A method and apparatus for quality monitoring and inspection of laser cutting of zirconium alloy plates

Abstract

This invention discloses a method and apparatus for quality monitoring and inspection of laser-cut zirconium alloy sheets, belonging to the field of laser cutting. A linear array CMOS camera is mounted on the side and below the laser-cut section. After the zirconium alloy sheet is laser-cut, the linear array CMOS camera acquires image information of the cut surface and the bottom surface of the sheet. Through a computer image correction and image defect recognition and classification program, the side-image section is not corrected, while the bottom surface image is corrected and merged into a single image. This image is then processed for grayscale conversion and mesh division, enabling rapid identification, classification, marking, early warning, judgment, and evaluation of defects. This invention uses a single linear array CMOS camera to merge the cut surface and the bottom surface of the sheet into a single image, simultaneously monitoring and inspecting three types of defects online. Compared to existing quality inspection methods for laser-cut zirconium alloy sheets, this invention develops a highly integrated online monitoring and inspection method and apparatus with lower hardware and software costs, improving the reliability of inspection and product quality.

Description

Technical Field

[0001] This invention belongs to the field of laser cutting technology, and more specifically, relates to a method and apparatus for quality monitoring and inspection of laser cutting of zirconium alloy plates. Background Technology

[0002] The working principle of laser cutting is to guide and focus a laser beam generated by a laser onto the surface of a material, causing it to melt and vaporize. Simultaneously, compressed gas coaxial with the laser beam blows away the molten material, and the laser beam and material move relative to each other along a specific trajectory, thus forming a cut of a certain shape. Laser cutting technology can be used in the processing of both metallic and non-metallic materials, significantly reducing processing time, lowering processing costs, and improving workpiece quality. It is widely used in the cutting of sheet metal in production. Zirconium alloys have good plasticity and can be made into pipes, sheets, etc., primarily used in the petroleum and nuclear technology fields.

[0003] Currently, the laser cutting process involves a complex kerf environment. Inappropriate selection of laser cutting parameters such as laser power, cutting speed, gas flow rate, and defocusing amount can easily lead to numerous defects, such as dross, spatter, and excessive surface roughness. These defects result in substandard key quality indicators of the laser-cut products, severely impacting the performance of the final assembled product and potentially causing serious safety accidents. However, traditional laser cutting quality inspection relies on manual checks, which are prone to misjudgments and omissions due to worker fatigue, consuming significant manpower and requiring improved reliability of results. Furthermore, traditional laser cutting lacks detection devices for defects such as dross, spatter, and excessive surface roughness. It cannot quickly evaluate products produced from laser-cut zirconium alloy plates, requiring manual inspection after cooling to ensure compliance with standards. This time-consuming evaluation of laser-cut product quality necessitates the development of a highly integrated, precise, and efficient online monitoring and inspection method and device for laser cutting quality. Summary of the Invention

[0004] In view of the shortcomings of the existing technology, the purpose of this invention is to provide a method and device for quality monitoring and inspection of laser-cut zirconium alloy plates, aiming to solve the problem of how to save a lot of manpower and improve the reliability of laser-cut product inspection.

[0005] To achieve the above objectives, in a first aspect, the present invention provides a method for quality monitoring and inspection of laser cutting of zirconium alloy plates, the method comprising:

[0006] S1. Obtain an image of the laser-cut zirconium alloy sheet containing the cut surface and the bottom surface. After correcting the bottom surface of the sheet, merge it with the uncorrected cut surface image into a single image, and then perform grayscale conversion and mesh division on the image.

[0007] S2. Calculate the ratio of the standard deviation of the grayscale histogram distribution to the root mean square height of the distribution for each grid, and compare it with the threshold corresponding to the maximum acceptable roughness. If it exceeds the threshold, the grid has a surface roughness exceeding the standard defect; otherwise, the grid does not have a surface roughness exceeding the standard defect.

[0008] S3. Input the grayscale image of each grid into the convolutional neural network to obtain the probability of slag-hanging defects and splashing defects in each grid, and label the defects according to the probability.

[0009] S4. Evaluate the cutting quality of the sheet metal by considering the quantity and area of ​​three types of defects: excessive surface roughness, slag, and splashing.

