Molten aluminum leakage identification and early warning method and device based on liquid level area abnormal change detection

Abstract

The invention discloses a liquid level area abnormal change detection-based molten aluminum leakage identification and early warning method and device, and the method comprises the steps: collecting a liquid level image of molten aluminum, carrying out the cutting and framing processing of the liquid level image, and carrying out the segmentation of the liquid level of the molten aluminum through an improved image semantic segmentation algorithm model; according to the molten aluminum liquid level segmentation image completed at the current moment, the molten aluminum liquid level area is calculated, and the molten aluminum liquid level area variation in a set time window is calculated; and an IQR method is adopted to determine an early warning threshold value of the change amount of the liquid level area, the change amount of the liquid level area of the molten aluminum is compared with the early warning threshold value, whether molten aluminum leakage exists or not is judged, and if the change amount of the liquid level area of the molten aluminum is larger than the early warning threshold value, it is judged that molten aluminum leakage exists. According to the invention, the accuracy and timeliness of judging molten aluminum leakage are effectively improved, and the intrinsic safety of aluminum processing deep well casting is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of safe production, and in particular relates to a method and a device for identifying and warning aluminum liquid leakage based on abnormal liquid surface area change detection. Background Art

[0002] In recent years, aluminum processing deep well casting explosion accidents have occurred frequently. The main reason is that the high-temperature aluminum liquid leakage in the casting well was not discovered in time. The high-temperature aluminum liquid came into contact with the cooling water and produced physical and chemical reactions, releasing a huge amount of energy in a short period of time in the narrow casting deep well space, thus causing explosion accidents, causing great casualties and economic losses to the company. Therefore, timely detection of aluminum liquid leakage and real-time warning can effectively prevent explosion accidents caused by aluminum liquid leakage.

[0003] By comparing and analyzing the characteristics of the change in the surface area of ​​molten aluminum in the normal state and the state of aluminum liquid leakage during the casting process, it is found that the surface area of ​​molten aluminum decreases in the state of aluminum liquid leakage. According to the scenarios and characteristics of abnormal changes in the surface area of ​​molten aluminum, a method for identifying and warning aluminum liquid leakage based on abnormal surface area change detection is proposed, and a set of aluminum liquid leakage identification and warning devices are established based on this method. For example, the high-temperature molten aluminum liquid and water contact explosion prevention and warning method based on numerical simulation technology with application number 201910775833.8, although this scheme can judge whether overflow leakage occurs in the aluminum liquid pouring process under the same aluminum liquid flow rate under the simulated state by simulating the aluminum liquid pouring state under different aluminum liquid flow rates and measuring the real-time aluminum liquid flow rate, the numerical simulation technology it uses has a relatively large time lag, and its modeling geometric parameters require a lot of manual adjustment.

[0004] In summary, how to improve the real-time recognition and judgment of aluminum liquid level changes and reduce manual intervention are issues that need to be addressed urgently. Summary of the invention

[0005] The main purpose of the present invention is to overcome the shortcomings and deficiencies of relying on operators to observe whether there is aluminum liquid leakage, and to provide an aluminum liquid leakage identification and early warning method and device based on abnormal liquid surface area change detection, and propose an improved image semantic segmentation algorithm model to perform aluminum liquid surface segmentation, and then calculate the liquid surface area change index, and use the IQR method to calculate the early warning threshold of the liquid surface area change, which effectively improves the accuracy and timeliness of judging aluminum liquid leakage and enhances the inherent safety of deep well casting of aluminum processing.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] In a first aspect, the present invention provides a method for identifying and warning aluminum liquid leakage based on abnormal change detection of liquid surface area, comprising the following steps:

[0008] Collect the aluminum liquid surface image, crop and frame the liquid surface image, and use the improved image semantic segmentation algorithm model to segment the aluminum liquid surface;

[0009] According to the aluminum liquid surface segmentation image completed at the current moment, the aluminum liquid surface area is calculated, and the aluminum liquid surface area change within the set time window is calculated;

[0010] The IQR method is used to determine the warning threshold of the liquid surface area change. The change in the aluminum liquid surface area is compared with the warning threshold to determine whether there is an aluminum liquid leakage. When the change in the aluminum liquid surface area is greater than the warning threshold, it is determined to be an aluminum liquid leakage.

