Chemical production site online inspection method and system based on closed circuit television system

Abstract

The embodiment of the invention provides a chemical production site online inspection method, device and system based on a closed circuit television system. The chemical production site on-line inspection method based on the closed circuit television system comprises the following steps: acquiring a target image shot for each to-be-acquired image area of a chemical production site from the closed circuit television system; according to a first name specifying rule and a first path specifying rule, storing each target image in a first path with a respective name; according to a comparison result between each target image and a matched comparison reference image, an inspection score for each to-be-collected image area is generated, and the matched comparison reference image is determined according to the name of the target image; and according to a second name specifying rule and a second path specifying rule, storing the generated inspection scores in a file of a second path which can be accessed by the client by respective names.

Description

Technical Field

[0001] The embodiments of this specification generally relate to the technical field of image monitoring and processing in the chemical production scenario, and particularly to an online inspection method, device, and system for chemical production sites based on a closed-circuit television system. Background Art

[0002] In the field of chemical production, traditional inspection methods usually rely on manual regular inspections, but they are often inefficient and difficult to conduct high-frequency, comprehensive, and detailed inspections on large-scale chemical production sites, resulting in some potential problems being difficult to detect in a timely manner. Moreover, manual judgment is easily affected by subjective factors, and the objectivity of inspection results is insufficient. There are also some methods that deploy various sensors around chemical production equipment to collect relevant data and transmit it to the monitoring center for analysis and judgment to determine whether the equipment operation status and environmental conditions are normal. For closed-circuit television systems, corresponding personnel are often required to be on duty in the control center to view the images transmitted from the production site at any time. With the expansion of chemical production scale and the progress of automation technology, a more efficient, objective, and labor-saving online inspection solution is needed. Summary of the Invention

[0003] In view of the above, the embodiments of this specification provide an online inspection method, device, and system for chemical production sites based on a closed-circuit television system. Using this online inspection solution, first, the target images obtained as needed are stored in an orderly manner according to the designed first name designation rule and first path designation rule. Then, an inspection score is generated for the target images by means of comparison reference images determined according to the target image names, and the inspection score is stored in a file accessible to the client using the second name designation rule and second path designation rule. Thus, it not only realizes fast, objective, and timely inspection of the image acquisition area to be inspected, but also provides a technical basis for the remote monitoring end or other relevant systems to obtain inspection result data in a timely manner, which is conducive to realizing data sharing and collaborative work among multiple systems, laying a foundation for improving the overall management level and emergency response ability of chemical production sites, and ensuring the safe and stable operation and production efficiency of chemical production.

[0004] According to one aspect of the embodiments of the present specification, there is provided an on-line inspection method for a chemical production site based on a closed-circuit television system, including: obtaining target images captured for each image area to be collected at the chemical production site from the closed-circuit television system; storing each target image in a first path with its own name according to a first name designation rule and a first path designation rule; generating an inspection score for each image area to be collected according to the comparison result between each target image and a matching comparison reference image, wherein the matching comparison reference image is determined according to the name of the target image; and storing each generated inspection score in a file in a second path accessible by a client with its own name according to a second name designation rule and a second path designation rule.

[0005] According to another aspect of the embodiments of the present specification, there is provided an on-line inspection device for a chemical production site based on a closed-circuit television system, including: an image acquisition module configured to obtain target images captured for each image area to be collected at the chemical production site from the closed-circuit television system; an image storage module configured to store each target image in a first path with its own name according to a first name designation rule and a first path designation rule; a score generation module configured to generate an inspection score for each image area to be collected according to the comparison result between each target image and a matching comparison reference image, wherein the matching comparison reference image is determined according to the name of the target image; and a file storage module configured to store each generated inspection score in a file in a second path accessible by a client with its own name according to a second name designation rule and a second path designation rule.

[0006] According to yet another aspect of the embodiments of the present specification, there is provided an on-line inspection system for a chemical production site based on a closed-circuit television system, including: a closed-circuit television system for collecting and storing images captured for each image area to be collected at the chemical production site; a server for executing the on-line inspection method for a chemical production site based on a closed-circuit television system as described above; and a client for extracting corresponding inspection scores from each file under the second path; determining area identifiers corresponding to the extracted inspection scores according to a name-area identifier comparison table matching the second name designation rule; and presenting an on-line inspection result in the form of area identifier-inspection score.

[0007] According to still another aspect of the embodiments of the present specification, there is provided an on-line inspection device for a chemical production site based on a closed-circuit television system, including: at least one processor, and a memory coupled to the at least one processor, the memory storing instructions that, when executed by the at least one processor, cause the at least one processor to execute the on-line inspection method for a chemical production site based on a closed-circuit television system as described above. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] By referring to the following drawings, a further understanding of the essence and advantages of the content of this specification can be achieved. In the drawings, similar components or features may have the same reference numerals.

[0009] Figure 1 An exemplary architecture of a chemical production site on - line inspection method, device and system based on a closed - circuit television system according to an embodiment of this specification is shown.

[0010] Figure 2 A flowchart of an example of a chemical production site on - line inspection method based on a closed - circuit television system according to an embodiment of this specification is shown.

[0011] Figure 3 A flowchart of an example of the generation process of inspection scores according to an embodiment of this specification is shown.

[0012] Figure 4 A flowchart of another example of a chemical production site on - line inspection method based on a closed - circuit television system according to an embodiment of this specification is shown.

[0013] Figure 5 A flowchart of an example of the determination process of similarity according to an embodiment of this specification is shown.

[0014] Figure 6 A flowchart of an example of the determination process of similarity according to an embodiment of this specification is shown.

[0015] Figure 7 A block diagram of an example of a chemical production site on - line inspection device based on a closed - circuit television system according to an embodiment is shown.

[0016] Figure 8 A signaling diagram of an example of the interaction between various devices in a chemical production site on - line inspection system according to an embodiment of this specification is shown.

[0017] Figure 9 A schematic diagram of an example of a chemical production site on - line inspection device based on a closed - circuit television system according to an embodiment of this specification is shown. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] The subject matter described herein will be discussed with reference to example embodiments. It should be understood that discussing these embodiments is only to enable those skilled in the art to better understand and thus implement the subject matter described herein, and is not a limitation on the scope of protection, applicability, or examples set forth in the claims. The functions and arrangements of the elements discussed may be changed without departing from the scope of protection of the embodiments of this specification. Each example may omit, substitute, or add various processes or components as needed. Additionally, the features described relative to some examples may be combined in other examples.

