A top-blown furnace smelting online monitoring method and device based on a forced air cooling vision system, an electronic device, and a storage medium

Abstract

The present application relates to non-ferrous metallurgy equipment and computer vision processing technical field, especially to top blown converter smelting online monitoring method based on air cooling vision system, including the following steps: first, through the air cooling industrial camera arranged in the observation port of the top blown converter, the smelting scene in the top blown converter is online photographed, and the photographed real-time video of smelting in the top blown converter is sent to the industrial computer, then the image in the photographed real-time video of smelting in the top blown converter is sequentially subjected to logarithmic transformation, fog removal processing, weighted fusion and color restoration / enhancement preprocessing through the industrial computer, and the image after logarithmic transformation, fog removal processing, weighted fusion and color restoration / enhancement preprocessing is subjected to abnormality identification in combination with the trained abnormal working condition identification model of smelting in the top blown converter; if the abnormal working condition is identified, the alarm is given and the corresponding adjustment of the lance is guided by the personnel in the central control room; in this way, the smelting scene in the furnace can be observed in real time without lifting the lance and stopping the air, and the image fog removal processing can also penetrate the thick smoke in the furnace, clearly restore the position of the lance and the fluctuation of the molten pool liquid level, so as to solve the problem that the image collected by the existing smelting furnace smelting scene is blurred and unclear due to the thick smoke and high temperature red light radiation in the top blown converter, and further unable to provide effective guidance to the workers responsible for the smelting of the top blown converter.

Description

Technical Field

[0001] This invention relates to the field of non-ferrous metal metallurgical equipment and computer vision processing technology, and in particular to an online monitoring method, device, electronic equipment and storage medium for top-blown furnace smelting based on an air-cooled vision system. Background Technology

[0002] In the top-blown smelting process, the flow of melt in the top-blown furnace, the splashing of the spray gun, and the erosion of refractory bricks are directly related to production safety and efficiency. Therefore, monitoring the top-blown smelting process is of utmost importance.

[0003] The traditional monitoring method for top-blown furnace smelting relies on manual observation through periodic opening of the furnace top observation holes. However, this method requires "raising the lance, stopping the air supply, and stopping the material supply" before observation can begin. This not only exposes workers to the risk of high-temperature radiation burns, but also makes it difficult to see the specific details of the molten pool inside the furnace due to the dense smoke. Although some factories have tried installing monitoring cameras, ordinary protective covers are insufficient to withstand high-temperature splashes, and the lenses are easily blocked by smoke and dust. More seriously, the dense smoke and high-temperature red light radiation inside the top-blown furnace make the acquired images blurry, appearing as a "white blur" or "red blur," failing to provide effective guidance to the workers responsible for top-blown furnace smelting and indirectly affecting the smelting process.

[0004] Therefore, it is necessary to improve and optimize the online monitoring method for top-blown furnace smelting to solve the problem that the images of the existing smelting scene are blurred due to the dense smoke and high-temperature red light radiation inside the top-blown furnace, which makes it impossible to provide effective guidance to the workers responsible for top-blown furnace smelting. Summary of the Invention

[0005] The purpose of this invention is to propose an online monitoring method, device, electronic equipment, and storage medium for top-blown furnace smelting based on an air-cooled vision system, in order to solve the problem that existing images of smelting furnace scenes are blurred due to the dense smoke and high-temperature red light radiation inside the top-blown furnace, thus failing to provide effective guidance to the workers responsible for top-blown furnace smelting.

[0006] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: This application proposes, in its first aspect, a method for online monitoring of top-blown furnace smelting based on an air-cooled vision system, comprising the following steps: An air-cooled industrial camera installed at the observation port of the top-blown furnace is used to capture the smelting scene inside the furnace online, and the captured real-time video of the smelting inside the furnace is sent to the industrial control computer. The industrial control computer preprocesses images from real-time video of smelting inside the top-blown furnace and combines them with a trained abnormal smelting condition recognition model to identify anomalies in the preprocessed images. If an abnormal condition is detected, an alarm is triggered, and personnel in the central control room are instructed to make corresponding adjustments to the spray guns. The preprocessing includes sequentially performing logarithmic transformation, defogging, weighted fusion, and color restoration / enhancement preprocessing on the images. The abnormal smelting condition recognition model for the top-blown furnace is trained based on images of typical abnormal smelting conditions inside the top-blown furnace and is constructed based on a convolutional neural network.

[0007] Preferably, the step of using an air-cooled industrial camera installed at the observation port of the top-blown furnace to capture online footage of the smelting process inside the furnace and sending the captured real-time video of the smelting process inside the furnace to an industrial control computer includes the following steps: The air-cooled gas assembly is controlled to deliver cooling air to the cooling air duct inside the air-cooled industrial camera located at the observation port of the top-blown furnace. The cooling air is then blown out through the dustproof air curtain nozzle near the shooting end of the air-cooled industrial camera, blowing away the molten slag that flies toward the shooting window of the air-cooled industrial camera. While the dustproof air curtain nozzles blow away the molten slag flying towards the viewing window of the air-cooled industrial camera in real time, the air-cooled industrial camera is controlled to shoot the smelting scene inside the top-blown furnace online, and the real-time video of the smelting inside the top-blown furnace is transmitted to the remote industrial control computer.

[0008] Preferably, the industrial control computer includes an image compression module, an image dehazing and enhancement module, a color restoration module, an abnormal smelting condition recognition module in the top-blown furnace, and an abnormal condition alarm module; The industrial control computer preprocesses images from real-time video of smelting inside the top-blown furnace and uses a trained abnormal smelting condition recognition model to identify anomalies in the preprocessed images. If an abnormal condition is detected, an alarm is triggered, and personnel in the central control room are instructed to make corresponding adjustments to the spray guns, including the following steps: The image compression module performs logarithmic transformation compression on any image in the real-time video of melting inside the top-blown furnace to obtain the compressed image. The compressed image is processed by an image dehazing enhancement module to remove smoke and dust interference, resulting in an image with smoke and dust interference removed. The image with smoke and dust interference removed is then weighted and fused with the image processed by the multi-scale Retinex algorithm to obtain the fused image. The color restoration module corrects the color cast in the fused image to obtain the corrected image. The abnormal working conditions of the smelting abnormal working conditions in the top blown furnace are detected and identified by the corrected image through the abnormal working condition identification module. If the smelting conditions inside the top-blown furnace are abnormal, an alarm will be triggered through the abnormal condition alarm module, and the smelting condition information inside the top-blown furnace will be pushed to the personnel in the central control room to guide them to make corresponding adjustments to the spray guns. The corresponding adjustments to the spray guns include adjusting the height of the spray guns or adjusting the air volume of the spray guns.