[0011] As an optimal technical solution, the improved image semantic segmentation algorithm model includes an encoder, a decoder and an attention mechanism, the encoder structure includes a feature network extraction module and a void space convolution pooling pyramid module, the attention mechanism includes a channel attention module and a spatial attention module, and the channel attention module and the spatial attention module are connected in series; the attention mechanism is introduced into the encoder structure, the deep feature map extracted by the feature network extraction module is input into the attention mechanism, and the output end of the attention mechanism is connected to the void space convolution pooling pyramid module; the attention mechanism is introduced before the decoder structure, the shallow feature map extracted by the feature network extraction module is input into the attention mechanism, and the output end of the attention mechanism is connected to the decoder; the output end of the void space convolution pooling pyramid module is connected to the decoder.

[0012] As a preferred technical solution, an improved offline training of the image semantic segmentation algorithm model is included, specifically:

[0013] Collecting RGB images of the aluminum liquid surface area and performing image processing to establish an aluminum liquid surface image data set; the image processing includes image cropping, image denoising and image annotation;

[0014] An improved image semantic segmentation algorithm model is constructed, and a transfer learning strategy of feature extraction is utilized; the transfer learning strategy includes a model pre-training stage and a transfer stage. In the model pre-training stage, the improved image semantic segmentation algorithm model is pre-trained using a VOC data set, a pre-training file in a .pth format is generated, and pre-training weights are obtained; in the transfer stage, the improved image semantic segmentation algorithm model is re-trained using an aluminum liquid surface image data set and according to the pre-training weights, and the training is stopped when the model converges.

[0015] As a preferred technical solution, the calculation of the aluminum liquid surface area is specifically as follows:

[0016] Read the aluminum liquid surface segmentation image completed at the current moment and adjust the image to its original size; extract the pixel points of the aluminum liquid surface part in the image, calculate the total liquid surface pixels based on OpenCV, use the total value of the liquid surface pixels to represent the liquid surface area, and calculate the image liquid surface area at the current moment.

[0017] As a preferred technical solution, the calculation of the change in the surface area of ​​the molten aluminum within the set time window includes:

[0018] The set time ΔT is used as the time window for the change in the surface area of ​​the aluminum liquid;

[0019] Calculate the absolute value of the change in liquid surface area as follows:

[0020] |ΔS|=|S t+ΔT -S t |

[0021] Among them, |ΔS| represents the absolute value of the change in liquid surface area, S t represents the liquid surface area at the tth second, S t+ΔT Represents the liquid surface area at t+ΔT seconds.

[0022] As a preferred technical solution, the IQR method is used to determine the warning threshold of the liquid surface area change, as shown in the following formula:

[0023] IQR=|Q1-Q3|

[0024] LB=λ×IQR+Q3

[0025] In the formula, Q1 represents the upper and lower quartiles, IQR represents the quartile range, LB represents the warning threshold of the liquid surface area change, and λ is a fixed constant.

[0026] In a second aspect, the present invention further provides an aluminum liquid leakage identification and early warning system based on abnormal liquid surface area change detection, which is applied to the aluminum liquid leakage identification and early warning method based on abnormal liquid surface area change detection, and includes a camera, a data acquisition box, a server and a workstation;

[0027] A camera, used to capture video images of the aluminum liquid surface area;

[0028] Data collection box, used for collecting and transmitting video image data, the data collection box is connected to the camera through network cable and optical fiber;

[0029] Server, used for data storage, data processing, aluminum liquid leakage identification and early warning;

[0030] The workstation is used to read the data stored in the server and display the aluminum liquid surface area, aluminum liquid surface area change and aluminum liquid leakage warning information within the set time window in the form of a visual interface.

[0031] As a preferred technical solution, the server includes a data storage module, a data processing module, and an aluminum liquid leakage identification and aluminum leakage warning module;

[0032] A data storage module is used to store the real-time collected aluminum liquid surface area image data, image processing process data, aluminum liquid surface area, aluminum liquid surface area change within a set time window, and the time window and liquid surface area change warning threshold;

[0033] The data processing module is used for cutting, framing and segmenting the aluminum liquid surface image, calculating the aluminum liquid surface area, and calculating the change in the aluminum liquid surface area within a set time window;

[0034] The aluminum liquid leakage identification and aluminum leakage warning module is used to determine the warning threshold of the liquid surface area change by using the IQR method, compare the aluminum liquid surface area change with the warning threshold, and judge whether there is aluminum liquid leakage. When the aluminum liquid surface area change is greater than the warning threshold, it is judged as aluminum liquid leakage.