[0019] As used herein, the term "comprising" and its variants denote open terms meaning "including but not limited to". The term "based on" means "at least partially based on". The term "one embodiment" and "an embodiment" mean "at least one embodiment". The term "another embodiment" means "at least one other embodiment". The terms "first", "second", etc. may refer to different or the same objects. Other definitions may be included below, whether explicit or implicit. Unless explicitly specified in the context, the definition of a term is consistent throughout the specification.

[0020] The flowcharts used in this specification illustrate the operations implemented by a system according to some embodiments of this specification. It should be clearly understood that the operations in the flowchart may not be implemented in sequence. On the contrary, the operations may be implemented in reverse order or simultaneously. In addition, one or more other operations may be added to the flowchart. One or more operations may be removed from the flowchart.

[0021] In view of this, the embodiments of this specification propose an on - line inspection scheme for chemical production sites based on a closed - circuit television system. Using this on - line inspection scheme, first, the target images obtained as needed are stored in an orderly manner according to the designed first name - specifying rule and the first path - specifying rule. Then, an inspection score is generated for the target images by means of a comparison reference image determined according to the target image name, and the inspection score is stored in a file accessible to the client using the second name - specifying rule and the second path - specifying rule. Thus, it not only realizes the fast, objective, and timely inspection of the area where images are to be collected, but also provides a technical basis for the remote monitoring end or other related systems to obtain inspection result data in a timely manner, which is conducive to realizing data sharing and collaborative work among multiple systems, laying a foundation for improving the overall management level and emergency response ability of chemical production sites, and ensuring the safe and stable operation and production efficiency of chemical production.

[0022] The on - line inspection method, device, and system for chemical production sites based on a closed - circuit television system according to the embodiments of this specification will be described in detail below with reference to the accompanying drawings.

[0023] Figure 1An exemplary architecture 100 of a chemical production site online inspection method, apparatus, and system based on a closed-circuit television system according to an embodiment of the present specification is shown.

[0024] In Figure 1 , a network 110 is applied to interconnect between a closed-circuit television (CCTV) system 120 and an application server 130.

[0025] The network 110 can be any type of network capable of interconnecting network entities. The network 110 can be a single network or a combination of various networks. In terms of coverage, the network 110 can be a local area network (LAN), a wide area network (WAN), etc. In terms of the bearer medium, the network 110 can be a wired network, a wireless network, etc. In terms of data exchange technology, the network 110 can be a circuit-switched network, a packet-switched network, etc.

[0026] The closed-circuit television system 120 can refer to an image communication system capable of video transmission in a specific area. In some examples, the closed-circuit television system 120 can include several parts such as an image acquisition device, an image transmission device, an image control device, and an image processing, display, and recording device. In some examples, the image acquisition device can be a camera. Different CCTV cameras can be respectively set at designated positions according to the plant layout, production process, and functional characteristics of different areas of the chemical enterprise. In one example, in the production workshop, high-definition cameras can be set for key production equipment, material stacking areas, operation channels, and other key areas. For example, cameras with high resolution (such as 4K), wide dynamic range, and low illumination functions can be selected to adapt to the complex lighting environment of chemical enterprises (such as production workshops with strong light irradiation and dark warehouse corners), so as to ensure the clarity of the image and retain image details. Another example is that cameras with high stability and durability can be selected, so that they can operate stably for a long time in the special environment of chemical enterprises (such as the presence of chemical gases, dust, etc.).

[0027] The server 130 can access the closed-circuit television system 120 through the network 110, so as to obtain the images collected by the closed-circuit television system 120. Furthermore, the server 130 can generate corresponding inspection scores based on the obtained images and store the inspection scores in the corresponding files under the specified path.

[0028] In one implementation, the terminal device used by the user ( Figure 1(not shown in the figure) can access the server 130 and read the content of the file at the specified path above. In some examples, the terminal device can be any type of electronic computing device capable of accessing the server 130, processing data or signals, etc. For example, the terminal device 120 can be a laptop computer, a tablet computer, a smart phone, etc.

[0029] It should be understood that Figure 1 all the network entities shown in the figure are exemplary, and according to specific application requirements, any other network entities may be involved in the architecture 100.

[0030] Figure 2 FIG. shows a flowchart of an example of an on-site online inspection method 200 for a chemical production site based on a closed-circuit television system according to an embodiment of the present specification.

[0031] As Figure 2 shown, at 210, target images are obtained from the closed-circuit television system for photographing each image acquisition area to be collected at the chemical production site.

[0032] In this embodiment, the image acquisition areas to be collected usually need to include key attention areas at the chemical production site. For example, the key attention areas can include areas where important production equipment is located, material storage areas, operation workstations, finished product packaging areas, emergency passage areas, etc. In some examples, the closed-circuit television system can collect current images of each image acquisition area to be collected at the chemical production site in real time. In some examples, the closed-circuit television system can store the images of each image acquisition area to be collected at the chemical production site and the corresponding acquisition times. In some examples, the collected images can be obtained by remotely accessing the closed-circuit television system.

[0033] In some examples, images can be obtained from the closed-circuit television system for photographing each image acquisition area to be collected at the chemical production site in multiple ways, such as based on a predetermined time interval (e.g., every 15 minutes), triggering by a predetermined type of event (e.g., detecting that a person or vehicle enters a specified area), or a predetermined type of instruction (e.g., receiving a start online inspection instruction sent by a client). For example, the obtained images can be directly used as the target images for the corresponding image acquisition areas to be collected. For another example, the obtained images can be subjected to image preprocessing (such as denoising the images, enhancing the contrast, etc.), and the preprocessed images can be used as the target images for the corresponding image acquisition areas to be collected.

[0034] In some implementations, it is possible to sequentially retrieve from a closed-circuit television system the images captured for each image area to be collected at a chemical production site by simulating mouse operations and perform image capture. Among them, the mouse operations may include at least one of the following: moving, clicking, and scroll wheel operation. In some examples, an automated scripting software can be used to simulate operations such as mouse movement, clicking, and scroll wheel rolling by choreographing task actions. For example, the scroll wheel operation of the mouse can be simulated to switch the camera, and the mouse click on the screenshot button can be simulated to capture the image captured by the current camera. Thus, it is possible to automatically obtain the target images from the closed-circuit television system.

[0035] At 220, according to the first name specification rule and the first path specification rule, store each target image in the first path with its respective name.

[0036] In this embodiment, the first name specification rule can be set according to actual needs. For example, it can be named according to the camera number or the area identifier of the image area to be collected. The area identifier can be, for example, an area code. Alternatively, the name of the target image can be determined based on information such as the image acquisition time, location, and area number. For example, the format "acquisition time_location number_area number.jpg" can be used. In some examples, the first path can be a server folder or a local disk path specifically designated for storing target images.

