Surface detection method for energy-saving and environment-friendly aluminum plate curtain wall based on machine vision

Abstract

The invention relates to the technical field of industrial detection, in particular to a surface detection method for an energy-saving and environment-friendly aluminum plate curtain wall based on machine vision. According to the method, the aluminum plate curtain wall is polished, shadow can be eliminated, misjudgment of surface detection of the aluminum plate curtain wall is reduced, meanwhile, image information of the aluminum plate curtain wall is obtained in real time, image preprocessing is conducted on the obtained image information, the defect boundary can be effectively reflected in real time, and therefore the defect area of the aluminum plate curtain wall is accurately determined; meanwhile, the detected aluminum plate curtain wall is divided into a plurality of subareas with the same area, the reason that the curtain wall does not meet the standard is determined according to the area ratio of all the subareas through a periodic detection result, and a corresponding processing mode is generated; according to the aluminum plate curtain wall surface detection method, the influence of preprocessing or irradiation light on aluminum plate curtain wall surface detection in the image acquisition process can be effectively avoided, so that the situation of erroneous judgment during aluminum plate curtain wall detection is effectively reduced, the accuracy of aluminum plate curtain wall detection is further improved, and the curtain wall detection efficiency is further improved.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial detection, and in particular to a surface detection method of an energy-saving and environment-friendly aluminum plate curtain wall based on machine vision. Background Art

[0002] Energy-saving and environmentally friendly aluminum curtain wall is a modern building exterior wall material. It combines the durability of aluminum panels with the characteristics of energy saving and environmental protection, becoming an important choice in modern building design. However, the detection and research of its surface corrosion is currently limited.

[0003] China Patent Publication No.: CN118332292B, discloses a surface detection method for energy-saving and environmentally friendly aluminum curtain wall based on machine vision, including: summarizing a set of steel types; summarizing a set of steel corrosion; forming a steel corrosion category; obtaining an alternative set; sorting and numbering the alternative detection schemes in the alternative set; forming a corrosion database and data analysis classification for the alternative detection schemes; identifying the material type and corrosion type of the steel to be detected; obtaining the alternative detection scheme with the smallest number in the alternative set corresponding to the characteristic steel corrosion category as the characteristic alternative detection scheme; using the corrosion database and data analysis classification corresponding to the characteristic alternative detection scheme to analyze the data to be analyzed; re-sorting and numbering the alternative detection schemes in the updated alternative set. By constructing an alternative set, matching the best detection scheme, and updating in the existing data, it is ensured that the corrosion detection of steel can be continuously optimized.

[0004] It can be seen that the above scheme has the following problems: the scheme cannot analyze the reasons why the surface inspection results do not meet the standards due to other reasons other than the material itself, and there is no periodic inspection of the curtain wall to make the detected reasons for non-compliance with the standards more accurate, thereby failing to effectively reduce the occurrence of misjudgments when inspecting aluminum curtain walls, and further failing to further improve the accuracy of aluminum curtain wall inspection, and failing to further improve the efficiency of curtain wall inspection. Summary of the invention

[0005] To this end, the present invention provides a surface detection method for an energy-saving and environmentally friendly aluminum plate curtain wall based on machine vision, so as to overcome the problem in the prior art that it is impossible to analyze the reasons why the surface detection results do not meet the standards other than the curtain wall itself, thereby failing to reduce the misjudgment of the surface not meeting the standards, and further failing to ensure the accuracy of the surface detection of the aluminum plate curtain wall.

[0006] To achieve the above object, the present invention provides a surface detection method for energy-saving and environmentally friendly aluminum curtain wall based on machine vision, comprising:

[0007] Light the aluminum curtain wall to eliminate shadows on its surface;

[0008] Acquiring image information of the surface of the aluminum curtain wall in real time;

[0009] Preprocessing the image information;

[0010] Based on the preprocessed image information, the image information is detected using a YOLOv5 model to obtain defect features in the image information;

[0011] Dividing the preprocessed image information into a plurality of partitions of the same area, and calculating the area ratio in each partition respectively, wherein the area ratio is the ratio of the area of ​​the defect feature in a single partition to the area of ​​the corresponding partition;

[0012] Based on the obtained area ratios, it is determined whether the aluminum curtain wall meets the standards, and when it is determined that the aluminum curtain wall does not meet the standards, a corresponding treatment method is generated based on the determined reasons for not meeting the standards, including issuing an anti-corrosion treatment notice and redetermining corresponding detection parameters.

