reflected after being projected by the infrared projector onto the target mold;

[0120] A fourth determination module, used to determine the second infrared ray reflected after being projected by the infrared projector onto the target reference plane;

[0121] The fifth determining module is used to determine the three-dimensional size of the target mold based on the parallax distance between the first projection point of the first infrared ray and the second projection point of the second infrared ray.

[0122] Optionally, in a possible implementation manner of the present application, after the step of determining the three-dimensional size of the target mold based on the parallax distance between the first projection point of the first infrared ray and the second projection point of the second infrared ray, the device includes:

[0123] Determine the Yolov5 model to be trained;

[0124] Get sample images to be trained;

[0125] A scaling module, used for scaling the sample image to be trained, wherein when scaling the sample image to be trained with an aspect ratio other than 1:1, the black border area of ​​the black border filled around the sample image is reduced;

[0126] The Yolov5 model to be trained is trained based on the scaled sample image to be trained to obtain an object detection model for detecting the mold type, wherein, during the training process, upsampling is performed from top to bottom to transfer high-level mold type feature information, and downsampling is performed from bottom to top to fuse multi-scale mold type features to obtain a feature map of the mold type after feature fusion;

[0127] The association module is used to associate the characteristic map of the mold type with the three-dimensional size of the mold to obtain the mold information library.

[0128] Optionally, in a possible implementation manner of the present application, when the number of molds to be inspected is more than one, the number of the information collected by the preset camera is more than one, and the number of three-dimensional mold dimensions corresponding to the mold type is more than one.

[0129] Optionally, in a possible implementation manner of the present application, the step of inputting the information collected by the preset camera into the object detection model, and processing the information based on the object detection model to obtain the mold type corresponding to the mold to be detected, the device includes:

[0130] A seventh determination module is used to determine that the photo of the mold to be detected collected by the RGB color camera in the preset camera is the information;

[0131] The first processing module is used to input the information into the object detection model, and process the information based on the object detection model to obtain the mold type corresponding to the mold to be detected.

[0132] Optionally, in a possible implementation manner of the present application, when receiving a detection instruction for mold size detection, after the step of determining the corresponding mold to be detected, the device includes:

[0133] An eighth determination module, used to determine a preset updated object detection model if it is necessary to expand the type of mold to be detected;

[0134] A second processing module, configured to input information collected by a preset camera into the updated object detection model, and process the information based on the updated object detection model to obtain a mold type corresponding to the mold to be detected;

[0135] The first search module is used to search, based on the mold type, in an expanded mold information library that associates the mold type with the mold three-dimensional size, to obtain the mold three-dimensional size corresponding to the mold type.

[0136] Optionally, in a possible implementation manner of the present application, if it is necessary to update the three-dimensional size of the mold corresponding to the mold type in the mold information library, the mold information library is updated based on the corresponding update requirements.

[0137] Optionally, in a possible implementation manner of the present application, after the step of searching in a mold information library that associates the mold type with the mold three-dimensional size based on the mold type to obtain the mold three-dimensional size corresponding to the mold type, the device includes:

[0138] The display module is used to display the three-dimensional size of the mold corresponding to the mold type in the form of a three-dimensional diagram.

[0139] The specific implementation of the mold size detection device of the present application is basically the same as the embodiments of the above-mentioned mold size detection method, and will not be repeated here.

[0140] Embodiment 7

[0141] Further, based on all the above embodiments, another embodiment of the present application is provided. In this embodiment, a mold size detection device is provided. The mold size detection device is a physical node device. The mold size detection device includes: a memory, a processor, and a program for implementing the mold size detection method stored in the memory. The memory is used to store the program for implementing the mold size detection method; the processor is used to execute the program for implementing the mold size detection method to implement the steps of the mold size detection method in the above embodiments.

[0142] Reference Figure 4 , Figure 4 It is a schematic diagram of the device structure of the hardware operating environment involved in the embodiment of the present application.

[0143] like Figure 4 As shown, the mold size detection device may include: a processor 1001, such as a CPU, a memory 1005, and a communication bus 1002. The communication bus 1002 is used to realize the connection and communication between the processor 1001 and the memory 1005. The memory 1005 may be a high-speed RAM memory or a stable memory (non-volatile memory), such as a disk memory. The memory 1005 may also be a storage device independent of the aforementioned processor 1001.

[0144] Optionally, the mold size detection equipment may also include a network interface, an audio circuit, a display, connecting lines, sensors, an input module, etc. The network interface may optionally include a standard wired interface, a wireless interface (such as a WI-FI interface, a Bluetooth interface), and the input module may optionally include a keyboard, a system soft keyboard, voice input, wireless receiving input, etc.

[0145] Those skilled in the art will appreciate that the structure of the mold size detection device does not constitute a limitation on the mold size detection device, and may include more or fewer components than shown in the figure, or a combination of certain components, or a different arrangement of components.

[0146] The memory as a computer medium may include an operating system, an information exchange module and a mold size detection program. The operating system is a program that manages and controls the hardware and software resources of the mold size detection equipment, and supports the operation of the mold size detection program and other software and / or programs. The information exchange module is used to realize the communication between the components inside the memory, and to communicate with other hardware and software in the mold size detection system.

[0147] In the mold size detection device, the processor is used to execute the mold size detection program stored in the memory to implement the above-mentioned mold size detection steps.

[0148] The specific implementation of the mold size detection equipment of the present application is basically the same as the above-mentioned embodiments of the mold size detection method, and will not be repeated here.