[0037] At 230, generate an inspection score for each image area to be collected according to the comparison result between each target image and the matching comparison reference image.

[0038] In this embodiment, the comparison reference image can be an image of the image area to be collected when it meets the inspection standard, such as an image with a full score for the inspection score. In some examples, the comparison reference images can correspond one-to-one with the image areas to be collected, that is, different image areas to be collected can correspond to different comparison reference images. In some examples, different comparison reference images can be distinguished by area identifiers. In some examples, for key areas in a chemical production site, such as a large equipment area, it can correspond to a group of comparison reference images. Each group of comparison reference images can contain multiple comparison reference images taken from different angles. In some examples, the number of target images and perspectives required can be determined according to the importance and complexity of the image area to be collected.

[0039] In this embodiment, comparison reference images corresponding to each image region to be collected can be pre-stored. After obtaining the target image, the matching comparison reference image can be determined according to the name of the target image. For example, if the name of the target image contains a region number, the comparison reference image corresponding to the image region to be collected indicated by the region number found according to the region number can be determined as the matching comparison reference image.

[0040] In this embodiment, the degree of difference between the target image and the matching comparison reference image can be determined in various ways, and an inspection score can be generated according to the degree of difference. In some examples, the target image and the matching comparison reference image can be compared pixel by pixel, and the inspection score can be determined by synthesizing the comparison results of each pixel. In some examples, preprocessing can be performed on the target image and the matching comparison reference image, such as normalization processing, grayscale processing, etc. After that, the structural similarity index (SSIM) between the preprocessed target image and the matching comparison reference image can be determined, and then the inspection score can be determined accordingly. It can be understood that generally, the smaller the degree of difference between the target image and the matching comparison reference image, the higher the inspection score.

[0041] At 240, according to the second name specification rule and the second path specification rule, each generated inspection score is stored in a file at the second path that can be accessed by the client under its respective name.

[0042] In this embodiment, the second name specification rule can adopt a naming method similar to that of the target image for the convenience of association and query. For example, the name of the inspection score can adopt the format of "region identifier.txt". Another example is that the name of the inspection score can adopt the format of "region number_inspection time.txt" or "region number_inspection time_score value.txt". And the inspection score can be written into the corresponding file. In some examples, the second path can be specified as a network shared folder or a specific table field in a database so that the client can conveniently access and query the inspection score.

[0043] It should be understood that all steps and their orders in process 200 are exemplary, and the embodiments of the present disclosure will also cover any modifications to process 200. For example, although Figure 2 step 230 shown is executed after step 220, it can also be executed before step 220.

[0044] Next, referring to Figure 3 , Figure 3 shows a flowchart of an example of the inspection score generation process 300 according to an embodiment of the present specification. Process 300 can be Figure 2An exemplary implementation of step 230 in

[0045] As Figure 3 shown, at 310, the similarity between each target image and the matching comparison reference image is determined.

[0046] In this embodiment, multiple image similarity algorithms can be used to determine the similarity between the target image and the comparison reference image. For example, methods such as based on color feature histograms, based on the difference between corresponding pixel values, and based on the structural similarity index can be used to determine the similarity between the target image and the matching comparison reference image. Another example is that a similarity calculation method based on the extracted feature vectors can be used to determine the similarity between the target image and the matching comparison reference image. In one example, the feature vectors corresponding to the similarity between the target image and the matching comparison reference image can be extracted respectively. For example, algorithms such as SURF (Speeded Up Robust Feature), SIFT (Scale-invariant feature transform), and ORB (Oriented FAST and Rotated BRIEF) can be used to extract feature points and construct feature vectors corresponding to the target image and the matching comparison reference image respectively, and then the similarity between the two constructed feature vectors (such as Euclidean distance, cosine similarity, etc.) is used to determine the similarity between the target image and the matching comparison reference image.

[0047] At 320, the determined similarity is calibrated using the calibration coefficients corresponding to each target image to obtain the calibrated similarity score.

[0048] In this embodiment, the calibration coefficients can be determined by fitting the similarity of historical images and the corresponding benchmark inspection scores. For example, historical scoring data corresponding to each comparison reference image can be collected in advance. For each comparison reference image, each piece of historical scoring data can be composed of the similarity determined for the comparison reference image and the corresponding historical image and the manually labeled scoring data. For example, for the image area 1 to be collected, images regularly collected every day in the past month can be selected as historical images. For each historical image and the comparison reference image of the image area 1 to be collected, the similarity can be determined and the manually labeled scoring data can be obtained. In some examples, the determined similarity can be fitted with the corresponding scoring data using the collected historical scoring data. For example, each similarity can be denoted as x, the corresponding scoring data can be denoted as y, and the regression equation y = ax + b can be fitted by the least squares method to determine the values of the calibration coefficients a and b for the comparison reference image. Furthermore, the similarity of the newly obtained target image corresponding to the comparison reference image can be converted into a calibrated similarity score using the values of the calibration coefficients a and b.

[0049] In some examples, the range of the determined similarity can be divided into several intervals (bins) using the binning principle, and then the mean or median of the scoring data in each interval can be statistically calculated and used as the calibration coefficient of the comparison reference image in each interval. For example, the range of the similarity determined for each historical image of the image area 1 to be collected and the comparison reference image of the image area 1 to be collected can be divided into [0, 10), [10, 20), …, [90, 100]. Taking the interval [60, 70) as an example, the average (or median) of the scores of the manually labeled scoring data corresponding to the historical images with similarities in this interval is statistically calculated. For example, if the mean is 67, the calibration coefficient for this interval can be 67. Similarly, the calibration coefficients for each interval can be obtained. Thus, when the similarity of the newly obtained target image corresponding to the comparison reference image falls into the corresponding interval, the calibration coefficient of this interval can be determined as the calibrated similarity score.

[0050] It can be understood that according to the above method, the calibration coefficients corresponding to each comparison reference image can be determined. After the target image is determined, the corresponding calibration coefficient can be determined according to the matched comparison reference image.

[0051] At 330, corresponding inspection scores are generated according to each similarity score.

[0052] In this embodiment, the inspection scores corresponding to each similarity score can be determined according to a pre-set correspondence table between the similarity scores and the inspection scores. For example, the similarity scores can be divided into several intervals, and each interval corresponds to a specific inspection score. In some examples, other factors can also be combined to adjust the inspection scores. For example, if specific abnormal features are detected in the target image, such as the presence of smoke, obstacles, etc., the inspection score can be correspondingly reduced. In some examples, a pre-set mapping function can be used to normalize the inspection scores. For example, the inspection scores can be mapped to the interval [0, 100].