[0035] In a third aspect, the present invention provides an electronic device, the electronic device comprising:

[0036] at least one processor; and,

[0037] a memory communicatively connected to the at least one processor; wherein,

[0038] The memory stores computer program instructions that can be executed by the at least one processor, and the computer program instructions are executed by the at least one processor so that the at least one processor can execute the aluminum liquid leakage identification and early warning method based on abnormal liquid surface area change detection.

[0039] In a fourth aspect, the present invention provides a computer-readable storage medium storing a program, which, when executed by a processor, implements the aluminum liquid leakage identification and early warning method based on abnormal liquid surface area change detection.

[0040] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0041] The present invention uses the ResNet50 network as the liquid surface area change feature extraction network, and introduces the CBAM attention mechanism to improve the DeepLapV3+ algorithm model. The improved DeepLapV3+ algorithm model is used to segment the aluminum liquid surface, which can reduce the complexity of the model while improving the real-time processing capability of the aluminum liquid surface image features, and enhance the network model's perception of the fine-grained features of the aluminum liquid surface; secondly, the feature extraction transfer learning strategy is adopted to improve the efficiency of model training; then, the aluminum liquid surface area change within the set time window is calculated and the IQR method is used to determine the warning threshold of the liquid surface area change; this method is conducive to improving the accuracy and timeliness of judging aluminum liquid leakage, and enhancing the inherent safety of deep well casting in aluminum processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0043] Figure 1 This is a flow chart of a method for identifying and warning aluminum liquid leakage based on abnormal change detection of liquid surface area according to an embodiment of the present invention;

[0044] Figure 2 The camera of the embodiment of the present invention collects the image of the aluminum liquid surface area during the deep well casting process of aluminum processing;

[0045] Figure 3 This is a schematic diagram of the image processing process according to an embodiment of the present invention;

[0046] Figure 4 This is a schematic diagram of the improved DeepLapV3+ algorithm model structure;

[0047] Figure 5 This is a flowchart of offline training of the DeepLapV3+ algorithm model improved by an embodiment of the present invention;

[0048] Figure 6 It is a structural schematic diagram of an aluminum liquid leakage identification and early warning system based on abnormal liquid surface area change detection according to an embodiment of the present invention;

[0049] Figure 7 A schematic diagram of a server according to an embodiment of the present invention;

[0050] Figure 8 4 is a structural diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0051] In order to enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without making creative work are within the scope of protection of the present application.

[0052] Reference to "embodiments" in this application means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described in this application may be combined with other embodiments.

[0053] See also Figure 1 This embodiment provides a method for identifying and warning aluminum liquid leakage based on abnormal change detection of liquid surface area, comprising the following steps:

[0054] S1. Collect the aluminum liquid surface image, crop and frame the liquid surface image, and use the improved image semantic segmentation algorithm model to segment the aluminum liquid surface.

[0055] like Figure 2 and Figure 3 As shown in the figure, in order to detect the liquid surface of molten aluminum, the camera is first used to collect the image of the liquid surface area of ​​molten aluminum in real time, and the image of the liquid surface area is cropped and pre-processed by framing, and then the trained improved DeepLapV3+ algorithm model is used to segment the liquid surface of molten aluminum.

[0056] This embodiment adopts the following steps:

[0057] S11, using a camera to shoot the aluminum liquid surface area during the deep well casting process of aluminum processing, and obtaining an RGB image, such as Figure 2 As shown;

[0058] S12, collecting RGB images of the aluminum liquid surface area and performing image processing to establish an aluminum liquid surface image data set; the image processing includes image cropping, image denoising and image annotation, such as Figure 3 As shown;

[0059] 1) Image cropping: Use the same coordinates to extract the real aluminum liquid surface area from the read RGB image and obtain the preliminary image data set in batches, such as Figure 3 As shown in part (a).

[0060] 2) Image denoising: Under the premise of retaining image detail features, Gaussian filtering is used to eliminate initial image noise and improve the quality of image data sets, such as Figure 3 As shown in part (b).

[0061] 3) Image annotation: Labelme is used to annotate the denoised image and complete the annotation of the aluminum liquid surface segmentation label. The annotation categories are divided into two categories: "liquid" and "background". The "liquid" label area is the aluminum liquid surface area, and the corresponding image pixel value is 1; the "background" label area is the blank background area, and the corresponding image pixel value is 0, which is not considered in the training model process. Figure 3 As shown in part (c).