[0010] Preferably, in step S1, bottom surface correction of the image includes:

[0011] 1) Based on perspective transformation, identify the intersection line between the bottom surface and the cutting surface and extend it to the edge of the image;

[0012] 2) Perform perspective transformation correction on the image below the resulting extension line;

[0013] 3) Merge the corrected image below along the extended line with the cutting surface to obtain the final corrected image.

[0014] Preferably, the horizontal distance between the shooting device and the cutting surface is obtained, and the acute angle between the vertical plane of the shooting device axis and the bottom surface of the plate is obtained based on trigonometric function relationships, which is used as the angle for bottom surface correction.

[0015] Preferably, before inputting the grayscale images of each grid into the convolutional neural network, a selective search algorithm is used to initially screen the suspicious grid images.

[0016] Preferably, in step S3, for the obtained defect judgment results, if the maximum probability exceeds 95%, the defect is identified as existing and a red box is directly added to the area; if the maximum probability is between 75% and 95%, the defect is considered to be suspected and a yellow box is added to the area; if the maximum probability is below 75%, no mark is added. This makes it easy to distinguish when the results are finally displayed on the computer interface through illustrations.

[0017] Preferably, the method further includes any of the following processing methods:

[0018] Method 1: A three-color indicator light and a buzzer warning sound are used to provide early warnings via a pop-up window on the visual display interface and the linear CMOS monitoring system. A red warning light indicates a serious defect, with the number or size of defects exceeding the normal range, accompanied by a continuous buzzer warning. A yellow warning light indicates that the quality of the cut surface and bottom surface is within the normal range, but the score is not excellent, accompanied by an intermittent buzzer warning. A green warning light indicates that the defect is normal and the score is excellent, with no warning sound.

[0019] Method 2: Count the number, size, and density of the detected defects, and compare them with the standard to determine whether the standard is met for each type of defect.

[0020] To achieve the above objectives, in a second aspect, the present invention provides a quality monitoring and inspection device for laser cutting of zirconium alloy plates, the device comprising: an imaging device, a guide rail, a support and connection device, and an image processing and analysis module;

[0021] The guide rail is located below the laser-cut surface and is parallel to the x-axis of the laser cutting. It is used to drive the imaging device to move smoothly and without obstruction, while ensuring that the imaging device can scan the entire surface of the laser-cut workpiece at the same time.

[0022] The shooting device is fixed on the guide rail by a support connection device. It is used to take pictures containing the bottom surface of the cut plate and the cut surface after the zirconium alloy plate is laser cut according to the control signal, and upload them to the image processing and analysis module.

[0023] The image processing and analysis module is used to process and analyze the captured images using the method described in the first aspect to obtain the defect location and type of the zirconium alloy plate and the cutting quality.

[0024] Preferably, the elevation angle of the shooting device is 45° to 65°.

[0025] Preferably, the device further includes a protective sheet;

[0026] The protective plate is installed on the lens surface of the imaging device. It is used to close during the laser cutting process according to the control signal to isolate the lens of the imaging device from the outside world, and to open during the scanning and collection process of the imaging device to ensure image acquisition.

[0027] Preferably, the device further includes: a miniature ventilation device;

[0028] The miniature ventilation device is installed around the lens of the imaging device and is used to blow out gas during the scanning and collection process of the imaging device according to the control signal, so as to remove residual smoke and dust within the field of view of the imaging device.

[0029] In summary, the technical solutions conceived by this invention have the following beneficial effects compared with the prior art:

[0030] (1) This invention proposes a method for quality monitoring and inspection of laser-cut zirconium alloy plates. It acquires and analyzes images of both the bottom and cut surfaces of the plate in a single photograph, reducing the number of photographs and analyses required and thus the time spent on quality monitoring and inspection. The acquired images are corrected using a computer correction program, focusing only on the bottom surface while maintaining the cut surface as a side view. The two parts are then merged into a final corrected image. The image is then divided and selected into regions, and various defects are classified and judged using a convolutional neural network. The defect information is displayed on the computer interface to assess the cutting quality of the laser-cut zirconium alloy plate. This method allows for the rapid identification of defect locations and types in the laser-cut zirconium alloy plate, and the defect information is fed back to the laser cutting system to adjust laser cutting parameters and monitor the quality of the laser-cut zirconium plate.