[0053] Through the above method, the proposed solution makes the inspection scores fit the manually marked standard to a certain extent by introducing the calibration coefficient. Compared with the supervised training process of machine learning models, it is simpler, more intuitive, and easier to implement, thus helping to make the inspection scores more objectively reflect the relationship between the similarity scores of the images and the actual production situation, providing strong support for the safety management of chemical production.

[0054] Continue to refer to Figure 4 , Figure 4 shows a flowchart of another example of the on-site online inspection method 400 for chemical production based on a closed-circuit television system according to an embodiment of the present specification.

[0055] As Figure 4 shown, at 410, target images are obtained from the closed-circuit television system for photographing each area of the chemical production site to be collected.

[0056] In some implementation manners, the comparison reference image can have a corresponding shooting time. In some examples, the shooting time of the comparison reference image can be used as an important reference factor for determining when to trigger the operation of obtaining the target image. Obtaining the target images for photographing the areas of the chemical production site to be collected can be performed in response to the current time matching the shooting time of the matching comparison reference image. In some examples, when the system detects that the current time is the same as the shooting time of a certain comparison reference image, the operation of obtaining the target image from the closed-circuit television system can be automatically triggered. For example, if the comparison reference image is taken within a specific time period (e.g., 20:00 - 21:00), then the target images can be obtained within the same time period to ensure the accuracy of the comparison. In some examples, a certain time tolerance can also be set, that is, when the current time is close to the shooting time of the comparison reference image within a certain range, it is also considered to match and the operation of obtaining the target image is performed. For example, the time tolerance can be set to plus or minus 10 minutes.

[0057] In some implementations, obtaining a target image for photographing a to-be-acquired image area at a chemical production site may be performed in response to sensor data in the to-be-acquired image area indicating an abnormal situation. In these implementations, a chemical production site often has multiple sensors for collecting data such as temperature, pressure, liquid level, gas leakage, etc. In some examples, the above sensors may establish a data connection with a closed-circuit television system and a device (such as the server 130 in Figure 1 to perform an on-line inspection method for a chemical production site based on the closed-circuit television system). In some examples, the closed-circuit television system may obtain sensor data at the corresponding moment while collecting images, and may perform time synchronization and associated storage of the image data of each to-be-acquired image area and the sensor data within the to-be-acquired image area.

[0058] In some examples, when the sensor detects abnormal data, it can immediately trigger an operation to obtain a target image for the corresponding to-be-acquired image area from the closed-circuit television system so as to timely understand the on-site situation. For example, if the temperature sensor detects an excessively high temperature value, or the pressure sensor detects an abnormal pressure change, or the liquid waste pool detects that the liquid level is approaching the warning threshold, the above operation of obtaining the target image can be automatically triggered. In some examples, different levels of sensor abnormal situations can also be preset to correspond to different target image acquisition priorities. For example, for serious abnormal situations, the target image is preferentially obtained and urgently processed; for general abnormal situations, the target image can be obtained within a certain time range (such as 10 minutes). At this time, step 420 can also be executed first, and then step 410.

[0059] In 420, for each target image, collect sensor data of each sensor in the corresponding to-be-acquired image area at the shooting moment of the target image.

[0060] In some examples, the current sensor data can be directly collected at the shooting moment of the determined target image. In some examples, the sensor data can be monitored and recorded in real time. In some examples, the shooting moment can be determined according to the timestamp of the target image, and the corresponding sensor data can be extracted accordingly.

[0061] In 430, store each target image in a first path with its own name according to a first name designation rule and a first path designation rule.

[0062] In 440, determine the similarity between each target image and a matching comparison reference image.

[0063] In 450, calibrate the determined similarity using a calibration coefficient corresponding to each target image to obtain a calibrated similarity score.

[0064] At 460, according to each similarity score and the corresponding sensor data, a corresponding patrol inspection score is generated.

[0065] In this embodiment, the similarity score and the sensor data can be combined in various ways to generate the patrol inspection score. In some examples, different weights can be assigned to the similarity score and different types of sensor data in advance, and then the weighted sum is calculated as the patrol inspection score. In some examples, the patrol inspection score can be dynamically adjusted according to the change trend of the sensor data. For example, if the sensor data shows that the condition of a certain area is gradually deteriorating, then a certain score can be correspondingly reduced based on the similarity score to obtain the patrol inspection score of that area. In some examples, a machine learning model can be pre-trained, and the patrol inspection score corresponding to the target image is obtained by inputting the similarity score and the corresponding sensor data into the machine learning model.

[0066] At 470, according to the second name specification rule and the second path specification rule, each generated patrol inspection score is stored in a file at the second path accessible to the client with its respective name.

[0067] It should be noted that the above steps 410, 430 - 450 can respectively refer to the corresponding descriptions of steps 210 - 220 in the foregoing Figure 2 embodiment and Figure 3 steps 310 - 320 in the

[0068] embodiment, which will not be elaborated here. It should also be noted that the above step 420 can be executed before steps 430 - 450, or after any one of steps 430 - 450, or before step 410.

[0069] Next, referring to Figure 5 , Figure 5 FIG. shows a flowchart of an example of the determination process 500 of the similarity according to an embodiment of the present specification. The process 500 can be Figure 3 an exemplary implementation of step 310 in

[0070] As Figure 5 shown, at 510, local texture features are extracted for each part of the target image.

[0071] In this embodiment, various texture feature extraction algorithms can be used, such as Gray-Level Co-occurrence Matrix (GLCM), Local Binary Patterns (LBP), etc. For example, the GLCM can be used to extract features such as texture roughness and contrast in different regions of each target image. In some examples, different weights can be preset according to the importance of different local regions of the target image. For example, for local images involving key equipment or high-risk areas, a larger weight coefficient can be corresponding when extracting texture features. In some examples, the extracted local texture features can also be preprocessed, such as normalization, to eliminate the scale differences between different locals. For example, the extracted texture feature values can be mapped to a specific interval so that the texture features of different locals are comparable. By combining the GLCM and LBP methods, typical scenarios such as the texture state of the production equipment surface (such as whether there is wear, scratches, etc.) and the cleaning texture of the ground (such as whether there are stains, special textures formed by debris accumulation) can be effectively identified.

[0072] At 520, the extracted local texture features are fused to obtain the texture feature vector of the target image.