[0062] It should be emphasized that the improved image semantic segmentation algorithm model of this embodiment particularly adopts the improved DeepLapV3+ algorithm model. The liquid surface segmentation task involves two labels, the liquid surface area and the background area, with simple semantic information. The more easily optimized ResNet50 network is used to replace the Xception network as the liquid surface area change feature extraction network. The CBAM attention mechanism is introduced into the encoder and decoder structure of the DeeplabV3+ algorithm model structure to increase the network model's perception of the fine-grained features of the aluminum liquid surface.

[0063] Specifically, if Figure 4 As shown, the improved DeepLapV3+ algorithm model in this embodiment includes an encoder, a decoder, and an attention mechanism. The encoder structure includes a feature network extraction module and a void space convolution pooling pyramid module. The attention mechanism includes a channel attention module and a spatial attention module. The channel attention module and the spatial attention module are connected in series; the attention mechanism is introduced into the encoder structure, and the deep feature map extracted by the feature network extraction module is input into the attention mechanism, and the output end of the attention mechanism is connected to the void space convolution pooling pyramid module; the attention mechanism is introduced before the decoder structure, and the shallow feature map extracted by the feature network extraction module is input into the attention mechanism, and the output end of the attention mechanism is connected to the decoder; the output end of the void space convolution pooling pyramid module is connected to the decoder.

[0064] After completing the above steps, this embodiment uses transfer learning to train the model, such as Figure 5 As shown, the following steps are included:

[0065] Construct an improved image semantic segmentation algorithm model and use the feature extraction transfer learning strategy; the transfer learning strategy includes a model pre-training stage and a transfer stage. In the model pre-training stage, the improved image semantic segmentation algorithm model is pre-trained using the VOC dataset, a pre-training file in .pth format is generated, and pre-training weights are obtained;

[0066] In the migration stage, the improved image semantic segmentation algorithm model is retrained using the aluminum liquid surface image dataset and according to the pre-trained weights, and the training is stopped when the model converges.

[0067] S2. Calculate the area of ​​the molten aluminum surface according to the segmented image of the molten aluminum surface completed at the current moment, and calculate the change in the area of ​​the molten aluminum surface within the set time window.

[0068] S21. Calculate the total pixels of the liquid surface based on OpenCV, use the total value of the liquid surface pixels to represent the surface area of ​​the aluminum liquid, and calculate the surface area of ​​the aluminum liquid. The specific steps are as follows:

[0069] Capture the surveillance video image at the current time t at a frame rate of 1 frame / s; pass the surveillance video liquid surface image at the current time t into the trained DeepLabV3+ algorithm model for segmentation; adjust the segmented image to the original size; extract the type of each pixel; write the real-time aluminum liquid surface area; obtain the image liquid surface area S at the current time t t .

[0070] S22, calculating the change in the surface area of ​​the molten aluminum within the set time window.

[0071] Set the time ΔT = 3s as the time window of the change in the surface area of ​​the aluminum liquid; calculate the absolute value of the change in the surface area |ΔS| as the judgment indicator of the liquid level drop, and the calculation formula is as follows:

[0072] |ΔS|=|S t+ΔT -S t |

[0073] In the formula, |ΔS| represents the absolute value of the change in liquid surface area, S t represents the liquid surface area at the tth second, S t+ΔT Represents the liquid surface area at t+ΔT seconds.

[0074] S3. Use the IQR method to determine the warning threshold of the liquid surface area change, compare the size of the aluminum liquid surface area change and the warning threshold, and judge whether there is aluminum liquid leakage. When the aluminum liquid surface area change is greater than the warning threshold, it is judged as aluminum liquid leakage.

[0075] S31. Use the IQR method to determine the warning threshold of the liquid surface area change. The calculation method is as follows:

[0076] IQR=|Q1-Q3|

[0077] LB=1.5×IQR+Q3

[0078] In the formula, Q1 is the upper and lower quartiles, IQR is the quartile difference, LB is the abnormal lower limit, that is, the warning threshold of the liquid surface area change. The calculation process takes the absolute value of the liquid surface area change and adopts the upper limit value of IQR.

[0079] S32. Compare the change in the surface area of ​​the aluminum liquid |ΔS| with the warning threshold LB to determine whether there is an aluminum liquid leakage; if the change in the surface area |ΔS| is greater than the warning threshold LB, it is determined to be an aluminum liquid leakage.