[0013] Furthermore, the process of determining whether the aluminum plate curtain wall meets the standards based on the obtained area ratios includes: calculating the average value of each of the area ratios, and determining whether the aluminum plate curtain wall meets the standards based on a comparison result between the obtained average value and a pre-stored preset area ratio, wherein: if the average value is less than or equal to the first preset average value, the aluminum plate curtain wall is determined to meet the standards; if the average value is greater than the first preset average value and less than the second preset average value, whether the aluminum plate curtain wall meets the standards is determined based on the variance of each of the area ratios; if the average value is greater than or equal to the second preset average value, the aluminum plate curtain wall is determined to not meet the standards, and the reason why the aluminum plate curtain wall does not meet the standards is determined based on the difference between the average value and the second preset average value.

[0014] Furthermore, the process of determining whether the aluminum plate curtain wall meets the standard based on the variance of each of the area ratios includes: comparing the obtained variance with the preset variance; if the variance is less than or equal to the preset variance, determining that the acquisition process for the image information does not meet the standard, and re-determining the detection parameters in the image acquisition process based on the obtained variance value; if the variance is greater than the preset variance, determining that the curtain wall does not meet the standard, and determining the reason for non-compliance with the standard based on the difference between the average value and the preset second preset average value.

[0015] Furthermore, the process of redetermining the detection parameters in the image acquisition process based on the variance value includes: obtaining the difference between the variance and the preset variance; increasing the illumination light intensity based on the obtained difference, and the difference is proportional to the increase in the illumination light intensity.

[0016] Furthermore, the process of determining the reason why the aluminum plate curtain wall does not meet the standard based on the difference between the average value and the second preset average value includes: determining whether the aluminum plate curtain wall is corroded based on the comparison result of the difference between the average value and the second preset average value and the pre-stored preset difference, wherein: if the difference is less than or equal to the first preset difference, it is determined that the aluminum plate curtain wall is corroded; if the difference is greater than the first preset difference and less than the second preset difference, it is determined whether the aluminum plate curtain wall is corroded based on the average value of the clarity of the boundaries of each defect feature; if the difference is greater than or equal to the second preset difference, it is determined whether the aluminum plate curtain wall is corroded based on the secondary ratio; wherein: the secondary ratio is the ratio of the average value of the area ratio in the current test result to the average value of the area ratio in the previous test result.

[0017] Further, the process of determining whether the aluminum plate curtain wall is corroded based on the average clarity value of the boundaries of each defect feature includes: comparing the average clarity value with a preset average clarity value; if the average clarity value is less than or equal to the preset average clarity value, determining that the preprocessing process for the image information does not meet the standard, and re-determining the preprocessing parameters based on the difference between the average clarity value and the preset average clarity value; wherein: the detection parameters include the preprocessing parameters. If the average clarity value is greater than the preset average clarity value, determining that the aluminum plate curtain wall is corroded, and issuing an anti-corrosion treatment notice.

[0018] Furthermore, the process of redetermining the preprocessing parameters based on the difference between the clarity average value and the preset clarity average value includes: increasing the noise reduction factor based on the difference between the clarity and the preset clarity, and the difference is proportional to the noise reduction factor; wherein: the preprocessing parameters include (noise reduction factor, grayscale transformation parameters).

[0019] Furthermore, the process of determining whether the aluminum curtain wall is corroded based on the secondary ratio includes: comparing the secondary ratio with a preset secondary ratio; if the secondary ratio is less than or equal to the preset secondary ratio, it is determined that the aluminum curtain wall is corroded, and an anti-corrosion treatment notice is issued; if the secondary ratio is greater than the preset secondary ratio, it is determined that there is a problem with the illumination angle, and the illumination angle is re-determined based on the difference between the secondary ratio and the preset secondary ratio; wherein: the illumination angle is the angle between the line connecting the illumination point and the aluminum curtain wall, and the line connecting the image acquisition point and the aluminum curtain wall; the detection parameters include the illumination angle.

[0020] Furthermore, the process of redetermining the illumination angle based on the difference between the secondary ratio and the preset secondary ratio includes: reducing the illumination angle based on the difference, and the difference is inversely proportional to the reduction degree of the illumination angle.

[0021] Further, the process of re-determining the illumination light intensity based on the illumination angle includes: reducing the illumination light intensity based on the illumination angle, and the angle is inversely proportional to the reduction degree of the illumination light intensity; wherein: the detection parameter includes the illumination light intensity.