[0073] In this embodiment, methods such as weighted average and principal components analysis (PCA) can be used to fuse the extracted local texture features, so that multiple local texture feature vectors can be dimensionally reduced and fused into a more representative texture feature vector. In some examples, the fusion method can be dynamically adjusted according to the reliability of different local texture features. For example, if there is significant noise interference (i.e., low reliability) in the extraction process of some local texture features, their weights in the fusion can be reduced. In some examples, the fused texture feature vector can also be further optimized, such as removing redundant features. For example, a feature selection algorithm can be used to select several features that are most valuable for similarity judgment from the fused texture feature vector.

[0074] At 530, according to the geometric shape detected from the target image, the layout feature vector of the target image is generated.

[0075] In this embodiment, algorithms such as edge detection (e.g., Canny edge detection) and Hough transform can be used to detect geometric shapes in an image. For example, an edge detection algorithm can be used to extract the contour of an object in the image, and then the Hough transform can be used to detect geometric shapes such as lines and circles (e.g., circular chemical containers, rectangular shelves, etc.), thereby constructing a layout feature vector. In some examples, different weights can be set for different types of geometric shapes to highlight important layout features. For example, for regular-shaped equipment in a chemical production site, a higher weight can be given, while a lower weight can be given to the irregular-shaped background area. In some examples, geometric parameters (such as area, perimeter, aspect ratio, etc.) can be further calculated based on the detected geometric shapes, and the calculated geometric parameters can be used as an element in the constructed layout feature vector.

[0076] At 540, extract the color features of the target image.

[0077] In this embodiment, methods such as color histograms and color moments can be used to extract color features. In some examples, a feature histogram extraction method based on the HSV color space can be adopted, where the hue, saturation, and brightness of the image are divided into multiple intervals respectively, and the number distribution of pixels in each interval is statistically counted to form a color feature histogram.

[0078] At 550, compare the extracted texture feature vector, layout feature vector, and color features with the corresponding features of the matching reference image to obtain texture feature comparison results, layout feature comparison results, and color feature comparison results.

[0079] In this embodiment, methods such as distance metrics and similarity functions can be used for comparison. In some examples, the Euclidean distance between the texture feature vectors of the target image and the reference image can be used as the texture feature comparison result. The texture feature comparison result can also be obtained based on the intersection distance of the LBP histogram and the difference degree of the GLCM feature parameters. In some examples, the Euclidean distance between the layout feature vectors of the target image and the reference image can be used as the layout feature comparison result. The difference degree of geometric parameters (such as area difference, perimeter difference, etc.) between the extracted shape and the corresponding shape extracted from the reference image can be further calculated, so that a layout feature comparison result integrating the difference degree of geometric parameters can be obtained. In some examples, the Bhattacharyya Distance between the color feature histograms of the target image and the reference image can be used as the color feature comparison result.

[0080] In some examples, normalization processing can be performed on the texture feature comparison result, layout feature comparison result, color feature comparison result, etc. For example, the texture feature comparison result, layout feature comparison result, and color feature comparison result can be respectively mapped to the interval [0, 1], and the closer to 1, the more similar.

[0081] At 560, according to the fusion result among the obtained texture feature comparison result, layout feature comparison result, and color feature comparison result, determine the similarity between the target image and the matching comparison reference image.

[0082] In this embodiment, methods such as weighted summation and machine learning models can be used for fusion. For example, a neural network model for outputting the comprehensive similarity can be pre-trained, taking the texture feature comparison result, layout feature comparison result, and color feature comparison result as inputs, and outputting the similarity between the target image and the comparison reference image. In some examples, the above neural network model can include an adaptive weight module based on the attention mechanism, which can dynamically generate the weights of various feature comparison results.

[0083] Through the above method, this solution can comprehensively utilize multi-dimensional features such as the texture feature, layout feature, and color feature of the target image, so as to more comprehensively and accurately judge the similarity between the target image and the comparison reference image, especially suitable for equipment operation status monitoring, production process compliance inspection, and 5S (Sorting Seiri, Straightening Seiton, Sweeping Seiso, Standardizing Seiketsu, and Sustaining Shitsuke) management, etc. in the complex environment of the chemical production site. For example, the change in the appearance texture of the equipment may indicate wear, corrosion, or material adhesion, etc., and can also be used to identify the cleanliness and smoothness of the equipment surface to judge whether the sorting and sweeping work is in place. Another example is that the difference in layout features can directly show whether the placement of equipment, tools, and materials meets the requirements of straightening, whether the space utilization is reasonable, and can also reflect key information such as whether the equipment has been displaced and whether the pipeline connection is normal. The color feature helps to distinguish the functional positioning of different areas and the integrity of warning signs, examines the implementation of cleaning and discipline, and the change in color feature may also imply potential dangers such as chemical leakage and discoloration caused by abnormal temperature.

[0084] Continue to refer to Figure 6 , Figure 6 shows a flowchart of an example of the similarity determination process 600 according to an embodiment of this specification. The patrol inspection score can include the 5S management score. The process 600 can be Figure 5Another exemplary implementation of step 310 in [the above]. Process 600 is applicable to a target image group corresponding to each to-be-acquired image area of a predetermined type. Process 600 can be executed for each target image group.

[0085] In this embodiment, for a to-be-acquired image area of a predetermined type, target images taken at the same moment from different perspectives of the to-be-acquired image area can be obtained to form a corresponding target image group. For example, for a key area in a chemical production site, such as a large equipment area, target images can be obtained from cameras set at multiple different angles at the same moment. In some examples, the number of target images and perspectives required can be determined according to the importance and complexity of the to-be-acquired image area. In some examples, the target image group can also be marked and classified. For example, the target image group can be classified and stored according to information such as the shooting angle and time of the image, facilitating quick retrieval and use.

[0086] As Figure 6 shown, at 610, according to the target image group, a three-dimensional model for the corresponding to-be-acquired image area is constructed.

[0087] In this embodiment, a three-dimensional reconstruction algorithm (such as structured light method, stereo vision method, etc.) can be used to fuse images from different perspectives in the target image group to construct a three-dimensional model. For example, the stereo vision method can be used to calculate the disparity of corresponding points in different images to restore the three-dimensional shape and spatial position of the object. In some examples, the constructed three-dimensional model can also be optimized and simplified to improve the processing efficiency and visualization effect. For example, unnecessary details and noises in the three-dimensional model are removed, and key structures and features are retained. In some examples, the three-dimensional model can also be rendered and annotated according to actual needs for better display and analysis. For example, key parts of the equipment are annotated, or different areas are color-coded to highlight specific information.

