Visual monitoring system and method for storage tank cleaning process and cleaning gun

Abstract

The invention provides a visual monitoring system and method for a storage tank cleaning process and a cleaning gun, and the system comprises a front-end device and an image capturing and collecting unit which is used for receiving a preprocessing image of a monitored region and comprises the monitored region and a combination body combined with the monitored region; the first monitoring area is used for determining target area information of an assembly combined with the monitoring equipment according to the preprocessed image and a primary preprocessing model; and the second monitoring area is used for preprocessing the image and determining a reference coordinate point of the monitored area according to the secondary preprocessing model, and determining a reference coordinate line of the monitored area according to the reference coordinate point and the area information of the target. According to the scheme, the moving and cleaning speed of the cleaning gun or the nozzle can be determined, and the moving and cleaning speed of the cleaning gun or the nozzle can be determined through the relative displacement of the cleaning gun or the nozzle. And the cleaning speed of the cleaning gun or the nozzle can still be accurately determined, so that the operation process of the cleaning gun or the nozzle can be monitored.

Description

Technical Field

[0001] The invention belongs to the field of mechanical cleaning, environmental protection and monitoring technology of storage tanks, and in particular relates to a visual monitoring system and method for a storage tank cleaning process and a cleaning gun. Background Art

[0002] During the process of collecting, storing and refining, oil carries a large amount of sludge. During the storage of oil, a black and thick gelatinous material layer will form on the bottom and wall of the tank. If this material is not removed in time, it will have an adverse effect on the quality of the oil in the tank and the effective storage volume of the tank, and accelerate the corrosion of the tank bottom and tank wall, thereby reducing the service life of the tank. As a technology widely used in the cleaning industry, the mechanical cleaning and sludge reduction technology of petroleum tanks has been regulated by existing technical specifications. However, it is often difficult to achieve the ideal working state during the actual operation process, and due to the inherent characteristics of the tank, the operator cannot always visually observe the cleaning process and the movement of the cleaning gun or nozzle, resulting in the cleaning process not achieving satisfactory results, resulting in unsatisfactory cleaning effects and affecting the efficiency of device cleaning. Summary of the invention

[0003] In view of this, the present invention aims to propose a visual monitoring system and method for a tank cleaning process and a cleaning gun, so as to solve the technical problem that the cleaning process cannot be effectively remotely monitored during the operation of a manual cleaning gun or a cleaning nozzle.

[0004] To achieve the above object, the technical solution of the present invention is achieved as follows:

[0005] A visual monitoring method for a tank cleaning process and a cleaning gun, comprising:

[0006] Receiving a pre-processed image sent by a front-end device, including a monitored area and a combination combined with the monitored area;

[0007] Determine target area information of a combination combined with the monitoring device according to the preprocessed image and the primary preprocessing model;

[0008] Determine the reference coordinate points of the monitored area based on the preprocessed image and the secondary preprocessing model, and determine the reference coordinate lines of the equipment to be monitored based on the reference coordinate points and the target area information;

[0009] According to the preprocessed image, the reference coordinate line and the three-dimensional preprocessing model, the reference coordinate information on the reference coordinate line is determined, and according to the reference coordinate information and the target area information, the positioning information of the monitored area is determined;

[0010] The relative displacement speed between the monitored area and the assembly is determined based on a plurality of positioning information in a plurality of pre-processed images of the monitored area and the corresponding time information and model.

[0011] Furthermore, before the preprocessed image and the primary preprocessed model determine the target area information of the combination combined with the monitoring device, it also includes:

[0012] Acquire reference image data; the reference image data includes reference target and reference region information combined with the reference target, the reference target includes a preprocessed image and a preprocessed model;

[0013] Acquire a first initial capture processing model, the first initial capture processing model being the regional structure and range formed within the cleaning range, and determine initial region information in the reference video data according to the first initial capture module and the reference image data;

[0014] According to the initial region information and the reference region information, a processing model is modified to determine a preprocessing model.

[0015] Furthermore, before the preprocessed image and the secondary preprocessed model determine the reference coordinate points of the monitored area, and before determining the reference coordinate lines of the system to be monitored according to the reference coordinate points and the area information of the target, the method further includes:

[0016] obtaining the secondary processing model;

[0017] Obtaining first initial captured image data, wherein the first initial captured image includes capture area information of a capture assembly and a capture element combination in the captured video data; the first captured image data includes capture coordinate points and capture coordinate lines;

[0018] Determine the initial coordinate point of the capture device according to the first captured image data and the second capture processing model, and determine the initial coordinate line of the capture device according to the initial coordinate point and the capture area information;

[0019] The secondary processing model is corrected according to the initial coordinate points, the initial coordinate lines, the captured coordinate points and the captured coordinate lines to determine the secondary preprocessing model.