[0149] Embodiment 8

[0150] An embodiment of the present application provides a medium, and the medium stores one or more programs, and the one or more programs can also be executed by one or more processors to implement the steps of the mold size detection method in the above embodiment.

[0151] The specific implementation of the medium of the present application is basically the same as the above-mentioned embodiments of the mold size detection method, and will not be repeated here.

[0152] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or system including the element.

[0153] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0154] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM or RAM, magnetic disk, optical disk) as described above, including a number of instructions for a terminal device (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in each embodiment of the present application.

[0155] The above are only preferred embodiments of the present application, and are not intended to limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A mold size detection method, characterized in that: The mold size detection method comprises: When receiving a detection instruction for mold size detection, determining a corresponding mold to be detected; Determine a preset object detection model; Inputting information collected by a preset camera into the object detection model, and processing the information based on the object detection model to obtain a mold type corresponding to the mold to be detected; Based on the mold type, searching in a mold information library that associates the mold type with the three-dimensional size of the mold, to obtain the three-dimensional size of the mold corresponding to the mold type; The preset camera is an RGB-D camera, and the RGB-D camera includes an infrared projector, an RGB color camera, and an infrared camera. When a detection instruction for mold size detection is received, before the step of determining the corresponding mold to be detected, the method includes: determining a first infrared ray reflected after being projected onto a target mold by the infrared projector; Determine a second infrared ray reflected after being projected by the infrared projector onto the target reference plane; Determining the three-dimensional size of the target mold based on a parallax distance between a first projection point of the first infrared ray and a second projection point of the second infrared ray; When a detection instruction for mold size detection is received, after the step of determining the corresponding mold to be detected, the method includes: If it is necessary to expand the type of molds to be inspected, determine the preset updated object inspection model; Inputting information collected by a preset camera into the updated object detection model, and processing the information based on the updated object detection model to obtain a mold type corresponding to the mold to be detected; Based on the mold type, the mold information library that has been expanded to associate the mold type with the mold three-dimensional size is searched to obtain the mold three-dimensional size corresponding to the mold type.

2. The mold size detection method according to claim 1, characterized in that: After the step of determining the three-dimensional size of the target mold based on the parallax distance between the first projection point of the first infrared ray and the second projection point of the second infrared ray, the method further comprises: Determine the Yolov5 model to be trained; Get sample images to be trained; Scaling the sample image to be trained, wherein when scaling the sample image to be trained with an aspect ratio other than 1:1, reducing the area of ​​the black border filled around the sample image; The Yolov5 model to be trained is trained based on the scaled sample image to be trained to obtain an object detection model for detecting the mold type, wherein, during the training process, upsampling is performed from top to bottom to transfer high-level mold type feature information, and downsampling is performed from bottom to top to fuse multi-scale mold type features to obtain a feature map of the mold type after feature fusion; The characteristic graph of the mold type is associated with the three-dimensional size of the mold to obtain the mold information library.

3. The mold size detection method according to claim 1, characterized in that: When the number of the molds to be inspected is more than one, the number of the information collected by the preset camera is more than one, and the number of the three-dimensional dimensions of the mold corresponding to the mold type is more than one.

4. The mold size detection method according to claim 1, characterized in that: If the three-dimensional size of the mold corresponding to the mold type in the mold information library needs to be updated, the mold information library is updated based on the corresponding update requirements.

5. The mold size detection method according to claim 1, characterized in that: After the step of searching in a mold information library that associates the mold type with the mold three-dimensional size based on the mold type to obtain the mold three-dimensional size corresponding to the mold type, the method further includes: The three-dimensional size of the mold corresponding to the mold type is displayed in the form of a three-dimensional diagram.

6. A mold size detection device, characterized in that: Mould size detection device, comprising: A first determination module, configured to determine a corresponding mold to be detected when receiving a detection instruction for mold size detection; A second determination module, used to determine a preset object detection model; A processing module, used for inputting the information collected by the preset camera into the object detection model, and processing the information based on the object detection model to obtain the mold type corresponding to the mold to be detected; A search module, configured to search, based on the mold type, in a mold information library that associates the mold type with the mold three-dimensional size, to obtain the mold three-dimensional size corresponding to the mold type; The preset camera is an RGB-D camera, which includes an infrared projector, an RGB color camera and an infrared camera. The mold size detection device is used to achieve: determining a first infrared ray reflected after being projected onto a target mold by the infrared projector; Determine a second infrared ray reflected after being projected by the infrared projector onto the target reference plane; Determining the three-dimensional size of the target mold based on a parallax distance between a first projection point of the first infrared ray and a second projection point of the second infrared ray; The mold size detection device is used to achieve: If it is necessary to expand the type of molds to be inspected, determine the preset updated object inspection model; Inputting information collected by a preset camera into the updated object detection model, and processing the information based on the updated object detection model to obtain a mold type corresponding to the mold to be detected; Based on the mold type, the mold information library that has been expanded to associate the mold type with the mold three-dimensional size is searched to obtain the mold three-dimensional size corresponding to the mold type.

7. A mold size detection device, characterized in that: It comprises a memory, a processor and a mold size detection program stored in the memory and executable on the processor, wherein the processor executes the mold size detection program to implement the steps of the mold size detection method according to any one of claims 1 to 5.

8. A medium, characterized in that The medium stores a program for implementing the mold size detection method, and the program for implementing the mold size detection method is executed by a processor to implement the steps of the mold size detection method according to any one of claims 1 to 5.

Images