[0022] Compared with the prior art, the beneficial effect of the present invention lies in that, by lighting the aluminum plate curtain wall, the present invention can eliminate shadows and reduce misjudgment of its surface detection, and at the same time, by real-time acquisition of image information of the aluminum plate curtain wall and image preprocessing of the acquired image information, the defect boundary can be effectively reflected in real time, thereby accurately determining the defect area of ​​the aluminum plate curtain wall; at the same time, by dividing the detected aluminum plate curtain wall into a number of partitions with the same area, and determining the reason why the curtain wall does not meet the standard through the result of periodic detection based on the area ratio of each partition and generating a corresponding processing method, the influence of preprocessing or irradiation light on the surface detection of the aluminum plate curtain wall during image acquisition can be effectively avoided, thereby effectively reducing the misjudgment when detecting the aluminum plate curtain wall, thereby further improving the accuracy of aluminum plate curtain wall detection, and further improving the efficiency of curtain wall detection.

[0023] Furthermore, the present invention determines whether the surface of the aluminum curtain wall meets the standards through the average value of each area ratio, and preliminarily determines the reason for not meeting the standards. This can more quickly determine whether the surface of the aluminum curtain wall meets the standards, thereby further improving the efficiency of determining whether the surface of the aluminum curtain wall is standard.

[0024] Furthermore, the present invention generates a corresponding processing method by comparing the variance of each area ratio with a preset variance. Based on the fluctuation of the area ratio at multiple moments, it can more accurately determine the reason why the surface of the aluminum curtain wall does not meet the standards and make adjustments, thereby further improving the accuracy of surface detection of the aluminum curtain wall.

[0025] Furthermore, the present invention adjusts the brightness of the irradiated light by the difference between the variance of each area ratio and a preset variance, thereby further avoiding the problem of inaccurate surface detection of the aluminum curtain wall due to the brightness of the irradiated light, and further improving the detection accuracy of the aluminum curtain wall surface.

[0026] Furthermore, the present invention determines whether there is corrosion on the aluminum curtain wall by the difference between the actual area ratio and the second preset average value, and preliminarily analyzes the cause of the corrosion. It can more quickly analyze whether there is corrosion on the aluminum curtain wall, thereby further improving the detection efficiency of the aluminum curtain wall surface.

[0027] Furthermore, the present invention analyzes the comparison results of the average clarity of the boundary of the corrosion position at the same time with the preset clarity, and can more accurately determine whether the corrosion is detected due to improper setting of the preprocessing parameters of the image information, thereby further improving the detection accuracy of the aluminum plate curtain wall surface.

[0028] Furthermore, the present invention adjusts the noise reduction ratio based on the difference between the clarity and the preset clarity, so that the noise reduction ratio can be adjusted more accurately, and image preprocessing can be performed more accurately, thereby further improving the detection accuracy of the aluminum curtain wall surface.

[0029] Furthermore, the present invention analyzes the comparison result based on the ratio of the area ratio in the detection result at the current moment to the area ratio in the detection result at the previous moment and the preset ratio, so as to more accurately judge whether the aluminum curtain wall is corroded during detection due to the problem of light irradiation angle, thereby further improving the detection accuracy of the aluminum curtain wall surface.

[0030] Furthermore, the present invention adjusts the angle based on the ratio of the area ratio in the detection result at the current moment to the area ratio in the detection result at the previous moment and a preset ratio, which can more accurately adjust the angle, thereby further avoiding the occurrence of corrosion detection due to inappropriate angle size, thereby further improving the detection accuracy of the aluminum plate curtain wall surface.

[0031] Furthermore, the present invention adjusts the intensity of the irradiated light based on the angle, which can avoid the influence of reflection during direct irradiation and save resources, thereby further improving the detection accuracy of the aluminum curtain wall surface. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a module schematic diagram of a surface detection method of an energy-saving and environmentally friendly aluminum plate curtain wall based on machine vision according to an embodiment of the present invention;

[0033] Figure 2 This is a flow chart of the steps of a surface detection method for energy-saving and environmentally friendly aluminum curtain wall based on machine vision according to an embodiment of the present invention;

[0034] Figure 3 A flowchart of the steps of determining the comparison result based on the average value of several area ratios at the same time and the preset area ratio in an embodiment of the present invention;

[0035] Figure 4 A flowchart of the steps of performing determination based on the variance of the area ratio at the same time and the preset variance according to an embodiment of the present invention;