[0020] Furthermore, before determining the reference coordinate information on the reference coordinate line using the preprocessed image, the reference coordinate line and the three-dimensional preprocessing model, and determining the positioning information of the monitored area according to the reference coordinate information and the target area information, the method further includes:

[0021] obtaining the three-times processing model;

[0022] Obtaining secondary reference image data of reference coordinate points and reference coordinate lines in the primary reference image data; the secondary reference image data includes reference positioning information of the reference device;

[0023] Determine initial coordinate information of the reference coordinate line according to the secondary reference image data, the reference coordinate line and the three-dimensional processing model, and determine initial positioning information of the reference target according to the initial coordinate information and the reference area information;

[0024] According to the initial positioning information and the reference positioning information, the tertiary processing model is modified to determine the tertiary preprocessing model.

[0025] Furthermore, the step of determining the target area information of the assembly combined with the monitoring device using the preprocessed image and the primary preprocessing model comprises:

[0026] Step A1, determining an image of the assembly according to the preprocessed image and the primary preprocessing model;

[0027] Step A2: searching for a preprocessing monitoring device located outside the assembly according to the image of the assembly and the preprocessing image;

[0028] Step A3, when searching for the monitored area outside the assembly, determining the boundary between the assembly and the monitored area;

[0029] Step A4: determining target area information combining the assembly and the monitored area based on the image of the assembly and the boundary between the assembly and the monitored area.

[0030] Furthermore, the steps of determining the reference coordinate points of the monitored area using the preprocessed image and the secondary preprocessing model, and determining the reference coordinate line of the system to be monitored according to the reference coordinate points and the target area information include:

[0031] Step B1, determining the two sides of the monitored area according to the preprocessed image and the secondary preprocessing model;

[0032] Step B2, determining the parallel portion between the two sides of the monitored area according to the two sides of the monitored area and the secondary preprocessing model;

[0033] Step B3, determining the reference coordinate points of the monitored area according to each parallel component and the secondary preprocessing model;

[0034] Step B4: Determine the reference coordinate line of the monitored area according to the reference coordinate point and area information of the target.

[0035] Furthermore, the step of determining the reference coordinate points of the monitored area by each parallel component and the secondary preprocessing model includes:

[0036] Step C1, determining a vertical line of each parallel part according to each parallel part and the secondary preprocessing model;

[0037] Step C2, determining the midpoint of each vertical line according to the vertical lines of the parallel parts and the secondary preprocessing model;

[0038] Step C3: Determine the reference coordinate point of the area to be monitored according to the midpoint of each vertical line.

[0039] Furthermore, the steps of determining reference coordinate information on the reference coordinate line using the preprocessed image, the reference coordinate line and the three-dimensional preprocessing model, and determining the location information of the monitored area according to the reference coordinate information and the target area information include:

[0040] Step D1, determining reference coordinate information on the reference coordinate line according to the preprocessed image, the reference coordinate line and the three-dimensional preprocessing model;

[0041] Step D2: when there are several reference coordinate information on the reference coordinate line, determine the distance between each reference coordinate information and the target area information according to the target area information;

[0042] Step D3: Select the reference coordinate information closest to the target edge from each reference coordinate information, and determine the positioning information of the area to be monitored according to the reference coordinate values ​​closest to the target edge and the target edge.

[0043] Furthermore, the step of determining the relative displacement speed between the monitored area and the assembly based on the plurality of positioning information of the monitored area in the plurality of preprocessed images, the corresponding time of each preprocessed image information and the simulation model thereof comprises:

[0044] Step E1, determining adjacent positioning information of the monitored device according to a plurality of positioning information of the monitored device in a plurality of pre-processed images;

[0045] Step E2: determining the relative displacement speed between the monitored device and the combination according to the adjacent positioning information of the monitored device, the time information of each processed image of the adjacent positioning information contained in the processed image, and the simulation model.