[0080] It should be noted that, for the sake of convenience, the aforementioned method embodiments are all expressed as a series of action combinations, but those skilled in the art should know that the present invention is not limited to the described order of actions, because according to the present invention, certain steps can be performed in other orders or simultaneously.

[0081] Based on the same idea as the aluminum liquid leakage identification and early warning method based on abnormal liquid surface area change detection in the above-mentioned embodiment, the present invention also provides an aluminum liquid leakage identification and early warning system based on abnormal liquid surface area change detection, which can be used to execute the above-mentioned aluminum liquid leakage identification and early warning method based on abnormal liquid surface area change detection. For ease of explanation, the structural schematic diagram of the embodiment of the aluminum liquid leakage identification and early warning system based on abnormal liquid surface area change detection only shows the parts related to the embodiment of the present invention. Those skilled in the art can understand that the illustrated structure does not constitute a limitation on the device, and may include more or fewer components than the illustrated structure, or combine certain components, or arrange the components differently.

[0082] See also Figure 6 In another embodiment of the present application, there is provided an aluminum liquid leakage identification and early warning system 10 based on abnormal liquid surface area change detection, the system comprising a camera 11, a data acquisition box 12, a server 13 and a workstation 14;

[0083] Camera 11, used for capturing video images of the aluminum liquid surface area;

[0084] The data acquisition box 12 is used for collecting and transmitting video image data. The data acquisition box is connected to the camera via a network cable or an optical fiber.

[0085] Server 13, used for data storage, data processing, aluminum liquid leakage identification and early warning;

[0086] The workstation 14 is used to read the data stored in the server and display the aluminum liquid surface area, the aluminum liquid surface area change and the aluminum liquid leakage warning information within the set time window in the form of a visual interface.

[0087] like Figure 7As shown, the server 13 includes a data storage module 131, a data processing module 132, and an aluminum liquid leakage identification and aluminum leakage warning module 133;

[0088] The data storage module 131 is used to store the real-time collected aluminum liquid surface area image data, the image processing process data, the aluminum liquid surface area, the aluminum liquid surface area change within the set time window, and the time window and the warning threshold value of the liquid surface area change;

[0089] The data processing module 132 is used for cutting, framing and segmenting the aluminum liquid surface image, calculating the aluminum liquid surface area, and calculating the aluminum liquid surface area change within a set time window;

[0090] The aluminum liquid leakage identification and aluminum leakage warning module 133 is used to determine the warning threshold of the liquid surface area change by using the IQR method, compare the aluminum liquid surface area change with the warning threshold, and judge whether there is aluminum liquid leakage. When the aluminum liquid surface area change is greater than the warning threshold, it is judged as aluminum liquid leakage.

[0091] It should be noted that the aluminum liquid leakage identification and early warning system based on abnormal liquid surface area change detection of the present invention corresponds one to one with the aluminum liquid leakage identification and early warning method based on abnormal liquid surface area change detection of the present invention. The technical features and beneficial effects described in the embodiment of the aluminum liquid leakage identification and early warning method based on abnormal liquid surface area change detection are applicable to the embodiment of the aluminum liquid leakage identification and early warning method based on abnormal liquid surface area change detection. For specific contents, please refer to the description in the embodiment of the method of the present invention, which will not be repeated here. This is hereby declared.

[0092] In addition, in the implementation of the aluminum liquid leakage identification and early warning system based on abnormal liquid surface area change detection in the above-mentioned embodiment, the logical division of each program module is only an example. In actual applications, the above-mentioned functions can be assigned to different program modules as needed, for example, for the convenience of corresponding hardware configuration requirements or software implementation. That is, the internal structure of the aluminum liquid leakage identification and early warning system based on abnormal liquid surface area change detection is divided into different program modules to complete all or part of the functions described above.

[0093] See also Figure 8 In one embodiment, an electronic device that implements an aluminum liquid leakage identification and early warning method based on abnormal liquid surface area change detection is provided. The electronic device 20 may include a first processor 21, a first memory 22 and a bus, and may also include a computer program stored in the first memory 22 and executable on the first processor 21, such as an aluminum liquid leakage identification and early warning program 23 based on abnormal liquid surface area change detection.