[0009] Preferably, the industrial control computer also includes: The image storage and query module is used to store the fused images processed by the image dehazing and enhancement module, and to display previous fused images in response to image query requests; among them, the previous fused images are high-definition videos after dehazing within the last 30 days.

[0010] Preferably, the image dehazing enhancement module is specifically used for: Based on the improved dark channel prior algorithm, smoke and dust interference removal processing is performed on the compressed image to obtain the processed image; the improved dark channel prior algorithm is constructed by combining the modified atmospheric light value selection strategy and the introduction of red light channel suppression factor to correct the transmittance calculation. Based on the guided filtering algorithm, the processed image is optimized to obtain the optimized image, that is, the image after removing smoke and dust interference; Multi-scale feature image extraction is performed on the compressed image based on the multi-scale Retinex algorithm. The image extracted by the multi-scale Retinex algorithm is then weighted and fused with the image after removing smoke and dust interference to obtain the fused image.

[0011] Preferably, the abnormal smelting condition identification module in the top-blown furnace is specifically used for: A model for identifying abnormal smelting conditions in a top-blown furnace is constructed based on a convolutional neural network. A dataset of smelting images inside the top-blown furnace is obtained. This dataset is then input into the top-blown furnace smelting abnormal condition recognition model for training, resulting in a trained model. The top-blown furnace smelting image dataset includes images of normal splashing inside the top-blown furnace, images of excessive foam slag inside the top-blown furnace, and images of nozzle formation on the top-blown furnace lance.

[0012] Preferably, the color restoration module is specifically used for: The color cast in the fused image is corrected by a white balance algorithm to obtain a corrected image. The color cast in the image is reddish due to high temperature radiation. The corrected image shows molten slag, molten metal and refractory bricks in a pseudo-color or grayscale level that is easy to identify.

[0013] This application proposes, in its second aspect, an online monitoring device for top-blown furnace smelting based on an air-cooled vision system, comprising: An air-cooled industrial camera is installed at the observation port of the top-blown furnace. It is used to record the smelting scene inside the top-blown furnace after blowing away the molten slag that flies towards the camera's viewing window during the process of filming the smelting scene inside the top-blown furnace. The air-cooled cooling component is connected to the air intake end of the cooling channel of the air-cooled industrial camera. It is used to provide cooling air to the cooling channel of the air-cooled industrial camera. The air is then sprayed out through the air curtain nozzle connected to the cooling channel and located at the shooting window of the air-cooled industrial camera, blowing away the molten slag flying towards the shooting window of the air-cooled industrial camera. The industrial control computer communicates with the air-cooled industrial camera to receive video of the smelting scene inside the top-blown furnace and identify whether there are any abnormal conditions in the images of the smelting scene inside the top-blown furnace; if there are abnormal conditions, an alarm will be triggered.

[0014] Preferably, the air-cooled industrial camera includes a cooling outer sleeve, an industrial camera, and an air-cooled inner sleeve. The cooling outer sleeve is a cavity structure with both ends closed. The cooling outer sleeve is connected to the observation port of the top-blown furnace. The air-cooled inner sleeve is connected inside the cooling outer sleeve. The industrial camera is fixedly installed inside the air-cooled inner sleeve. The shooting end of the industrial camera is on the same axis as the open end of the air-cooled inner sleeve and the transparent window connected to the cooling outer sleeve. A cooling channel is defined between the air-cooled inner sleeve and the cooling outer sleeve. The air inlet end of the cooling channel is connected to the air outlet end of the air-cooled cooling component. The cooling channel is connected to multiple air curtain nozzles embedded in the transparent window connected to the cooling outer sleeve. The multiple air curtain nozzles are arranged in a ring along the transparent window. The extension line of the shooting end of the industrial camera is located at the midpoint of the ring.

[0015] Preferably, the air-cooled assembly includes a pressure regulating air pump, an air pipe, an air valve, and an air pressure detector. The air pressure detector is electrically connected to the air valve via a controller. Both the air valve and the air pressure detector are mounted on the air pipe. The first end of the air pipe is connected to the air outlet of the pressure regulating air pump, and the second end of the air pipe is connected to the air inlet on the outer cooling sleeve. The air inlet is interconnected with the cooling channel defined between the inner air-cooled sleeve and the outer cooling sleeve.

[0016] This application proposes an electronic device in a third aspect, including... One or more processors; A memory is used to store one or more programs; when one or more programs are executed by one or more processors, the one or more processors implement an online monitoring method for top-blown furnace melting based on an air-cooled vision system.

[0017] In its fourth aspect, this application proposes a computer-readable storage medium storing computer instructions for causing a processor to execute an online monitoring method for top-blown furnace melting based on an air-cooled vision system.

[0018] Compared with the prior art, the present invention has at least one of the following beneficial effects: This invention discloses an online monitoring method for top-blown furnace smelting based on an air-cooled vision system. First, an air-cooled industrial camera installed at the observation port of the top-blown furnace captures the smelting scene inside the furnace online. The captured real-time video of the smelting process is then sent to an industrial control computer. The industrial control computer then performs logarithmic transformation, dehazing, weighted fusion, and color restoration / enhancement preprocessing on the images from the captured real-time video. Finally, a trained abnormal smelting condition recognition model for the top-blown furnace is used to perform logarithmic transformation, dehazing, and weighted fusion preprocessing. The image, after weight fusion and color restoration / enhancement preprocessing, is used for anomaly identification. If an abnormal condition is detected, an alarm is triggered and the control room personnel are instructed to make corresponding adjustments to the spray gun. In this way, the smelting scene inside the furnace can be observed in real time without picking up the gun and stopping the air supply. Moreover, the image defogging process can penetrate the dense smoke inside the furnace and clearly restore the position of the spray gun and the fluctuation of the molten pool surface. This solves the problem that existing images of smelting furnace scenes are blurred due to the dense smoke and high-temperature red light radiation inside the top-blown furnace, which makes it impossible to provide effective guidance to the workers responsible for smelting in the top-blown furnace.