[0036] Figure 5 This is a flow chart of the steps of determining based on the difference between the actual area ratio and the second preset average value and the preset difference in an embodiment of the present invention. DETAILED DESCRIPTION

[0037] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0038] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0039] See also Figure 1 As shown, it is a module schematic diagram of a surface detection system for energy-saving and environmentally friendly aluminum plate curtain wall based on machine vision according to an embodiment of the present invention;

[0040] Lighting units are used to illuminate the aluminum curtain wall to eliminate shadows on its surface;

[0041] An image acquisition unit, which is connected to the lighting unit and is used to acquire image information of the curtain wall surface;

[0042] An image processing unit, connected to the image acquisition unit, for preprocessing image information;

[0043] A feature extraction unit connected to the image processing unit, used to obtain defect features of the pre-processed image information, and to perform partition processing on the pre-processed image information;

[0044] An analysis unit connected to the feature extraction unit, configured to determine whether the curtain wall meets the standards based on the acquired defect features, and to determine the reasons for non-compliance with the standards based on the determination results, and to generate corresponding instructions according to the reasons;

[0045] The processing unit is connected to the analyzing unit and is used to issue an anti-corrosion treatment notification based on the received instruction, or to redetermine the corresponding detection parameters.

[0046] Specifically, in this embodiment, the lighting unit and the image acquisition unit are both fixed on the guide rail, and the illumination angle between the lighting unit and the image acquisition unit is achieved by adjusting the position of the lighting unit on the guide rail. The background transmits the image information obtained by the image acquisition unit to the image processing unit, performs noise reduction and other preprocessing operations on it, and inputs the image information after the preprocessing operation into the feature extraction unit, uses the YOLOv5 model to extract features, obtains the defect features of the image information, and divides it into multiple partitions of the same area. Subsequently, the multiple partitions will be periodically inspected to analyze the reasons why the curtain wall does not meet the standards, and generate corresponding instructions or redetermine the detection parameters according to the reasons, wherein the detection parameters include preprocessing parameters (noise reduction ratio, grayscale transformation parameters), the illumination angle of the lighting unit and the illumination light intensity.

[0047] See also Figure 2 As shown, it is a step flow chart of a surface detection method for energy-saving and environmentally friendly aluminum plate curtain wall based on machine vision in an embodiment of the present invention.

[0048] The steps of the system in the embodiment of the present invention during actual operation include:

[0049] S1, the lighting unit illuminates the aluminum curtain wall;

[0050] S2, the image acquisition unit acquires image information of the surface of the aluminum curtain wall in real time;

[0051] S3, the image processing unit performs image preprocessing on the image information;

[0052] S4, the feature extraction unit detects the image information using the YOLOv5 model based on the image information after image preprocessing, obtains the corrosion position on the surface of the aluminum plate curtain wall, and detects the clarity of the boundary of the corrosion position;

[0053] S5, the analysis unit divides the aluminum plate curtain wall at each moment into a plurality of partitions with the same area, and calculates the area ratio of the area of ​​the corroded part in each partition to the area of ​​the corresponding partition;

[0054] S6, the analysis unit further determines whether the aluminum curtain wall meets the standards based on the several area ratios at the same time, and analyzes the reasons why the surface of the aluminum curtain wall does not meet the standards according to the determination result.

[0055] Specifically, the aluminum curtain wall is first illuminated, and the main operation is to select a suitable light source type (such as natural light, artificial light source, point light source, parallel light source, etc.), and determine the position, angle and intensity of the light source to control shadows, highlights and overall brightness. Secondly, an industrial camera is used to capture images of the aluminum curtain wall after lighting, and then the collected image information is preprocessed, wherein the preprocessing includes grayscale processing and noise reduction operations. Then, the preprocessed image information is input into the trained YOLOV5 model for detection, and the output detection result is the corrosion position on the surface of the aluminum curtain wall, and the boundary of the corrosion position is detected using a clarity tester. Then, the aluminum curtain wall is divided into several partitions of the same area, and the ratio of the area of ​​the corroded part in each partition to the area of ​​the corresponding partition is calculated respectively. Finally, based on these ratios, it is judged whether the surface of the aluminum curtain wall does not meet the standards, and the reasons why the aluminum curtain wall does not meet the standards are preliminarily analyzed to avoid incorrect detection results caused by other factors.