[0031] (2) This invention proposes a quality monitoring and inspection device for laser cutting of zirconium alloy plates. Based on the existing laser cutting quality, a linear array CMOS camera and its motion assistance and protection device are added to the lower side of the laser cutting seam. With the help of computer image processing technology, the device can monitor and inspect three types of defects on two sides through a single photograph. Furthermore, it does not require major modifications to the overall processing equipment. Therefore, it saves a lot of manpower and improves the reliability of the inspection, thus saving a lot of costs in the long run. Attached Figure Description

[0032] Figure 1 The present invention provides a flowchart of a method for quality monitoring and inspection of laser cutting of zirconium alloy plates.

[0033] Figure 2 This is a schematic diagram of camera mounting provided for an embodiment of the present invention.

[0034] Figure 3 This is a schematic diagram of bottom surface correction provided for an embodiment of the present invention.

[0035] Figure 4 This is a schematic diagram of the defect marking results provided in an embodiment of the present invention. Detailed Implementation

[0036] To make the objectives, technical solutions, and advantages of this invention clearer, the 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 and not intended to limit the invention.

[0037] Figure 1 This invention provides a flowchart for a quality monitoring and inspection method for laser cutting of zirconium alloy plates. (See flowchart for example.) Figure 1As shown, the method includes:

[0038] Step S1: Set up the camera position and angle.

[0039] Figure 2 This is a schematic diagram of camera mounting provided for an embodiment of the present invention. Figure 2 As shown, a linear CMOS camera is installed on the bottom side of the zirconium plate. The camera is then covered with a protective lens, preparing for laser cutting of the zirconium plate. The main viewing direction of the linear CMOS camera is perpendicular to the line of intersection between the cutting surface 1 and the bottom surface 2. The optimal elevation angle for the linear CMOS camera is between 45° and 65°. Within this angle range, the outlines of slag and surface roughness defects on the cutting surface are highlighted, while also covering a larger area of ​​the bottom surface of the plate, which is beneficial for the subsequent neural network model to accurately identify and classify defects.

[0040] A guide rail is mounted on the lower side of the laser-cut surface for moving the line-scan CMOS camera. The guide rail is installed parallel to the x-axis of the laser cutting. The line-scan CMOS camera is connected to the guide rail via a support device to ensure unobstructed, smooth movement of the camera on the guide rail without being damaged by the laser.

[0041] Step S2: Laser cutting of zirconium alloy sheet.

[0042] Set the laser cutting parameters, including cutting power, cutting speed, and air pressure. Clamp the zirconium plate and perform laser cutting.

[0043] Step S3: Linear CMOS image acquisition and reconstruction.

[0044] After the laser cutting is completed, the fume extraction system absorbs most of the smoke and dust generated during the cutting process. Then, the protective cover of the camera is opened, and the ventilation system is turned on to disperse the residual smoke and dust in the field of view of the linear CMOS camera. The linear CMOS camera is then turned on, and it simultaneously acquires optical images of the cutting surface and the bottom surface of the workpiece along the pre-installed guide rail.

[0045] The protective sheet of the line array CMOS camera can isolate the lens of the imaging device from the outside world through the kit during the laser cutting process. When the line array CMOS camera scans and collects information, the protective sheet is opened by a motor pulling a soft rope, without interfering with the acquisition of image information by the imaging device. The protective sheet is made of glass or rubber and is rectangular in shape.

[0046] A small ventilation and smoke removal system is installed around the linear CMOS camera to remove residual smoke and dust within the camera's field of view during image acquisition, ensuring that the captured image information is not interfered with by residual smoke and dust.

[0047] A linear CMOS camera moves along the cutting trajectory from the starting point of the material cutting process to the cutting end via a servo motor on a guide rail, reconstructing the image information of the entire cutting surface and the bottom surface of the material. The reconstructed cutting surface and bottom surface are not front views of the cutting surface and bottom surface; the presented images are formed by the connection of two trapezoids.

[0048] Step S4: Correction of bottom surface image and fusion of cross-sectional image.

[0049] Since the camera is placed below the cut, the captured images are both side views with angles on the bottom and the cut surface. The images need to be corrected. The cut surface is not corrected, but the bottom surface of the board is corrected. The two parts are then merged to obtain the corrected image.