[0019] This invention places an air-cooled industrial camera at the observation port of a top-blown furnace. The air-cooling component is connected to the air inlet of the cooling channel of the air-cooled industrial camera. The cooling channel is connected to multiple air curtain nozzles located on the camera's viewing window. Before the air-cooled industrial camera captures video of the smelting scene inside the top-blown furnace, the air-cooling component can be controlled to input cooling air into the cooling channel of the air-cooled industrial camera. The cooling air entering the cooling channel cools the air-cooled industrial camera along the cooling channel to prevent damage from the high temperature radiated by the top-blown furnace. On the other hand, it flows along the cooling channel to the air curtain nozzles connected to the cooling channel and located at the camera's viewing window, blowing away the molten slag flying towards the camera's viewing window. This not only effectively prevents molten slag and dust from adhering in front of the air-cooled industrial camera, achieving maintenance-free operation, but also improves the clarity of the video captured by the air-cooled industrial camera.

[0020] 3. An online monitoring method for top-blown furnace smelting based on an air-cooled vision system. Personnel responsible for the top-blown furnace smelting process do not need to approach the high-temperature furnace opening to observe the smelting scene inside the top-blown furnace. Furthermore, the image storage and query module is connected to the image defogging and enhancement module, which can store and query videos of the top-blown furnace smelting scene in the most recent time period, providing objective basis for process optimization and accident analysis. Attached Figure Description

[0021] Figure 1 This is a flowchart of an online monitoring method for top-blown furnace smelting based on an air-cooled vision system, as described in this invention.

[0022] Figure 2 This is a schematic diagram of the structure of the present invention. Figure 1 .

[0023] Figure 3 This is a schematic diagram of the internal structure of the air-cooled industrial camera in the present invention.

[0024] Figure 4 This is a schematic diagram of the structure of the present invention. Figure 2 .

[0025] Figure 5 This is a schematic diagram of the structure of an electronic device according to the present invention.

[0026] In the diagram, 1-air-cooled industrial camera, 2-top-blown furnace observation port, 3-top-blown furnace, 4-air-cooled cooling component, 5-cooling channel, 6-air curtain nozzle, 7-industrial computer, 8-cooled outer sleeve, 9-industrial camera, 10-air-cooled inner sleeve, 11-transparent window, 12-pressure regulating air pump, 13-air pipe, 14-air valve, 15-air pressure detector, 16-air inlet. Detailed Implementation

[0027] like Figure 1-5 As shown, to make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present invention and are not intended to limit the present invention.

[0028] Example 1 Figure 1 This invention provides a flowchart of an online monitoring method for top-blown furnace smelting based on an air-cooled vision system, as described in Embodiment 1. This embodiment is applicable to online monitoring of top-blown furnace smelting scenarios. The method can be executed by an online monitoring device for top-blown furnace smelting based on an air-cooled vision system. This device can be implemented in hardware and / or software and can be configured in an electronic device with data processing capabilities. Figure 1 As shown, the method includes: S110. The air-cooled industrial camera installed at the observation port of the top-blown furnace is used to capture the smelting scene inside the top-blown furnace online, and the captured real-time video of the smelting inside the top-blown furnace is sent to the industrial control computer.

[0029] The observation port of the top-blown furnace is specifically located on the upper right side of the furnace wall and is interconnected with the furnace interior. The air-cooled industrial camera is an industrial camera with a cooling duct. The industrial camera can be selected based on its heat resistance. The melting scene inside the top-blown furnace includes the spray gun, molten metal, and melting process. The real-time video of melting inside the top-blown furnace is an integrated video showing the spray gun, molten metal, and melting process.

[0030] In this technical solution, an air-cooled industrial camera installed at the observation port of the top-blown furnace is used to capture the smelting scene inside the furnace online. The captured real-time video of the smelting inside the top-blown furnace is then sent to the industrial control computer. The air-cooled industrial camera can not only penetrate the top-blown furnace to capture the entire smelting process inside the furnace, but also monitor the real-time scene of material reaction, flame shape, material layer distribution, and molten pool status. Furthermore, the real-time smelting video is simultaneously transmitted to the industrial control computer, so operators do not need to be close to the high-temperature furnace body.

[0031] Optionally, an air-cooled industrial camera installed at the observation port of the top-blown furnace is used to capture online footage of the smelting process inside the furnace, and the captured real-time video of the smelting process inside the furnace is sent to an industrial control computer, including steps A1 to A2: A1. Control the air-cooled gas assembly to deliver cooling air to the cooling air duct inside the air-cooled industrial camera located at the observation port of the top-blown furnace. The cooling air is then blown out through the dustproof air curtain nozzle near the shooting end of the air-cooled industrial camera, blowing away the molten slag flying towards the shooting window of the air-cooled industrial camera.

[0032] The dustproof air curtain nozzles are arranged around the viewing window of the air-cooled industrial camera, which faces the observation port of the top-blown furnace. The dustproof air curtain nozzles at the camera's shooting end are connected to the observation port of the top-blown furnace. The cooling air duct is arranged around the air-cooled industrial camera; that is, the cooling air duct surrounds the air-cooled industrial camera. For details on the specific arrangement of the cooling air duct and the air-cooled industrial camera, please refer to [reference needed]. Figure 3 , Figure 3 A schematic diagram of the internal structure of the air-cooled industrial camera in this invention; In this technical solution, cooling air is supplied to the cooling duct inside the air-cooled industrial camera located at the observation port of the top-blown furnace by controlling the air-cooled gas assembly. The cooling air is then blown out through the dustproof air curtain nozzle near the shooting end of the air-cooled industrial camera, blowing away the molten slag flying towards the shooting window of the air-cooled industrial camera. This not only effectively prevents the molten slag from hitting and scratching the lens glass and accelerating the aging of components due to high temperature, but also saves on the manpower and material costs of later maintenance. In addition, it effectively resists the strong heat radiation and high temperature baking at the observation port of the top-blown furnace.