[0056] Specifically, the core idea of ​​YOLOv5 is to regard the entire object detection task as a regression problem. It divides the image into multiple regions (called grid cells) and predicts the bounding box and category probability for each grid cell. The architecture of YOLOv5 mainly includes the following parts: Backbone is responsible for extracting features from the image, usually using CSPDarknet53 as its backbone network; Neck is responsible for integrating and transmitting the extracted features, and Panoptic Feature Pyramid Network (PANet) is used in YOLOv5; Head is responsible for generating bounding boxes and category probabilities, using YOLO's original prediction method.

[0057] See also Figure 3 As shown, it is a flowchart of the steps of comparing the average value of several area ratios with the preset area ratio at the same time in an embodiment of the present invention. The process of judging whether the aluminum plate curtain wall meets the standard based on the obtained area ratios in the embodiment of the present invention includes: calculating the average value of each area ratio, and judging whether the aluminum plate curtain wall meets the standard based on the comparison result of the obtained average value and the pre-stored preset area ratio, wherein: if the average value is less than or equal to the first preset average value, the aluminum plate curtain wall is judged to meet the standard; if the average value is greater than the first preset average value and less than the second preset average value, whether the aluminum plate curtain wall meets the standard is judged based on the variance of each area ratio; if the average value is greater than or equal to the second preset average value, the aluminum plate curtain wall is judged to be non-compliant with the standard, and the reason why the aluminum plate curtain wall does not meet the standard is determined based on the difference between the average value and the second preset average value.

[0058] Specifically, in this embodiment, the average value L0 of several area ratios at the same time can be divided into a first preset average value L1 and a second preset average value L2, and the average value standard L3=0.5, L1=L3, L2=1.5×L3 is set. It should be noted that in other embodiments, the values ​​of L1, L2 and L3 can also be determined according to the relative component quality requirements, and the vertical values ​​of the matching items are rounded up; the comparison process based on the average value L with L1 and L2 is as follows:

[0059] If the average value L is less than or equal to the first preset average value L1, it can be determined that the surface of the current aluminum curtain wall meets the standard;

[0060] If the average value L is greater than the first preset average value L1 and less than the second preset average value L2, it is impossible to determine whether the detection result is caused by other factors, so it is necessary to perform a secondary determination on the variance P of the area ratio of the aluminum plate curtain wall surface at the same time;

[0061] If the average value L is greater than or equal to the second preset average value L2, it is determined that the current aluminum curtain wall surface does not meet the standard, and the reason why the aluminum curtain wall does not meet the standard is analyzed based on the difference between the average value of the actual area ratio and the second preset average value.

[0062] See also Figure 4 As shown, it is a flow chart of the steps of judging based on the variance of the area ratio at the same time and the preset variance in an embodiment of the present invention. The process of judging whether the aluminum plate curtain wall meets the standard based on the variance of the average value of each area ratio in an embodiment of the present invention includes: comparing the obtained variance with the preset variance; if the variance is less than or equal to the preset variance, judging that the acquisition process for the image information does not meet the standard, and re-determining the detection parameters in the image acquisition process based on the obtained variance value; if the variance is greater than the preset variance, judging that the curtain wall does not meet the standard, and determining the reason for not meeting the standard based on the difference between the average value and the preset second preset average value.

[0063] Specifically, when the average value L is greater than the first preset average value L1 and less than the second preset average value L2, the variance of the average values ​​of the area ratios is calculated and the reason why the curtain wall does not meet the standard is determined based on the comparison result of the obtained variance and the preset variance; in this embodiment, the preset variance P0 is set to 1, and the comparison process of the variance of the area ratio at the same time and the preset variance is specifically as follows:

[0064] If the variance P is less than or equal to the preset variance P0, it means that the numerical discreteness of the area ratio of each partition is low. At this time, it is determined that there is a problem with the illumination light intensity during the image information acquisition process, and the illumination brightness during the image information acquisition process needs to be re-determined.

[0065] If the variance P is greater than the preset variance P0, it means that there is no problem in the corrosion position identification part, and it is determined that the surface of the aluminum curtain wall meets the standard. It is necessary to further determine the reason why the surface of the aluminum curtain wall does not meet the standard.

[0066] Specifically, the process of redetermining the detection parameters in the image acquisition process based on the variance value in an embodiment of the present invention includes: obtaining the difference between the variance and the preset variance; increasing the illumination light intensity based on the obtained difference, and the difference is proportional to the increase in the illumination light intensity.