[0050] S4.1: Image Correction of the Bottom Surface of the Board: The correction process adopts a perspective transformation based on the improved geometric intersection of the bottom surface and the cutting surface. Perspective transformation involves rotating the original projection plane around the trace line by a certain angle to obtain the corrected front view. This method is based on perspective transformation. First, the intersection of the bottom surface and the cutting surface is identified and extended to the edge of the image. Perspective transformation correction is performed on the image below the obtained extension line. Since perspective transformation extends and scales the image, the stretched and enlarged parts will cause pixels to change from continuous to discontinuous. Here, bilinear interpolation is used, and the gray values ​​of the four adjacent pixels of the pixel to be obtained are linearly interpolated in two directions to obtain the final corrected image. Since the detection of the cutting surface requires attention to the roughness value, correction can easily affect the roughness detection. Moreover, the correction of the cutting surface has little impact on the detection of slag, so it is possible to choose not to correct the cutting surface.

[0051] In the specific correction calculation, take any point (x, y, 1) on the image before correction, and its corresponding point (x′, y′, 1) on the image after correction. Let the transformation matrix be H, then the transformation relationship is as follows:

[0052]

[0053] Because the width of the bottom surface of the zirconium plate being cut is much greater than its thickness, and the bottom surface occupies most of the inspection area, the camera needs to be positioned more towards the bottom surface for imaging. After the guide rail is installed, the height of the camera from the bottom surface of the workpiece can be determined. The guide rail is parallel to the x-axis of the laser cutting. After the machining program is loaded, the horizontal distance between the camera and the cutting surface can be obtained. Figure 3 This is a schematic diagram of bottom surface correction provided for an embodiment of the present invention. Figure 3 As shown, based on trigonometric relationships, the angle between the camera axis and the intersection of the cutting surface and the bottom surface can be obtained. This angle is the bottom surface correction angle θ.

[0054] S4.2: Image fusion of the bottom surface and the cut surface of the board: The corrected bottom surface image and the image directly captured by the line scan CMOS camera are placed in one image. The bottom surface and the cut surface of the board are distinguished by their boundary line. The corrected bottom surface image is rectangular, but the uncorrected cut surface image is trapezoidal. The top edge of the rectangle coincides with the bottom edge of the trapezoid, and a dashed line is reconstructed on the coincident line for visual distinction.

[0055] Step S5: Defect identification, classification, early warning, marking, judgment and evaluation.

[0056] In the identification and classification of defects, based on the quality monitoring and inspection method and device for laser cutting of zirconium alloy plates, the image input to the computer is converted into a grayscale image through grayscale processing. Specifically, a weighted average method is used, but not limited to, to perform grayscale processing on the RGB three-color image to obtain a grayscale image with pixel values ​​in the range [0, 255]. The algorithm used is I(x, y) = a*I_R(x, y) + b*I_G(x, y) + c*I_B(x, y), where a + b + c = 1.

[0057] S5.1: Defect Identification and Classification: For surface roughness detection, the surface roughness of the cut surface is large-scale and continuous. The surface roughness of the workpiece is related to the standard deviation and root mean square (RMS) of the image gray-level histogram distribution. Specifically, the surface roughness increases with the increase of the ratio of the standard deviation to the RMS height of the image gray-level histogram. The ratio varies for each metal. Specifically, for normal laser-cut zirconium alloy surfaces, it is necessary to take pictures, analyze the gray-level histogram, obtain the standard deviation S and the RMS height H of the gray-level histogram distribution, and determine the threshold by calculating the ratio S / H. 100 images are selected for testing to determine the threshold. Among the test results of 100 images, the data with roughness within the normal range are selected as the corresponding maximum S / H value as the threshold. The corrected image is cut into small areas and the gray-level histogram is analyzed. Based on the threshold corresponding to the maximum acceptable roughness obtained through pre-testing, it is determined whether the surface roughness of the area exceeds the standard. The judgment criterion is that the calculated value is greater than the threshold, and the surface roughness exceeds the standard. In this embodiment, the standard for the surface roughness value of the zirconium alloy plate laser cutting is set to Ra 3.2.