[0088] At 620, the spatial position state of the target entity in the three-dimensional model is determined.

[0089] In this embodiment, the spatial position state of the target entity can be determined by analyzing information such as the object coordinates and directions in the three-dimensional model. In some examples, the spatial position state can, for example, include the position coordinates and tilt angle of a chemical equipment in three-dimensional space. In some examples, historical data can also be combined to perform a trend analysis on the spatial position state of the target entity. For example, by comparing three-dimensional models at different time points, the change trend of the spatial position state of the chemical equipment is analyzed to predict possible problems in advance.

[0090] At 630, according to whether the determined spatial position state conforms to the corresponding 5S management rules, a spatial score corresponding to the target image group is generated.

[0091] In this embodiment, the rules of sorting, straightening, cleaning, standardizing, and discipline in the 5S management rules can be pre-converted into specific spatial position state indicators, and then the spatial score is generated according to the degree of compliance of the actual spatial position state of the target entity with these indicators. For example, if the position of the chemical equipment in the three-dimensional model is neat and the direction is consistent, meeting the requirements of straightening, a relatively high spatial score can be given. In some examples, different weights can be set according to different 5S management rules to reflect their importance in the spatial score. For example, in areas with high safety requirements in the chemical production site, the weights of the cleaning and discipline rules can be relatively high. In some examples, a specific analysis and feedback can be carried out on the spatial position state that does not conform to the 5S management rules, so as to take corresponding improvement measures. For example, if the cleaning score of a certain area is low, the specific dirty and messy areas and problems can be marked.

[0092] At 640, for each target image in the target image group, a feature extraction process is performed.

[0093] In this embodiment, the above feature extraction process can refer to the corresponding descriptions in steps 510-540 of the foregoing embodiment, and will not be elaborated here.

[0094] At 650, for each target image in the target image group, the extracted texture feature vector, layout feature vector, and color feature are compared with the corresponding features of the matching comparison reference image to obtain the texture feature comparison result, layout feature comparison result, and color feature comparison result of each target image.

[0095] The above step 650 can refer to the corresponding description in step 550 of the foregoing embodiment, and will not be elaborated here.

[0096] At 660, according to the fusion result among the texture feature comparison result, layout feature comparison result, and color feature comparison result of each target image in the target image group obtained, the feature comprehensive comparison result of the target image group is determined.

[0097] In this embodiment, the texture feature comparison result, layout feature comparison result, and color feature comparison result of each target image in the target image group can be fused by various methods (such as weighted summation, using machine learning models, etc.), and finally the feature comprehensive comparison result is obtained to evaluate the comprehensive similarity of the target image group and the corresponding comparison reference image group in terms of the extracted various features.

[0098] At 670, based on the determined comprehensive comparison result of features and the generated spatial score, determine the similarity between the target image group and the matching comparison reference image group.

[0099] In this embodiment, the comprehensive comparison result of features and the spatial score can be weighted and summed to obtain the similarity between the target image group and the comparison reference image group. In some examples, the weights of the comprehensive comparison result of features and the spatial score can be adjusted according to the requirements of the actual application scenario. For example, in the appearance inspection of the chemical production site, the weight of the comprehensive comparison result of features can be increased; while in the evaluation of the on-site layout and space management, the weight of the spatial score can be increased.

[0100] Through the above method, this solution can combine the three-dimensional model construction, feature extraction, and spatial score evaluation of the multi-perspective target image group to analyze the situation of the chemical production site more comprehensively and accurately.

[0101] Using Figures 1 - 6 the on-line inspection method for chemical production sites based on closed-circuit television systems disclosed in [reference], provides fully automated operations such as image acquisition, storage, comparison, and scoring in combination with the closed-circuit television system, realizes real-time and efficient inspection of chemical production sites, timely discovers potential problems and safety hazards, and improves the safety, standardization, and management efficiency of chemical production. At the same time, by comprehensively considering the texture, layout, color features of the images, as well as the spatial position state, and combining technical means such as calibration systems, it can further optimize and improve the on-line inspection of chemical production sites.

[0102] Figure 7 FIG. shows a block diagram of an example of an on-line inspection device 700 for chemical production sites based on a closed-circuit television system according to an embodiment.

[0103] The on-line inspection device 700 may include: an image acquisition module 710 configured to acquire target images captured for each image area to be collected at the chemical production site from the closed-circuit television system; an image storage module 720 configured to store each target image in a first path with its own name according to a first name designation rule and a first path designation rule; a scoring generation module 730 configured to generate an inspection score for each image area to be collected according to the comparison result between each target image and the matching comparison reference image, where the matching comparison reference image is determined according to the name of the target image; and a file storage module 740 configured to store each generated inspection score in a file in a second path accessible to the client with its own name according to a second name designation rule and a second path designation rule.

[0104] In addition, the on-line inspection device 700 may further include any other module configured to perform any operation of the on-line inspection method for the chemical production site based on the closed-circuit television system according to the above embodiments of the present disclosure.

[0105] The above has referred to Figures 1 to 7 , and the embodiments of the on-line inspection method and device for the chemical production site based on the closed-circuit television system according to the embodiments of the present specification have been described.

[0106] Figure 8 The signaling diagram shows an example of the interaction between various devices in the on-line inspection system 800 for the chemical production site according to the embodiments of the present specification.

[0107] As Figure 8 shown, at 810, the closed-circuit television system acquires and stores the images taken for each image area to be acquired at the chemical production site.

[0108] At 820, the server obtains the target images taken for each image area to be acquired at the chemical production site from the closed-circuit television system.

[0109] At 830, the server stores each target image in the first path with its respective name according to the first name specification rule and the first path specification rule.

[0110] At 840, the server generates an inspection score for each image area to be acquired according to the comparison result between each target image and the matching comparison reference image.

[0111] At 850, the server stores each generated inspection score in a file in the second path accessible to the client with its respective name according to the second name specification rule and the second path specification rule.

[0112] It should be noted that the above steps 810-850 may refer to the corresponding steps of the on-line inspection method for the chemical production site based on the closed-circuit television system described in the foregoing Figures 1 - 6 embodiments, and will not be elaborated here.

[0113] At 860, the client extracts the corresponding inspection scores from the respective files under the second path.

[0114] In this embodiment, the client can establish a communication connection with the above-mentioned server for storing inspection scores through a pre-configured network connection and corresponding permission verification mechanism, so as to directly access the files under the second path and extract each inspection score data according to the file format specification of the second path. In some examples, the client can preliminarily screen and sort the extracted inspection score data according to actual needs. For example, only extract the inspection scores corresponding to specific time periods or specific types of image areas to be collected for more efficient subsequent analysis and display.