[0067] Specifically, in this embodiment, the preset difference is Q0=0.5, and the comparison process based on the difference between the variance and the preset variance is as follows:

[0068] If the difference Q is less than the preset difference Q0, the illumination light intensity of the processing unit is adjusted to 1.5 times the original illumination light intensity;

[0069] If the difference Q is greater than or equal to the preset difference Q0, the illumination light intensity of the processing unit is adjusted to 3.5 times the original illumination light intensity.

[0070] See also Figure 5 As shown, it is a flowchart of the steps of the present invention based on the difference between the average value of the actual area ratio and the second preset average value and the preset difference. The process of determining the reason why the aluminum plate curtain wall does not meet the standard based on the difference between the average value and the second preset average value in the embodiment of the present invention includes: determining whether the aluminum plate curtain wall is corroded based on the comparison result of the difference between the average value and the second preset average value and the pre-stored preset difference, wherein: if the difference is less than or equal to the first preset difference, it is determined that the aluminum plate curtain wall is corroded; if the difference is greater than the first preset difference and less than the second preset difference, it is determined whether the aluminum plate curtain wall is corroded based on the average value of the clarity of the boundaries of each defect feature; if the difference is greater than or equal to the second preset difference, it is determined whether the aluminum plate curtain wall is corroded based on the secondary ratio; wherein: the secondary ratio is the ratio of the average value of the area ratio in the current detection result to the average value of the area ratio in the previous detection result.

[0071] Specifically, in this embodiment, the difference M0 can be divided into a first preset difference M1 and a second preset difference M2, and the average value standard M3=0.5, M1=M3, M2=1.5×M3 is set. It should be noted that in other embodiments, the values ​​of M1, M2 and M3 can also be determined according to the relative component quality requirements, and the vertical values ​​of the matching items are rounded up; the comparison process based on the difference M with M1 and M2 is as follows:

[0072] If the difference M0 is less than or equal to the first preset difference M1, it means that corrosion occurs on the surface of the aluminum curtain wall, and an anti-corrosion treatment notice is issued;

[0073] If the difference M0 is greater than the first preset difference M1 and less than the second preset difference M2, it means that there may be a problem in the image preprocessing process, and the reason for the corrosion on the surface of the aluminum curtain wall is analyzed based on the clarity of the boundary of the corrosion position;

[0074] If the difference M0 is greater than or equal to the second preset difference M2, the reason why the aluminum curtain wall is corroded is analyzed based on the detection results at the historical moment and the current moment.

[0075] Specifically, the process of determining whether the aluminum curtain wall is corroded based on the average clarity of the boundaries of each defect feature in the embodiment of the present invention includes:

[0076] Comparing the clarity average value with a preset clarity average value;

[0077] If the average clarity value is less than or equal to the preset average clarity value, it is determined that the preprocessing process for the image information does not meet the standard, and the preprocessing parameters are re-determined based on the difference between the average clarity value and the preset average clarity value; wherein: the detection parameters include the preprocessing parameters.

[0078] If the average clarity value is greater than the preset average clarity value, it is determined that the aluminum curtain wall is corroded, and an anti-corrosion treatment notice is issued.

[0079] Specifically, in this embodiment, the average value of the preset clarity is N0=3.4, and the comparison process of the average value of the clarity of the boundary of the corrosion position at the same time and the preset clarity is as follows:

[0080] If the average value N of the clarity is less than or equal to the average value N0 of the preset clarity, it means that there is a problem with the image acquisition, and the acquired image cannot be effectively processed, resulting in the use of blurred image information for feature recognition in actual situations, so the preprocessing parameters for the image information are adjusted;

[0081] If the average value N of the clarity is greater than the average value N0 of the preset clarity, it means that there is no problem in the image processing process, and corrosion occurs on the aluminum curtain wall, and an anti-corrosion treatment notice is issued.

[0082] Specifically, the process of re-determining the preprocessing parameters based on the difference between the clarity average value and the preset clarity average value in the embodiment of the present invention includes: increasing the noise reduction factor based on the difference between the clarity and the preset clarity, and the difference is proportional to the noise reduction factor; wherein: the preprocessing parameters include (noise reduction factor, grayscale transformation parameters).

[0083] Specifically, in this embodiment, the preset difference value H0=1, and the comparison process based on the difference between the definition and the preset definition is as follows:

[0084] If the actual difference H is less than or equal to the preset difference H0, the noise reduction ratio is adjusted to 1.2 times the original noise reduction ratio;

[0085] If the actual difference H is greater than the preset difference H0, the noise reduction ratio is adjusted to 3 times the original noise reduction ratio.