[0033] A2. While the dustproof air curtain nozzles blow away the molten slag flying towards the viewing window of the air-cooled industrial camera in real time, control the air-cooled industrial camera to shoot the smelting scene inside the top-blown furnace online, and transmit the real-time video of the smelting inside the top-blown furnace to the remote industrial control computer.

[0034] Among them, the air-cooled industrial camera takes online pictures of the smelting scene inside the top-blown furnace, which is a real-time online picture of the smelting scene inside the top-blown furnace. At the same time that the dust curtain nozzles blow away the molten slag flying towards the camera's viewing window, that is, when the top-blown furnace smelting scene is first photographed, the dust curtain nozzles blow away the molten slag flying towards the camera's viewing window before the air-cooled industrial camera starts shooting video of the top-blown furnace smelting scene.

[0035] In this technical solution, while the dustproof air curtain nozzles blow away the molten slag flying towards the viewing window of the air-cooled industrial camera in real time, the air-cooled industrial camera is controlled to shoot the smelting scene inside the top-blown furnace online, and the real-time video of the smelting inside the top-blown furnace is transmitted to the remote industrial control computer. This not only effectively avoids molten slag adhesion and dust obstructing the viewing window, ensuring a clear field of view for the camera from the source, but also allows for intuitive monitoring of the smelting situation inside the furnace from the remote operating area, avoiding the safety hazards of high temperature and burns on site.

[0036] S120: The industrial control computer preprocesses the images in the real-time video of smelting in the top-blown furnace and combines them with the trained abnormal working condition recognition model of smelting in the top-blown furnace to identify abnormalities in the preprocessed images; if an abnormal working condition is identified, an alarm is triggered and the personnel in the central control room are instructed to make corresponding adjustments to the spray gun.

[0037] The preprocessing includes sequentially performing logarithmic transformation, dehazing, weighted fusion, and color restoration / enhancement preprocessing on the image; the top-blown furnace smelting abnormal condition identification model is trained based on typical abnormal smelting images in the top-blown furnace; the top-blown furnace smelting abnormal condition identification model is constructed based on a convolutional neural network.

[0038] In this technical solution, the industrial control computer sequentially performs logarithmic transformation, defogging, weighted fusion, and color restoration / enhancement preprocessing on the images captured in the real-time video of smelting inside the top-blown furnace. Combined with a trained abnormal smelting condition recognition model for the top-blown furnace, the preprocessed images are used for anomaly identification. If an abnormal condition is detected, an alarm is triggered, and the control room personnel are instructed to adjust the spray gun accordingly. This not only effectively eliminates on-site interference such as strong light, dense smoke, dust, uneven lighting, and color casts inside the furnace, restoring the details of the real smelting scene and laying a solid image foundation for accurate identification, but also immediately triggers an alarm signal and pushes abnormal information upon detection of abnormal smelting conditions, avoiding the lag that can occur with manual detection.

[0039] Optionally, the industrial control computer includes an image compression module, an image dehazing and enhancement module, a color restoration module, an abnormal smelting condition recognition module for the top-blown furnace, and an abnormal smelting condition alarm module. The industrial control computer preprocesses images from the real-time video of smelting in the top-blown furnace and uses a trained abnormal smelting condition recognition model to identify anomalies in the preprocessed images. If an abnormal condition is detected, an alarm is triggered, and the control room personnel are instructed to make corresponding adjustments to the spray guns, including steps B1-B5. B1. The image compression module performs logarithmic transformation compression on any image in the real-time video of melting inside the top-blown furnace to obtain the compressed image.

[0040] The image logarithmic transformation compression process involves mapping the grayscale values ​​of all pixels in the image using a logarithmic function, compressing the dynamic range of grayscale in high-brightness areas, while simultaneously stretching and brightening low-grayscale dark details, significantly reducing the overall brightness difference in the image, and suppressing overexposure in strong light and pixel overflow in bright areas.

[0041] In this technical solution, the image compression module performs logarithmic transformation compression on any image in the real-time video of melting inside the top-blown furnace to obtain the compressed image. This not only suppresses overexposure of highlights and weakens the interference of strong flame light, but also brightens dark areas, highlights hidden working conditions inside the furnace, balances the grayscale of the image, and reduces the dynamic difference between light and dark.

[0042] B2. The compressed image is processed by the image dehazing enhancement module to remove smoke and dust interference, so as to obtain an image with smoke and dust interference removed. The image with smoke and dust interference removed is then weighted and fused with the image processed by the multi-scale Retinex algorithm to obtain the fused image.

[0043] Among them, the multi-scale Retinex algorithm is an image enhancement algorithm developed based on Edwin Land's Retinex vision theory. By fusing the processing results of multiple single-scale Retinex (SSR) algorithms at different scales, it achieves a balanced optimization of image dynamic range compression, local detail enhancement and color constancy preservation.

[0044] The specific processing steps of the image dehazing enhancement module are as follows: Based on the improved dark channel prior algorithm, the compressed image is processed to remove smoke and dust interference to obtain the processed image; the improved dark channel prior algorithm is constructed based on the modified atmospheric light value selection strategy and the introduction of a red light channel suppression factor to correct the transmittance calculation; based on the guided filtering algorithm, the processed image is optimized to obtain the optimized image, that is, the image after removing smoke and dust interference; based on the multi-scale Retinex algorithm, multi-scale feature image extraction is performed on the compressed image, and the image extracted by the multi-scale Retinex algorithm is weighted and fused with the image after removing smoke and dust interference to obtain the fused image.

[0045] It should be noted that the guided filtering algorithm uses a guide image to guide the filtering process of the input image, so that the output image can both smooth noise and preserve the edge structure in the guide image.