[0115] At 870, the client determines the area identifier corresponding to each extracted inspection score according to the name-area identifier comparison table that matches the second name specification rule.

[0116] In this embodiment, a name-area identifier comparison table can be maintained inside the client. In some examples, the above-mentioned name-area identifier comparison table can be synchronously obtained from the server side during the system initialization phase and updated regularly. The area identifier corresponding to the file name that matches each extracted inspection score can be determined according to the name-area identifier comparison table. For example, the name of the inspection score file can contain specific coding fields, and the client can determine the corresponding area identifier according to the correspondence between the coding fields and area identifiers in the name-area identifier comparison table. In some examples, the name-area identifier comparison table can also have corresponding indexes. For example, establish indexes according to rules such as the first letter of the name, number range, etc., to accelerate the search process, especially when the data volume is large, it can significantly improve the speed of determining the area identifier.

[0117] At 880, the client presents the online inspection results in the form of area identifier-inspection score.

[0118] In this embodiment, the client can use an intuitive and clear visualization interface to display the online inspection results. For example, the area identifiers and the corresponding inspection scores can be listed one by one in a table form, and different color identifiers can be set according to the high and low of the inspection scores to facilitate managers to quickly understand the situation of each area. For example, green can be used to indicate a higher score and good situation, and red can be used to indicate a lower score and possible problems, etc.

[0119] In some examples, the client can also provide a chart display function. For example, a bar chart can be used to show the comparison of inspection scores in different areas, or a line chart can be used to reflect the change trend of inspection scores in a specific area at different time periods, so that managers can more intuitively understand the dynamic changes and overall situation of each area at the chemical production site.

[0120] In some examples, the client can implement an interaction function. When a management staff member clicks on a certain area identifier, more detailed image information, historical inspection data, etc. of that area can be further displayed, facilitating in-depth analysis of the problems existing in that area or observing the improvement effects.

[0121] Through the above method, this solution provides a complete solution for automating, informatizing, and visualizing the on-site inspection process in chemical production. By leveraging the image acquisition of the closed-circuit television system and the functions of image storage, comparison, and analysis of the server, efficient and accurate inspection scores can be generated. And through a series of data extraction, identifier matching, and result presentation operations on the client side, management staff can conveniently and intuitively grasp the actual situation of each area in the chemical production site, promptly discover potential problem areas, and make targeted decisions, effectively improving the safety, standardization, and overall management efficiency of chemical production, and ensuring the stable and orderly operation of chemical production.

[0122] The above references Figures 1 to 8 have described the embodiments of the on-site online inspection method, device, and on-site online inspection system for chemical production based on a closed-circuit television system according to the embodiments of this specification.

[0123] The on-site online inspection device for chemical production based on a closed-circuit television system according to the embodiments of this specification can be implemented in hardware, or can be implemented using software or a combination of hardware and software. Taking software implementation as an example, as a logically meaningful device, it is formed by the processor of its host device reading the corresponding computer program instructions in the memory into the memory for operation. In the embodiments of this specification, the on-site online inspection device for chemical production based on a closed-circuit television system can be implemented using an electronic device, for example.

[0124] Figure 9 FIG. shows a schematic diagram of an example of the on-site online inspection device 900 for chemical production based on a closed-circuit television system according to the embodiments of this specification.

[0125] As Figure 9 shown, the on-site online inspection device 900 for chemical production based on a closed-circuit television system may include at least one processor 910, a memory (e.g., non-volatile memory) 920, a memory 930, and a communication interface 940, and at least one processor 910, the memory 920, the memory 930, and the communication interface 940 are connected together via a bus 950. At least one processor 910 executes at least one computer-readable instruction stored or encoded in the memory (i.e., the above elements implemented in software form).

[0126] In one embodiment, computer-executable instructions are stored in a memory, which when executed cause at least one processor 910 to: obtain target images captured for respective image regions to be collected at a chemical production site from a closed-circuit television system; store each target image in a first path with its own name according to a first name specification rule and a first path specification rule; generate an inspection score for each image region to be collected according to a comparison result between each target image and a matching comparison reference image, where the matching comparison reference image is determined according to the name of the target image; and store each generated inspection score in a file in a second path accessible to a client with its own name according to a second name specification rule and a second path specification rule.

[0127] It should be understood that the computer-executable instructions stored in the memory, when executed, cause at least one processor 910 to perform the various operations and functions described above in the respective embodiments of this specification. Figures 1 - 6 described.

[0128] Specifically, a system or device equipped with a readable storage medium can be provided, on which software program code for implementing the functions of any one of the above embodiments is stored, and the computer or processor of the system or device is caused to read and execute the instructions stored in the readable storage medium.

[0129] In this case, the program code read from the readable medium itself can implement the functions of any one of the above embodiments, so the machine-readable code and the readable storage medium storing the machine-readable code constitute a part of the present invention.

[0130] The computer program code required for the operations of each part of this specification can be written in any one or more programming languages, including object-oriented programming languages such as Java, Scala, Smalltalk, Eiffel, JADE, Emerald, C++, C#, VB, NET, and Python, conventional procedural programming languages such as C, Visual Basic 2003, Perl, COBOL 2002, PHP, and ABAP, dynamic programming languages such as Python, Ruby, and Groovy, or other programming languages. The program code can run on a user's computer, or run as a stand-alone software package on a user's computer, or part on a user's computer and part on a remote computer, or all on a remote computer or server. In the latter case, the remote computer can be connected to the user's computer in any network form, such as a local area network (LAN) or a wide area network (WAN), or connected to an external computer (for example, via the Internet), or in a cloud computing environment, or used as a service, such as software as a service (SaaS).

[0131] Examples of readable storage media include floppy disks, hard disks, magneto-optical disks, optical disks (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD-RW), magnetic tapes, non-volatile memory cards, and ROMs. Optionally, program code can be downloaded from a server computer or the cloud via a communication network.

[0132] The above describes specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the acts or steps recited in the claims can be performed in a different order than in the embodiments and still achieve the desired result. Additionally, the processes depicted in the figures do not necessarily require the particular order or sequential order shown to achieve the desired result. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0133] Not all steps and units in the above-mentioned processes and system structure diagrams are necessary, and certain steps or units can be omitted according to actual needs. The execution order of each step is not fixed and can be determined as required. The device structures described in the above embodiments can be physical structures or logical structures, that is, some units may be implemented by the same physical entity, or some units may be implemented separately by multiple physical entities, or some components in multiple independent devices may be jointly implemented.