[0086] Specifically, the process of determining whether the aluminum curtain wall is corroded based on the secondary ratio in the embodiment of the present invention includes: comparing the secondary ratio with a preset secondary ratio; if the secondary ratio is less than or equal to the preset secondary ratio, it is determined that the aluminum curtain wall is corroded, and an anti-corrosion treatment notification is issued; if the secondary ratio is greater than the preset secondary ratio, it is determined that there is a problem with the illumination angle, and the illumination angle is re-determined based on the difference between the secondary ratio and the preset secondary ratio; wherein: the illumination angle is the angle between the line connecting the illumination point and the aluminum curtain wall, and the line connecting the image acquisition point and the aluminum curtain wall; the detection parameters include the illumination angle.

[0087] Specifically, in this embodiment, the preset ratio G0=1, based on the ratio of the average value of the area ratio in the detection result at the current moment to the average value of the area ratio in the detection result at the previous moment and the preset ratio, the comparison process is as follows:

[0088] If the ratio G is less than or equal to the preset ratio G0, it means that the corrosion area on the surface of the aluminum curtain wall is large and an anti-corrosion treatment notice needs to be issued;

[0089] If the ratio G is greater than the preset ratio G0, it means that there is a deviation in the illumination angle, resulting in too many shadows, so that the system misjudges the shadows as corrosion, and therefore it is necessary to adjust the illumination angle during the image acquisition process.

[0090] Specifically, the process of redetermining the illumination angle based on the difference between the secondary ratio and the preset secondary ratio in the embodiment of the present invention includes: reducing the illumination angle based on the difference, and the difference is inversely proportional to the reduction degree of the illumination angle.

[0091] Specifically, in this embodiment, the preset difference K0=1, and the processing process based on the difference between the ratio and the preset ratio is as follows:

[0092] If the difference K is greater than the preset difference KO, the angle is adjusted to 0.7 times the original angle;

[0093] If the difference K is less than or equal to the preset difference K0, the angle is adjusted to 0.5 times the original angle.

[0094] Specifically, the process of re-determining the illumination light intensity based on the illumination angle in an embodiment of the present invention includes: reducing the illumination light intensity based on the illumination angle, and the angle is inversely proportional to the degree of reduction of the illumination light intensity; wherein: the detection parameters include the illumination light intensity.

[0095] Specifically, in this embodiment, the preset angle R0=90°, and the comparison process based on the angle and the preset angle is as follows:

[0096] If the angle R is less than or equal to the preset angle R0, the intensity of the irradiation light is reduced to 0.7 times of the original irradiation light intensity;

[0097] If the angle R is greater than the preset angle R0, the intensity of the irradiation light is reduced to 0.4 times of the original intensity of the irradiation light.

[0098] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

[0099] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A surface detection method for energy-saving and environmentally friendly aluminum curtain wall based on machine vision, characterized in that: include: Light the aluminum curtain wall to eliminate shadows on its surface; Acquiring image information of the surface of the aluminum curtain wall in real time; Preprocessing the image information; Based on the preprocessed image information, the image information is detected using a YOLOv5 model to obtain defect features in the image information; Dividing the preprocessed image information into a plurality of partitions of the same area, and calculating the area ratio in each partition respectively, wherein the area ratio is the ratio of the area of ​​the defect feature in a single partition to the area of ​​the corresponding partition; Based on the obtained area ratios, it is determined whether the aluminum curtain wall meets the standards, and when it is determined that the aluminum curtain wall does not meet the standards, a corresponding treatment method is generated based on the determined reasons for not meeting the standards, including issuing an anti-corrosion treatment notice and redetermining corresponding detection parameters.

2. The surface detection method of energy-saving and environmentally friendly aluminum curtain wall based on machine vision according to claim 1 is characterized in that: The process of determining whether the aluminum plate curtain wall meets the standard based on the obtained area ratios includes: Calculate the average value of each area ratio, and determine whether the aluminum curtain wall meets the standard based on the comparison result of the obtained average value and the pre-stored preset area ratio, wherein: If the average value is less than or equal to the first preset average value, it is determined that the aluminum curtain wall meets the standard; If the average value is greater than the first preset average value and less than the second preset average value, determining whether the aluminum curtain wall meets the standard based on the variance of each area ratio; If the average value is greater than or equal to the second preset average value, it is determined that the aluminum plate curtain wall does not meet the standard, and the reason why the aluminum plate curtain wall does not meet the standard is determined based on the difference between the average value and the second preset average value.