[0094] The first memory 22 includes at least one type of readable storage medium, and the readable storage medium includes flash memory, mobile hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), magnetic memory, disk, optical disk, etc. In some embodiments, the first memory 22 can be an internal storage unit of the electronic device 20, such as a mobile hard disk of the electronic device 20. In other embodiments, the first memory 22 can also be an external storage device of the electronic device 20, such as a plug-in mobile hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), etc. equipped on the electronic device 20. Further, the first memory 22 can also include both an internal storage unit of the electronic device 20 and an external storage device. The first memory 22 can not only be used to store application software and various types of data installed in the electronic device 20, such as the code of the aluminum liquid leakage identification and early warning program 23 based on the abnormal change detection of the liquid surface area, but also can be used to temporarily store data that has been output or is to be output.

[0095] In some embodiments, the first processor 21 may be composed of an integrated circuit, for example, a single packaged integrated circuit, or a plurality of packaged integrated circuits with the same or different functions, including one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and combinations of various control chips. The first processor 21 is the control core (ControlUnit) of the electronic device, and uses various interfaces and lines to connect various components of the entire electronic device, and executes various functions and processes data of the electronic device 20 by running or executing programs or modules stored in the first memory 22, and calling data stored in the first memory 22.

[0096] Figure 8 Only an electronic device with components is shown, and those skilled in the art will understand that Figure 8 The structure shown does not constitute a limitation on the electronic device 20 , and may include fewer or more components than shown in the figure, or combine certain components, or arrange the components differently.

[0097] The aluminum liquid leakage identification and early warning program 23 based on abnormal liquid surface area change detection stored in the first memory 22 of the electronic device 20 is a combination of multiple instructions. When running in the first processor 21, it can achieve:

[0098] Collect the aluminum liquid surface image, crop and frame the liquid surface image, and use the improved image semantic segmentation algorithm model to segment the aluminum liquid surface;

[0099] According to the aluminum liquid surface segmentation image completed at the current moment, the aluminum liquid surface area is calculated, and the aluminum liquid surface area change within the set time window is calculated;

[0100] The IQR method is used to determine the warning threshold of the liquid surface area change. The change in the aluminum liquid surface area is compared with the warning threshold to determine whether there is an aluminum liquid leakage. When the change in the aluminum liquid surface area is greater than the warning threshold, it is determined to be an aluminum liquid leakage.

[0101] Furthermore, if the module / unit integrated in the electronic device 20 is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a non-volatile computer-readable storage medium. The computer-readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a mobile hard disk, a magnetic disk, an optical disk, a computer memory, and a read-only memory (ROM).

[0102] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0103] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0104] The above embodiments are preferred implementation modes of the present invention, but the implementation modes of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, and simplifications that do not deviate from the spirit and principles of the present invention should be equivalent replacement methods and are included in the protection scope of the present invention.

Claims

1. A method for identifying and warning aluminum liquid leakage based on abnormal change detection of liquid surface area, characterized in that: The steps include: Collect the aluminum liquid surface image, crop and frame the liquid surface image, and use the improved image semantic segmentation algorithm model to segment the aluminum liquid surface; According to the aluminum liquid surface segmentation image completed at the current moment, the aluminum liquid surface area is calculated, and the aluminum liquid surface area change within the set time window is calculated; The IQR method is used to determine the warning threshold of the liquid surface area change. The change in the aluminum liquid surface area is compared with the warning threshold to determine whether there is an aluminum liquid leakage. When the change in the aluminum liquid surface area is greater than the warning threshold, it is determined to be an aluminum liquid leakage.

2. The aluminum liquid leakage identification and early warning method based on abnormal liquid surface area change detection according to claim 1 is characterized in that: The improved image semantic segmentation algorithm model comprises an encoder, a decoder and an attention mechanism, wherein the encoder structure comprises a feature network extraction module and a void space convolution pooling pyramid module, the attention mechanism comprises a channel attention module and a space attention module, and the channel attention module and the space attention module are connected in series; the attention mechanism is introduced into the encoder structure, the deep feature map extracted by the feature network extraction module is input into the attention mechanism, and the output end of the attention mechanism is connected to the void space convolution pooling pyramid module; the attention mechanism is introduced before the decoder structure, the shallow feature map extracted by the feature network extraction module is input into the attention mechanism, and the output end of the attention mechanism is connected to the decoder; the output end of the void space convolution pooling pyramid module is connected to the decoder.