[0046] It should be noted that after acquiring the video frames from inside the furnace, highlight and flame suppression must be performed on the images. Because the molten metal and flames inside the furnace create a large number of specular reflections and self-illuminating areas, they severely interfere with the estimation of traditional atmospheric light values. If the pixels with the lowest brightness of the first 0.1% are directly selected, it is highly likely that the boiling molten steel will be mistakenly identified as "atmospheric light from infinity," leading to errors in the reconstruction of the entire image.

[0047] Therefore, this step uses multi-scale Retinex or color ratio-based flame segmentation. First, the pixels of the flame and the molten pool are marked and removed. Only in the remaining smoke-filled area, a fast quadtree search is used to iteratively lock the uniform gray-white area representing the "most dense smoke". This is used as a reliable atmospheric light value A. Then, an inter-frame first-order hysteresis filter is applied to this value to prevent the estimated value from jittering due to the churning inside the furnace.

[0048] In actual transmittance calculations, the physical meaning of the dark channel needs to be redefined. The traditional "dark channel prior" assumes that fog-free images always have very dark pixels, but the furnace background consists of hot refractory bricks and dark red or bright white furnace charge, which does not conform to this assumption. Therefore, this application uses an inverted dark channel or a correction based on physical imaging. Specifically, based on the spectral absorption characteristics of typical furnace dust (industrial dust often exhibits gray-white broadband absorption), the color image is first converted to a color space that highlights the contrast between the flue gas and the background. Then, a method called "joint estimation of bright and dark channels" is used. Specifically, for areas that are inherently luminescent, such as the furnace wall, the bright channel is used to constrain the lower bound of transmittance; for darker areas of furnace charge, the dark channel constraint is retained. The two are fused pixel-by-pixel through a weighted map determined by the local saturation and gradient of the image, resulting in an initial transmittance map that can accurately distinguish between "dust obstruction" and "naturally dark" areas.

[0049] This technical solution uses an image dehazing enhancement module to remove smoke and dust interference from the compressed image, resulting in an image free of smoke and dust interference. The image free of smoke and dust interference is then weighted and fused with an image processed by the multi-scale Retinex algorithm to obtain a fused image. This fused image not only smooths the edges of the rough transmittance map, thus preserving the edge details of the spray gun and the molten pool surface, but also penetrates the dense smoke inside the furnace, clearly restoring the position of the spray gun and the fluctuation of the molten pool surface.

[0050] It should be noted that the industrial control computer also includes an image storage and query module, which is connected to the image dehazing and enhancement module. The image storage and query module is used to store the fused images processed by the image dehazing and enhancement module, and to respond to image query requests by displaying previous fused images; among them, the previous fused images are high-definition videos after dehazing within the last 30 days.

[0051] B3. The color restoration module corrects the color cast in the fused image to obtain the corrected image.

[0052] Specifically, the color restoration module corrects the color cast in the fused image using a white balance algorithm to obtain a corrected image. The color cast in the image is caused by a reddish tint due to high-temperature radiation. The corrected image presents molten slag, molten metal, and refractory bricks in a pseudo-color or grayscale level that is easy to identify. The color restoration module uses standard white as the color reference benchmark and adjusts the pixel color gain and pixel value ratio of the RGB three-channel pixels in the image to correct the color cast and color distortion caused by abnormal ambient light color temperature, furnace flame color light, and special lighting in high-temperature smelting. It restores the true and natural inherent colors of the objects in the scene and eliminates color distortions such as reddish, yellowish, and dark colors, thereby achieving a unified overall image tone and accurate color restoration.

[0053] This technical solution corrects color casts in the fused image using a color restoration module to obtain a corrected image. This not only effectively corrects issues such as reddish or yellowish tones and color temperature imbalances caused by fire, high-temperature light sources, and smoke scattering in the smelting environment, but also accurately restores the original true colors of the molten pool, materials, furnace walls, and flames inside the furnace. At the same time, it eliminates feature interference caused by color casts, allowing subsequent abnormal smelting condition recognition models to accurately make judgments based on standard color features, significantly reducing misjudgments and missed judgments caused by color distortion.

[0054] B4. The abnormal working condition detection and identification process is performed on the corrected image through the abnormal working condition identification module in the top blown furnace.

[0055] The specific processing steps of the top-blown furnace smelting abnormal condition identification module are as follows: Based on the convolutional neural network, a top-blown furnace smelting abnormal condition identification model is constructed; a top-blown furnace smelting image dataset is obtained, and the top-blown furnace smelting image dataset is input into the top-blown furnace smelting abnormal condition identification model for training, so as to obtain the trained top-blown furnace smelting abnormal condition identification model. The top-blown furnace smelting image dataset includes images of normal splashing inside the top-blown furnace, images of excessive foam slag inside the top-blown furnace, and images of nozzle formation on the spray gun inside the top-blown furnace.

[0056] This technical solution uses an abnormal smelting condition identification module in a top-blown furnace to detect and identify abnormal conditions in the corrected images. This eliminates the need for manual frame-by-frame observation, as the system automatically performs image analysis and condition judgment, thus overcoming judgment biases caused by human visual fatigue and experience differences, and achieving intelligent and autonomous detection of smelting conditions.

[0057] B5. If the smelting conditions in the top-blown furnace are abnormal, an alarm will be triggered through the abnormal conditions alarm module, and the smelting conditions information in the top-blown furnace will be pushed to the personnel in the central control room to guide them to make corresponding adjustments to the spray guns.

[0058] Among them, the corresponding adjustments to the spray gun are to adjust the spray gun height or adjust the spray gun airflow.

[0059] This technical solution, if the smelting conditions inside the top-blown furnace are abnormal, will trigger an alarm through the abnormal condition alarm module and push the smelting condition information inside the top-blown furnace to the central control room personnel, guiding them to make corresponding adjustments to the spray guns. It can not only simultaneously push detailed abnormal condition data and screen information, eliminating the need for step-by-step investigation and accurately pinpointing the root cause of the problem, but also promptly handle abnormal conditions inside the furnace, reducing the probability of equipment damage and on-site safety accidents.