[0134] The term "exemplary" used throughout this specification means "serving as an example, instance, or illustration" and does not mean "preferred" or "advantageous" over other embodiments. For the purpose of providing an understanding of the described technology, the specific embodiments include specific details. However, these technologies can be implemented without these specific details. In some instances, well-known structures and devices are shown in block diagram form to avoid obscuring the concepts of the described embodiments.

[0135] The optional embodiments of the embodiments of this specification have been described in detail above in conjunction with the accompanying drawings. However, the embodiments of this specification are not limited to the specific details in the above embodiments. Within the technical concept scope of the embodiments of this specification, various simple modifications can be made to the technical solutions of the embodiments of this specification, and these simple modifications all fall within the protection scope of the embodiments of this specification.

[0136] The foregoing description of the content of this specification is provided to enable any ordinary person skilled in the art to implement or use the content of this specification. Various modifications to the content of this specification will be apparent to those of ordinary skill in the art, and the general principles defined herein can also be applied to other variations without departing from the scope of protection of the content of this specification. Therefore, the content of this specification is not limited to the examples and designs described herein, but is consistent with the broadest scope that conforms to the principles and novel features disclosed herein.

Claims

1. A chemical production site online inspection method based on a closed-circuit television system, comprising: Acquire target images for each image area to be collected at the chemical production site from the closed-circuit television system; According to the first name specifying rule and the first path specifying rule, each target image is stored in the first path with its own name; Generating an inspection score for each image area to be collected according to a comparison result between each target image and a matched comparison reference image, wherein the matched comparison reference image is determined according to the name of the target image; and According to the second name specifying rule and the second path specifying rule, each generated inspection score is stored with a respective name in a file of the second path accessible to the client.

2. The online inspection method according to claim 1, wherein: The method of acquiring target images for each image area to be collected at the chemical production site from the closed-circuit television system includes: By simulating mouse operation, the pictures taken for each image area to be collected at the chemical production site are sequentially retrieved from the closed-circuit television system and image capture is performed, wherein the mouse operation includes at least one of the following: moving, clicking, and scroll wheel operation.

3. The online inspection method according to claim 1, wherein: The step of generating an inspection score for each image area to be collected according to the comparison result between each target image and the matched comparison reference image includes: Determining the similarity between each target image and the matched comparison reference image; Calibrate the determined similarities using calibration coefficients corresponding to each target image to obtain a calibrated similarity score, wherein the calibration coefficients are determined by fitting the similarities for the historical images and the corresponding benchmark inspection scores; and According to each similarity score, a corresponding inspection score is generated.

4. The online inspection method according to claim 3, wherein: The online inspection method further comprises: For each target image, the sensor data of each sensor in the corresponding image area to be collected at the time of shooting the target image is collected. Generating corresponding inspection scores according to the similarity scores includes: According to each similarity score and the corresponding sensor data, a corresponding inspection score is generated.

5. The online inspection method according to claim 4, wherein: The comparison reference image has a corresponding shooting time, and acquiring the target image shot for the image area to be collected at the chemical production site is executed in response to the current time being consistent with the shooting time of the matching comparison reference image, or Acquiring a target image captured for an image region to be captured at a chemical production site is performed in response to sensor data in the image region to be captured indicating the presence of an abnormality.

6. The online inspection method according to any one of claims 3 to 5, wherein: The inspection score includes a 5S management score, and the determination of the similarity between each target image and the matched comparison reference image includes: For each target image, Extracting local texture features from each local area of ​​the target image; The extracted local texture features are fused to obtain the texture feature vector of the target image; Generating a layout feature vector of the target image according to the geometric shape detected from the target image; Extracting color features of the target image; Comparing the extracted texture feature vector, layout feature vector and color feature with corresponding features of the matched comparison reference image to obtain a texture feature comparison result, a layout feature comparison result and a color feature comparison result; and The similarity between the target image and the matched comparison reference image is determined based on the fusion result among the obtained texture feature comparison result, the layout feature comparison result and the color feature comparison result.

7. The online inspection method according to claim 6, wherein: The target image includes target image groups corresponding to respective image regions to be collected of a predetermined type, and each target image group includes target images photographed at different viewing angles at the same time for the image regions to be collected of the predetermined type. Determining the similarity between each target image and the matched comparison reference image also includes: According to each target image group, construct a three-dimensional model for each corresponding image area to be collected; For each three-dimensional model, determine the spatial position state of the target entity in the three-dimensional model; and generate a spatial score corresponding to the target image group according to whether the determined spatial position state complies with the corresponding 5S management rules, Determining the similarity between the target image and the matched comparison reference image according to the fusion result among the obtained texture feature comparison result, the layout feature comparison result and the color feature comparison result comprises: For each target image group, Determining a comprehensive feature comparison result of the target image group according to a fusion result of the texture feature comparison result, the layout feature comparison result, and the color feature comparison result of each target image in the target image group; and The similarity between the target image group and the matched comparison reference image group is determined based on the determined feature comprehensive comparison result and the generated spatial score.

8. An online inspection device for chemical production site based on a closed-circuit television system, comprising: An image acquisition module is configured to acquire target images for each image area to be collected at the chemical production site from the closed-circuit television system; An image storage module is configured to store each target image in a first path with its own name according to a first name specifying rule and a first path specifying rule; A score generating module is configured to generate an inspection score for each image area to be collected according to a comparison result between each target image and a matched comparison reference image, wherein the matched comparison reference image is determined according to the name of the target image; as well as The file storage module is configured to store the generated inspection scores with respective names in files of the second path accessible to the client according to the second name designation rule and the second path designation rule.

9. A chemical production site online inspection system, comprising: A closed-circuit television system for collecting and storing images taken of various image areas to be collected at the chemical production site; A server, used to execute the on-line inspection method for chemical production site based on a closed-circuit television system as described in any one of claims 1 to 7; as well as The client is used to extract the corresponding inspection score from each file under the second path; determine the area identifier corresponding to each extracted inspection score according to the name-area identifier comparison table matching the second name specification rule; and present the online inspection result represented in the form of area identifier-inspection score.

10. A chemical production site online inspection device based on a closed-circuit television system, comprising: At least one processor, a memory coupled to the at least one processor, and a computer program stored in the memory, wherein the at least one processor executes the computer program to implement the on-line inspection method for chemical production sites based on a closed-circuit television system as described in any one of claims 1 to 7.

Images