3. The aluminum liquid leakage identification and early warning method based on abnormal liquid surface area change detection according to claim 2 is characterized in that: Including offline training of improved image semantic segmentation algorithm model, specifically: Collecting RGB images of the aluminum liquid surface area and performing image processing to establish an aluminum liquid surface image data set; the image processing includes image cropping, image denoising and image annotation; Construct an improved image semantic segmentation algorithm model and use a transfer learning strategy for feature extraction; the transfer learning strategy includes a model pre-training stage and a transfer stage. In the model pre-training stage, the improved image semantic segmentation algorithm model is pre-trained using the VOC dataset, a pre-training file in .pth format is generated, and pre-training weights are obtained; In the migration stage, the improved image semantic segmentation algorithm model is retrained using the aluminum liquid surface image dataset and according to the pre-trained weights, and the training is stopped when the model converges.

4. The aluminum liquid leakage identification and early warning method based on abnormal liquid surface area change detection according to claim 1 is characterized in that: The calculation of the aluminum liquid surface area is specifically as follows: Read the aluminum liquid surface segmentation image completed at the current moment and adjust the image to its original size; extract the pixel points of the aluminum liquid surface part in the image, calculate the total liquid surface pixels based on OpenCV, use the total value of the liquid surface pixels to represent the liquid surface area, and calculate the image liquid surface area at the current moment.

5. The aluminum liquid leakage identification and early warning method based on abnormal liquid surface area change detection according to claim 1 is characterized in that: The calculation of the change in the surface area of ​​the molten aluminum within the set time window includes: The set time ΔT is used as the time window for the change in the surface area of ​​the aluminum liquid; Calculate the absolute value of the change in liquid surface area as follows: |ΔS|=|S t+ΔT -S t | Among them, |ΔS| represents the absolute value of the change in liquid surface area, S t represents the liquid surface area at the tth second, S t+ΔT Represents the liquid surface area at t+ΔT seconds.

6. The aluminum liquid leakage identification and early warning method based on abnormal liquid surface area change detection according to claim 1 is characterized in that: The IQR method is used to determine the warning threshold of the liquid surface area change, as follows: IQR=|Q1-Q3| LB=λ×IQR+Q3 In the formula, Q1 represents the upper and lower quartiles, IQR represents the quartile range, LB represents the warning threshold of the liquid surface area change, and λ is a fixed constant.

7. An aluminum liquid leakage identification and early warning system based on abnormal liquid surface area change detection, characterized in that: The aluminum liquid leakage identification and early warning method based on abnormal liquid surface area change detection applied to any one of claims 1 to 6 comprises a camera, a data acquisition box, a server and a workstation; A camera, used to capture video images of the aluminum liquid surface area; Data collection box, used for collecting and transmitting video image data, the data collection box is connected to the camera through network cable and optical fiber; Server, used for data storage, data processing, aluminum liquid leakage identification and early warning; The workstation is used to read the data stored in the server and display the aluminum liquid surface area, aluminum liquid surface area change and aluminum liquid leakage warning information within the set time window in the form of a visual interface.

8. The aluminum liquid leakage identification and early warning system based on abnormal liquid surface area change detection according to claim 7 is characterized in that: The server includes a data storage module, a data processing module, and an aluminum liquid leakage identification and aluminum leakage early warning module; A data storage module is used to store the real-time collected aluminum liquid surface area image data, image processing process data, aluminum liquid surface area, aluminum liquid surface area change within a set time window, and the time window and liquid surface area change warning threshold; The data processing module is used for cutting, framing and segmenting the aluminum liquid surface image, calculating the aluminum liquid surface area, and calculating the change in the aluminum liquid surface area within a set time window; The aluminum liquid leakage identification and aluminum leakage warning module is used to determine the warning threshold of the liquid surface area change by using the IQR method, compare the aluminum liquid surface area change with the warning threshold, and judge whether there is aluminum liquid leakage. When the aluminum liquid surface area change is greater than the warning threshold, it is judged as aluminum liquid leakage.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores computer program instructions that can be executed by the at least one processor, and the computer program instructions are executed by the at least one processor so that the at least one processor can execute the aluminum liquid leakage identification and early warning method based on abnormal liquid surface area change detection as described in any one of claims 1-6.

10. A computer-readable storage medium storing a program, characterized in that: When the program is executed by the processor, the aluminum liquid leakage identification and early warning method based on abnormal liquid surface area change detection as described in any one of claims 1 to 6 is implemented.

Images