[0060] The technical solution of this application embodiment (technical solution of this application embodiment: S110-S120) firstly uses an air-cooled industrial camera installed at the observation port of the top-blown furnace to capture the smelting scene inside the top-blown furnace online, and sends the captured real-time video of the smelting inside the top-blown furnace to an industrial control computer. Then, the industrial control computer performs logarithmic transformation, dehazing, weighted fusion, and color restoration / enhancement preprocessing on the images in the captured real-time video of the smelting inside the top-blown furnace, and combines the trained abnormal working condition recognition model of the smelting inside the top-blown furnace to perform logarithmic transformation and dehazing. The image, after preprocessing including weighted fusion and color restoration / enhancement, is used for anomaly identification. If an abnormal condition is detected, an alarm is triggered and the control room personnel are instructed to make corresponding adjustments to the spray gun. In this way, the smelting scene inside the furnace can be observed in real time without picking up the gun and stopping the air supply. Moreover, the image defogging process can penetrate the dense smoke inside the furnace and clearly restore the position of the spray gun and the fluctuation of the molten pool surface. This solves the problem that existing images of smelting furnace scenes are blurred due to the dense smoke and high-temperature red light radiation inside the top-blown furnace, which makes it impossible to provide effective guidance to the workers responsible for smelting in the top-blown furnace.

[0061] Example 2 Figure 2 This invention provides a structural block diagram of an online monitoring device for top-blown furnace smelting based on an air-cooled vision system. This embodiment is applicable to online monitoring of top-blown furnace smelting scenarios. The online monitoring device for top-blown furnace smelting based on an air-cooled vision system can be implemented in hardware and / or software, and can be configured in an electronic device with data processing capabilities.

[0062] like Figure 2 As shown, the online monitoring device for top-blown furnace smelting based on an air-cooled vision system in this embodiment includes an air-cooled industrial camera 1, an air-cooling component 4, and an industrial control computer 7. The air-cooled industrial camera 1 is installed at the observation port 2 of the top-blown furnace and is used to capture images of the smelting scene inside the top-blown furnace 3 after blowing away the molten slag flying towards the camera's viewing window during the video recording process. The air-cooling component 4 is connected to the air inlet of the cooling channel 5 of the air-cooled industrial camera 1 and is used to provide cooling air to the cooling channel 5 of the air-cooled industrial camera 1. The air is then sprayed out through the air curtain nozzle 6, which is connected to the cooling channel 5 and located at the viewing window of the air-cooled industrial camera 1, to blow away the molten slag flying towards the camera's viewing window. The industrial control computer 7 is communicatively connected to the air-cooled industrial camera 1 and is used to receive the video of the smelting scene inside the top-blown furnace 3 and identify whether there are any abnormal conditions in the images of the smelting scene video inside the top-blown furnace 3. If there are any abnormal conditions, an alarm is triggered.

[0063] In practical applications, when the air-cooled industrial camera 1 captures video of the smelting scene inside the top-blown furnace, the air-cooling component 4 is first controlled to input cooling air into the cooling channel 5 of the air-cooled industrial camera 1. The cooling air entering the cooling channel 5 cools the air-cooled industrial camera 1 along the cooling channel 5 to prevent damage from the high temperature radiated by the top-blown furnace. On the other hand, it flows along the cooling channel 5 to the air curtain nozzle 6 connected to the cooling channel 5 and located at the shooting window of the air-cooled industrial camera 1, blowing away the molten slag flying towards the shooting window of the air-cooled industrial camera 1. Then, the air-cooled industrial camera 1 is controlled to capture the smelting scene inside the top-blown furnace 3. The captured video of the top-blown furnace 3 smelting is transmitted to the industrial control computer 7. The industrial control computer 7 then identifies and detects each image in the top-blown furnace 3 smelting video. If the top-blown furnace 3 is detected, the system will detect the slag. If the smelting conditions of Furnace 3 are abnormal, an alarm will be triggered and abnormal condition information will be pushed to the control room personnel. The height and airflow of the spray guns in the top-blown furnace will then be adjusted promptly. In this way, before the smelting scene inside the top-blown furnace is captured by the air-cooled industrial camera 1, the air curtain nozzles 6 at the camera's viewing window will blow away the molten slag flying towards the camera's viewing window. This not only effectively prevents molten slag and dust from adhering to the lens of the air-cooled industrial camera 1, achieving maintenance-free operation, but also improves the clarity of the video captured by the air-cooled industrial camera 1. Furthermore, it allows for timely push notifications of abnormal smelting conditions in the top-blown furnace to the control room personnel for adjustment of the spray gun height and airflow. This addresses the problem that existing images of the smelting scene captured by the smelting furnace are blurry and cannot provide effective guidance to the workers responsible for smelting in the top-blown furnace.

[0064] Please see Figure 3 The air-cooled industrial camera 1, used to capture images of the smelting scene in the top-blown furnace 3, includes a cooling outer sleeve 8, an air-cooled inner sleeve 10, and an industrial camera 9. The cooling outer sleeve 8 and the air-cooled inner sleeve 10 are assembled together to define a cooling channel 5 for cooling the air-cooled camera 9. The air-cooled industrial camera 1 is placed in the air-cooled inner sleeve 10. The air-cooled industrial camera 1 is used to capture images of the smelting scene in the top-blown furnace 3.

[0065] Specifically, the cooling outer tube 8 is a cavity structure with closed ends. The cooling outer tube 8 is connected to the observation port 2 of the top-blown furnace. The cooling inner sleeve 10 is connected inside the cooling outer tube 8. An industrial camera 9 is fixedly installed inside the air-cooled inner sleeve 10. The shooting end of the industrial camera 9 is on the same axis as the open end of the air-cooled inner sleeve 10 and the transparent window 11 connected to the cooling outer tube 8. A cooling channel 5 is defined between the air-cooled inner sleeve 10 and the cooling outer tube 8. The air inlet end of the cooling channel 5 is connected to the air outlet end of the air-cooling component 4. The cooling channel 5 is connected to multiple air curtain nozzles 6 embedded in the transparent window 11 connected to the cooling outer tube 8. The multiple air curtain nozzles 6 are arranged in a ring along the transparent window 11. The extension line of the shooting end of the industrial camera 9 is located at the midpoint of the ring.