[0058] For detecting slag and splashes, the image is first divided into multiple 84×84 pixel grids, and the regions are filtered to select all suspicious areas. A selective search algorithm is used here. The specific steps are as follows: First, the input image is divided by initialization with double thresholds to recommend target regions. The similarity between adjacent sub-regions in the image is calculated based on the texture features and overlap of adjacent regions. Finally, by continuously merging adjacent similar regions, a smaller number of target regions are obtained, thereby narrowing the target search range. Compared with the traditional sliding window exhaustive detection method, this method reduces redundant regions and improves detection efficiency.

[0059] The obtained suspicious areas are cropped into multiple 42×42 pixel images. The defects captured by the linear CMOS camera are at the 0.1mm level, and the 42×42 pixel images can include the defects. Using a convolutional neural network, the 42×42 pixel images of the cropped areas are input and processed through 4 to 6 layers of convolution, pooling, activation, etc., to extract the features of the defects. Flattening is used between the convolutional layers and the fully connected layers to achieve the transition between the convolutional and fully connected layers. The input is then fed into a 3-layer fully connected layer, and finally the Softmax activation function is used to output the probability value of the presence of various types of defects in the image.

[0060] The neural network is trained using images of the cut products. The input data consists of 1,000 42×42 pixel images with slag defects and 1,000 42×42 pixel images with splash defects. 80% of the data is used to train the CNN convolutional neural network, and the other 20% is used for prediction. When the prediction accuracy reaches 99% or higher, the trained neural network is considered reliable and can be used in actual production.

[0061] S5.2: Defect Warning: A three-color indicator light and a buzzer warning are used to provide warnings via a pop-up window on the visual display interface and the linear CMOS monitoring system. A red warning light indicates a serious defect, with the number or size of defects exceeding the normal range, accompanied by a continuous buzzer warning. A yellow warning light indicates that the quality of the cut surface and bottom surface is within the normal range, but the score is not excellent, accompanied by an intermittent buzzer warning. A green warning light indicates that the defect condition is normal and the score is excellent, with no warning sound.

[0062] S5.3: Defect Marking: The defect-marked areas are displayed on the computer interface, using red for the defect areas. For the obtained defect judgment results, areas with a maximum probability exceeding 95% are directly marked with a red box; areas with a maximum probability between 75% and 95% are marked with a yellow box; areas with a maximum probability below 75%, since only two types of defects are detected, have a maximum probability of no less than 50%. If the maximum probability value of one type of defect is significantly higher than that of another, it is considered not a defect and is not marked. This is for easy differentiation when finally displayed graphically on the computer interface. Figure 4 This is a schematic diagram of defect marking results provided in an embodiment of the present invention. Figure 4 As shown, small circles represent slag defects, ovals represent splashing defects, and squares represent roughness exceeding the standard.

[0063] S5.4: Defect Judgment: Compare with the standard to see if a single defect meets the standard. Perform quantity, size, and density statistics on the detected defects. The number of defects with slag should not exceed 4, and the density should not exceed 2mm. 2 / dm 2 The number of splashes must not exceed 10, and the density must not exceed 3mm. 2 / dm 2 The area with excessive surface roughness must not exceed 12mm. 2 .

[0064] S5.5: Defect Assessment: A comprehensive assessment of the board quality is conducted by combining measured dimensions, the area of ​​surface roughness exceeding standards, and quality indicators such as the size and quantity of slag and splashes. Based on key indicators such as surface roughness, slag, and splashes, a weighted average is used to obtain a comprehensive quality score. The aforementioned quality assessment standards for slag, splashes, and surface roughness are set at 60 points, with no defects defined as a full score of 100 points. Each quality indicator for a defect has an equal proportion in the defect score. The scores for each indicator are based on a standard value of 60 points and no defects as 100 points. The scores are evenly distributed across the quality data. The overall quality score is calculated by weighting the scores of the three types of defects in a 1:1:1 ratio, and the quality is divided into four levels: 90 points and above is Excellent, 75 points and above is Good, 60 points and above is Acceptable, and below 60 points is Unacceptable. Furthermore, if any data exceeds the standard value, it is classified as Unacceptable.

[0065] Furthermore, after evaluating the laser-cut products, any problems detected can be fed back to the laser cutting system, allowing for timely adjustments to the laser cutting parameters, thereby improving product quality and achieving the goal of monitoring product quality.