3. The surface detection method of energy-saving and environmentally friendly aluminum curtain wall based on machine vision according to claim 2 is characterized in that: The process of determining whether the aluminum curtain wall meets the standard based on the variance of each area ratio includes: Compare the obtained variance with the preset variance; If the variance is less than or equal to the preset variance, it is determined that the acquisition process of the image information does not meet the standard, and the detection parameters in the image acquisition process are re-determined based on the obtained variance value; If the variance is greater than the preset variance, it is determined that the curtain wall does not meet the standard, and the reason for not meeting the standard is determined based on the difference between the average value and the preset second preset average value.

4. The surface detection method of energy-saving and environmentally friendly aluminum curtain wall based on machine vision according to claim 3 is characterized in that: The process of re-determining the detection parameters in the image acquisition process based on the variance value includes: Obtaining a difference between the variance and the preset variance; The illumination light intensity of the lighting is increased based on the obtained difference, and the difference is proportional to the increase amplitude of the illumination light intensity.

5. The surface detection method of energy-saving and environmentally friendly aluminum curtain wall based on machine vision according to claim 2 is characterized in that: The process of determining the reason why the aluminum panel curtain wall does not meet the standard based on the difference between the average value and the second preset average value includes: Based on the comparison result of the difference between the average value and the second preset average value and the pre-stored preset difference value, it is determined whether the aluminum plate curtain wall is corroded, wherein: If the difference is less than or equal to the first preset difference, it is determined that the aluminum curtain wall is corroded; If the difference is greater than the first preset difference and less than the second preset difference, determining whether the aluminum curtain wall is corroded based on the average value of the clarity of the boundaries of each defect feature; If the difference is greater than or equal to the second preset difference, it is determined whether the aluminum curtain wall is corroded based on the secondary ratio; wherein: the secondary ratio is the ratio of the average value of the area ratio in the current test result to the average value of the area ratio in the previous test result.

6. The surface detection method of energy-saving and environmentally friendly aluminum curtain wall based on machine vision according to claim 5 is characterized in that: The process of determining whether the aluminum plate curtain wall is corroded based on the average value of the clarity of the boundaries of each defect feature includes: Comparing the clarity average value with a preset clarity average value; If the average clarity value is less than or equal to the preset average clarity value, it is determined that the preprocessing process for the image information does not meet the standard, and the preprocessing parameters are re-determined based on the difference between the average clarity value and the preset average clarity value; If the average clarity value is greater than the preset average clarity value, it is determined that the aluminum curtain wall is corroded and an anti-corrosion treatment notice is issued.

7. The surface detection method of energy-saving and environmentally friendly aluminum curtain wall based on machine vision according to claim 6 is characterized in that: The process of re-determining the preprocessing parameters based on the difference between the average clarity value and the preset average clarity value includes: The noise reduction ratio is increased based on the difference between the definition and the preset definition, and the difference is proportional to the noise reduction ratio.

8. The surface detection method of energy-saving and environmentally friendly aluminum curtain wall based on machine vision according to claim 5 is characterized in that: The process of determining whether the aluminum curtain wall is corroded based on the secondary ratio includes: comparing the secondary ratio with a preset secondary ratio; If the secondary ratio is less than or equal to the preset secondary ratio, it is determined that the aluminum curtain wall is corroded, and an anti-corrosion treatment notice is issued; If the secondary ratio is greater than the preset secondary ratio, it is determined that there is a problem with the illumination angle, and the illumination angle is redetermined based on the difference between the secondary ratio and the preset secondary ratio; wherein: the illumination angle is the angle between the line connecting the illumination point and the aluminum curtain wall, and the line connecting the image acquisition point and the aluminum curtain wall.

9. The surface detection method of energy-saving and environmentally friendly aluminum curtain wall based on machine vision according to claim 8 is characterized in that: The process of redetermining the irradiation angle based on the difference between the secondary ratio and the preset secondary ratio includes: The illumination angle is reduced based on the difference, and the difference is inversely proportional to the reduction of the illumination angle.

10. The surface detection method of energy-saving and environmentally friendly aluminum curtain wall based on machine vision according to claim 9 is characterized in that: The process of re-determining the illumination light intensity based on the illumination angle includes: The illumination light intensity is reduced based on the illumination angle, and the angle is inversely proportional to the reduction amplitude of the illumination light intensity.

Images