[0066] In practical applications, as multiple air curtain nozzles 6 blow away the molten slag flying towards the viewing window of the air-cooled industrial camera 1, the air-cooling component 4 first introduces cooling air into the cooling channel 5. The cooling air then flows at high speed within the cooling channel 5, carrying away the heat generated by the industrial camera 9 and the heat radiated from the outside of the top-blown furnace. On the other hand, it flows along the cooling channel 5 to the air curtain nozzles 6 at the viewing window of the air-cooled industrial camera 1 and is ejected to form an air wall. This air wall forms a high-pressure protection zone in front of the viewing window of the air-cooled industrial camera 1. Molten slag splashes and dust particles near the viewing window of the air-cooled industrial camera 1 are instantly blown away by the airflow, thus ensuring the long-term cleanliness of the viewing window of the air-cooled industrial camera 1.

[0067] Please see Figure 4 The air-cooling assembly 4, used to supply cool air to the cooling channel 5 of the air-cooled industrial camera 1, includes a pressure regulating air pump 12, an air pipe 13, an air valve 14, and an air pressure detector 15. The pressure regulating air pump 12 is used to generate cooling air; the air pipe 13 is used to transport the cooling air generated by the pressure regulating air pump 12 into the cooling channel 5 of the air-cooled industrial camera 1; the air valve 14 is used to control the opening and closing of the air pipe 13; and the air pressure detector 15 is used to monitor the pressure of the cooling air in the air pipe 13.

[0068] Specifically, the air pressure detector 15 is electrically connected to the air valve 14 via the controller. Both the air valve 14 and the air pressure detector 15 are mounted on the air pipe 13. The first end of the air pipe 13 is connected to the outlet of the pressure regulating air pump 12, and the second end of the air pipe 13 is connected to the air inlet 16 on the cooling outer sleeve 8. The air inlet 16 is interconnected with the cooling channel 5 defined between the air-cooled inner sleeve 10 and the cooling outer sleeve 8.

[0069] In practical applications, when the air-cooling component 4 supplies cooling air to the cooling channel 5 of the air-cooled industrial camera 1, the pressure regulating pump 12 can be put into operation. The gas generated by the pressure regulating pump 12 is first transported along the air pipe 13 to the cooling channel 5 of the air-cooled industrial camera 1, and then along the cooling channel 5 to the air curtain nozzle 6 at the shooting window of the air-cooled industrial camera 1, blowing away the molten slag flying towards the shooting window of the air-cooled industrial camera 1, thus completing the process of supplying cooling air to the cooling channel 5 of the air-cooled industrial camera 1 using the air-cooling component 4. During this process, if the air pressure detector 15 detects that the pressure on the air pipe is lower than 0.4MPa, the power supply to the camera can be cut off to prevent the air-cooled industrial camera 1 from being damaged due to insufficient cooling.

[0070] It should be noted that the pressure regulating air pump 12, air valve 14 and air pressure detector 15 are all existing technologies, and no improvements or designs have been made to them. Therefore, their structures will not be described in detail.

[0071] The online monitoring device for top-blown furnace smelting based on an air-cooled vision system provided in this embodiment of the invention can execute the online monitoring method for top-blown furnace smelting based on an air-cooled vision system provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method.

[0072] Example 3 Figure 5 A schematic diagram of an electronic device 10 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0073] like Figure 4As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 may also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0074] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0075] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, central processing unit (CPU), graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as the online monitoring method for top-blown furnace melting based on an air-cooled vision system.

[0076] In some embodiments, the online monitoring method for top-blown furnace melting based on an air-cooled vision system can be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the online monitoring method for top-blown furnace melting based on an air-cooled vision system described above can be performed. Alternatively, in other embodiments, processor 11 can be configured to perform the online monitoring method for top-blown furnace melting based on an air-cooled vision system by any other suitable means (e.g., by means of firmware).

[0077] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0078] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0079] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0080] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0081] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0082] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0083] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0084] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for online monitoring of top-blown furnace smelting based on an air-cooled vision system, characterized in that: Includes the following steps: An air-cooled industrial camera installed at the observation port of the top-blown furnace is used to capture the smelting scene inside the furnace online, and the captured real-time video of the smelting inside the furnace is sent to the industrial control computer. The industrial control computer preprocesses images from real-time video of smelting inside the top-blown furnace and combines them with a trained abnormal smelting condition recognition model to identify anomalies in the preprocessed images. If an abnormal condition is detected, an alarm is triggered, and personnel in the central control room are instructed to make corresponding adjustments to the spray guns. The preprocessing includes sequentially performing logarithmic transformation, defogging, weighted fusion, and color restoration / enhancement preprocessing on the images. The abnormal smelting condition recognition model in the top-blown furnace is trained based on images of typical abnormal smelting conditions inside the top-blown furnace. The abnormal smelting condition identification model in the top-blown furnace is based on a convolutional neural network.

2. The method for online monitoring of top-blown furnace smelting based on an air-cooled vision system according to claim 1, characterized in that: The process of using an air-cooled industrial camera installed at the observation port of the top-blown furnace to capture online footage of the smelting process inside the furnace and sending the captured real-time video to an industrial control computer includes the following steps: The air-cooled gas assembly is controlled to deliver cooling air to the cooling air duct inside the air-cooled industrial camera located at the observation port of the top-blown furnace. The cooling air is then blown out through the dustproof air curtain nozzles facing the shooting end of the air-cooled industrial camera, blowing away the molten slag that flies towards the shooting window of the air-cooled industrial camera. While the dustproof air curtain nozzles blow away the molten slag flying towards the viewing window of the air-cooled industrial camera in real time, the air-cooled industrial camera is controlled to shoot the smelting scene inside the top-blown furnace online, and the real-time video of the smelting inside the top-blown furnace is transmitted to the remote industrial control computer.