[0066] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for monitoring the quality of laser cutting of a zirconium alloy plate, characterized in that, The method comprises: S1. Obtain a picture containing a cutting surface and a bottom surface after laser cutting of a zirconium alloy plate, correct the bottom surface of the plate, and fuse the corrected bottom surface with the picture of the uncorrected cutting surface to form a picture, and then perform grayscale processing and grid division on the picture; S2. Calculate the ratio of the distribution standard deviation and the distribution root mean square height value of the gray histogram of each grid, and compare it with the threshold value corresponding to the maximum acceptable roughness, respectively. If it exceeds, the grid has a surface roughness defect, otherwise, the grid does not have a surface roughness defect; S3. Input the grayscale picture of each grid into a convolutional neural network to obtain the probability of each grid having slag hanging and splashing defects, and label the defects according to the probability size; S4. Evaluate the cutting quality of the plate by comprehensively considering the number and area of the three types of defects, i.e., surface roughness exceeding, slag hanging, and splashing; In step S1, the bottom surface correction of the picture comprises: 1) Based on perspective transformation, identify the intersection line of the bottom surface and the cutting surface and extend it to the edge of the picture; 2) Perform perspective transformation correction on the picture below the obtained extended line; 3) Fuse the corrected picture below along the extended line with the cutting surface to obtain the final corrected picture; Obtain the horizontal distance of the shooting device from the cutting surface, and according to the trigonometric relationship, obtain the acute angle between the vertical plane of the shooting device axis and the plate bottom surface as the angle of bottom surface correction.

2. The method of claim 1, wherein, Before inputting the grayscale picture of each grid into the convolutional neural network, a selective search algorithm is used for preliminary screening of suspicious grid pictures.

3. The method of claim 1, wherein, In step S3, for the obtained defect judgment result, if the maximum probability exceeds 95%, it is determined that the defect exists, and a red frame is added to the grid; if the maximum probability is between 75% and 95%, it is considered that the defect may exist, and a yellow frame is added to the grid; if the maximum probability is below 75%, no mark is added. Finally, in the computer interface display, it is convenient to distinguish.

4. The method of claim 1, wherein, The method further comprises any one of the following processing methods: Processing method one: three-color light pre-warning and buzzer warning sound pre-warning are performed on the visual display interface pop-up window and the linear array CMOS monitoring system; wherein, the red pre-warning light indicates that the defect problem is serious, the number or size of the defect exceeds the normal range, and the buzzer emits continuous warning sound; the yellow pre-warning light indicates that the quality of the cutting surface and the bottom surface is within the normal range, but the score does not reach excellent, and the buzzer emits intermittent warning sound; the green pre-warning light indicates that the defect condition is normal and the score is excellent, and there is no warning sound; Processing method two: the number, size, and defect density of the detected defects are counted and compared with the standard to determine whether each type of defect meets the standard.

5. A device for monitoring the quality of laser cutting of a zirconium alloy plate, characterized in that it comprises: The device comprises a shooting device, a guide rail, a support connection device, and an image processing and analysis module; The guide rail is located below the laser cutting surface and is parallel to the x-axis of the laser cutting, used to drive the shooting device to move smoothly without obstacles, and ensure that the shooting device can scan the overall appearance of the laser cutting workpiece cutting surface and bottom surface at the same time. The photographing device is fixed on the guide rail through the support connecting device, and is used for photographing pictures containing the bottom surface and the cutting surface of the cutting plate according to the control signal after the laser cutting of the zirconium alloy plate is completed, and uploading the pictures to the image processing analysis module; The image processing analysis module is used for processing and analyzing the photographed pictures by the method in any one of claims 1 to 4, to obtain the defect position and type of the zirconium alloy plate and the cutting quality.

6. The apparatus of claim 5, wherein, The elevation angle of the photographing device is 45°-65°.

7. The apparatus of claim 5, wherein, The device further comprises a protective sheet; The protective sheet is installed on the lens surface of the photographing device, and is used for closing to isolate the lens of the photographing device from the outside environment according to the control signal during the laser cutting process, and opening to ensure the collection of the image during the scanning collection of the photographing device.

8. The apparatus of claim 5, wherein, The device further comprises a micro ventilation device; The micro ventilation device is installed on the periphery of the lens of the photographing device, and is used for blowing out gas to remove the residual smoke and dust in the view angle of the photographing device according to the control signal during the scanning collection of the photographing device.

Images