3. The method for online monitoring of top-blown furnace smelting based on an air-cooled vision system according to claim 1, characterized in that: The industrial control computer includes an image compression module, an image dehazing and enhancement module, a color restoration module, an abnormal smelting condition recognition module in the top-blown furnace, and an abnormal condition alarm module. The industrial control computer preprocesses images from real-time video of smelting inside the top-blown furnace and uses a trained abnormal smelting condition recognition model to identify anomalies in the preprocessed images. If an abnormal condition is detected, an alarm is triggered, and personnel in the central control room are instructed to make corresponding adjustments to the spray guns, including the following steps: The image compression module performs logarithmic transformation compression on any image in the real-time video of melting inside the top-blown furnace to obtain the compressed image. The compressed image is processed by an image dehazing enhancement module to remove smoke and dust interference, and then the image with smoke and dust interference removed is weighted and fused with the image processed by the multi-scale Retinex algorithm to obtain the fused image. The color restoration module corrects the color cast in the fused image to obtain the corrected image. The abnormal working conditions of the smelting abnormal working conditions in the top blown furnace are detected and identified by the corrected image through the abnormal working condition identification module. If the smelting conditions inside the top-blown furnace are abnormal, an alarm will be triggered through the abnormal conditions alarm module, and the smelting conditions information inside the top-blown furnace will be pushed to the personnel in the central control room to guide them to make corresponding adjustments to the spray guns. The corresponding adjustments to the spray guns include adjusting the height of the spray guns or adjusting the air volume of the spray guns.

4. The method for online monitoring of top-blown furnace smelting based on an air-cooled vision system according to claim 3, characterized in that: The industrial control computer also includes: The image storage and query module is used to store the fused images processed by the image dehazing and enhancement module, and to display previous fused images in response to image query requests; among them, the previous fused images are high-definition videos after dehazing within the last 30 days.

5. The method for online monitoring of top-blown furnace smelting based on an air-cooled vision system according to claim 3, characterized in that: The image dehazing enhancement module is specifically used for: Based on the improved dark channel prior algorithm, smoke and dust interference removal is performed on the compressed image to obtain the processed image; the improved dark channel prior algorithm is constructed by combining the modified atmospheric light value selection strategy and the introduction of red light channel suppression factor to correct the transmittance calculation. Based on the guided filtering algorithm, the processed image is optimized to obtain the optimized image, that is, the image after removing smoke and dust interference; Multi-scale feature image extraction is performed on the compressed image based on the multi-scale Retinex algorithm. The image extracted by the multi-scale Retinex algorithm is then weighted and fused with the image after removing smoke and dust interference to obtain the fused image.

6. The method for online monitoring of top-blown furnace smelting based on an air-cooled vision system according to claim 3, characterized in that: The abnormal smelting condition identification module in the top-blown furnace is specifically used for: A model for identifying abnormal smelting conditions in a top-blown furnace is constructed based on a convolutional neural network. A dataset of smelting images inside the top-blown furnace is obtained. This dataset is then input into the top-blown furnace smelting abnormal condition recognition model for training, resulting in a trained model. The top-blown furnace smelting image dataset includes images of normal splashing inside the top-blown furnace, images of excessive foam slag inside the top-blown furnace, and images of nozzle formation on the top-blown furnace lance.

7. The method for online monitoring of top-blown furnace smelting based on an air-cooled vision system according to claim 3, characterized in that: The color restoration module is specifically used for: The color cast in the fused image is corrected by a white balance algorithm to obtain a corrected image. The color cast in the image is reddish due to high temperature radiation. The corrected image shows molten slag, molten metal and refractory bricks in a pseudo-color or grayscale level that is easy to identify.

8. An online monitoring device for top-blown furnace smelting based on an air-cooled vision system, characterized in that: include: An air-cooled industrial camera (1) is set at the observation port (2) of the top-blown furnace. It is used to film the slag flying towards the shooting window of the air-cooled industrial camera (1) after it is blown away during the process of filming the smelting scene inside the top-blown furnace (3). The air-cooled component (4) is connected to the air intake end of the cooling channel (5) of the air-cooled industrial camera (1) and is used to provide cooling air to the cooling channel (5) of the air-cooled industrial camera (1). The air is then sprayed out through the air curtain nozzle (6) connected to the cooling channel (5) and located at the shooting window of the air-cooled industrial camera (1) to blow away the molten slag flying towards the shooting window of the air-cooled industrial camera. The industrial control computer (7) is connected to the air-cooled industrial camera (1) to receive the smelting scene video inside the top-blown furnace (3) and identify whether there are any abnormal working conditions in the images of the smelting scene video inside the top-blown furnace (3); if there are any abnormal working conditions, an alarm will be triggered.

9. The online monitoring device for top-blown furnace smelting based on an air-cooled vision system according to claim 8, characterized in that: The air-cooled industrial camera (1) includes: The cooling outer sleeve (8) is a cavity structure with closed ends. The cooling outer sleeve (8) is connected to the observation port (2) of the top blown furnace. The air-cooled inner sleeve (10) is connected to the air-cooled outer sleeve (8). An industrial camera (9) is fixedly installed inside the air-cooled inner sleeve (10). The shooting end of the industrial camera (9) is on the same axis as the open end of the air-cooled inner sleeve (10) and the transparent window (11) connected to the cooling outer sleeve (8). A cooling channel (5) is defined between the air-cooled inner sleeve (10) and the cooling outer sleeve (8). The air inlet end of the cooling channel (5) is connected to the air outlet end of the air-cooling component (4). The cooling channel (5) is connected to multiple air curtain nozzles (6) embedded in the transparent window (11) connected to the cooling outer sleeve (8). The multiple air curtain nozzles (6) are arranged in a ring along the transparent window (11). The extension line of the shooting end of the industrial camera (9) is located at the midpoint of the ring.

10. The online monitoring device for top-blown furnace smelting based on an air-cooled vision system according to claim 9, characterized in that: The air-cooled component (4) includes: The air pressure detector (15) and the air valve (14) are electrically connected via a controller; The trachea (13), air valve (14) and air pressure detector (15) are all installed on the trachea (13); The first end of the pressure regulating air pump (12) and the air pipe (13) are connected to the air outlet of the pressure regulating air pump (12), and the second end of the air pipe (13) is connected to the air inlet (16) on the cooling outer sleeve (8). The air inlet (16) is connected to the cooling channel (5) defined between the air-cooled inner sleeve (10) and the cooling outer sleeve (8).

Images