[0046] Furthermore, the present invention discloses a visual monitoring system for a tank cleaning process and a cleaning gun, comprising:

[0047] Using the front-end device and the image capture and acquisition unit to receive a pre-processed image of the monitored area, the pre-processed image includes the monitored area and a combination combined with the monitored area;

[0048] A first monitoring area, for use in preprocessed images and a primary preprocessing model to determine target area information of a combination combined with a monitoring device;

[0049] The second monitoring area is used to preprocess the image and the secondary preprocessing model to determine the reference coordinate points of the monitored area, and determine the reference coordinate line of the monitored area according to the reference coordinate points and the area information of the target;

[0050] The third monitoring area is used to preprocess the image, the reference coordinate line and the three-dimensional preprocessing model to determine the reference coordinate information on the reference coordinate line, and determine the positioning information of the monitored area according to the reference coordinate information and the area information of the target;

[0051] The identification and analysis unit is used to determine the relative displacement speed between the monitored area and the assembly based on the positioning information in the processed images of the monitored area, the corresponding time information of each processed image and the simulation model.

[0052] Compared with the prior art, the visual monitoring system for the tank cleaning process and the cleaning gun described in the present invention has the following advantages:

[0053] (1) The visual monitoring system of a tank cleaning process and a cleaning gun described in the present invention, after receiving the preprocessed image information sent by the front-end device, the system determines the target area information of the assembly combined with the preprocessing equipment based on the preprocessed image and the primary preprocessing model. Then, the reference coordinate point of the preprocessing equipment is determined based on the preprocessing image and the secondary preprocessing model, and the reference coordinate line of the preprocessing device is determined based on the reference coordinate point and the target area information. Furthermore, the reference coordinate information on the reference coordinate line is determined based on the preprocessed image and the tertiary preprocessing model, and the positioning information of the monitored area is determined based on the reference coordinate information and the target area information. Then, the relative speed between the monitored area and the assembly is determined based on the multiple positioning information of the monitored area in the multiple preprocessed images, the time message corresponding to each preprocessed image information, and the simulation model;

[0054] (2) A visual monitoring system for a tank cleaning process and a cleaning gun according to the present invention includes a preprocessed image of a device to be monitored during operation, such as a preprocessed image during the cleaning process. By using the preprocessed image, as well as the primary preprocessing model, the secondary preprocessing model, and the tertiary preprocessing model, the target area information of the combination with the monitored device, the reference coordinate line of the monitored device, and the position information of the monitored device can be gradually determined. By monitoring the position information of the device at different times and the simulation model, the relative displacement speed between the monitoring device and the combination can be determined. That is, during the cleaning process, the cleaning speed of the cleaning gun or the cleaning nozzle can be determined, and the cleaning speed of the cleaning gun can be determined by the relative displacement of the cleaning gun or the cleaning nozzle. Even if jitter occurs during the cleaning process of the cleaning gun or the cleaning nozzle, the cleaning speed of the cleaning gun or the cleaning nozzle can still be accurately determined, thereby realizing the monitoring of the cleaning speed of the cleaning gun or the cleaning nozzle. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] The accompanying drawings constituting a part of the present invention are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the accompanying drawings:

[0056] Figure 1 It is a schematic diagram of the system composition of the present invention;

[0057] Figure 2 A first flow chart of the method for using the system of the present invention;

[0058] Figure 3 Another flowchart of the method for using the system of the present invention is provided;

[0059] Figure 4 A schematic diagram of an assembly and a monitored area in a method for using the system of the present invention;

[0060] Figure 5 It is a schematic diagram of the structure of the system of the present invention;

[0061] Figure 6 It is a schematic diagram of the composition of the workstation of the present invention. DETAILED DESCRIPTION

[0062] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other.

[0063] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and the like are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, features defined as "first", "second", and the like may explicitly or implicitly include one or more of the features. In the description of the present invention, unless otherwise specified, "multiple" means two or more.

[0064] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood by specific circumstances.

[0065] The present invention will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0066] Example 1, reference Figure 1 The present invention provides a monitoring system, which includes a front-end device 1 and a workstation 2. The front-end device 1, the workstation 2 and the front-end device 1 and the workstation 2 are connected by communication. The front-end device 1 includes but is not limited to an embedded high-definition camera device, a high-definition camera device, a mobile terminal, etc., and is used to capture images of the monitoring target.

[0067] The workstation 2 includes a local workstation and / or a remote workstation. The workstation 2 receives the pre-processed image information sent by the front-end device 1, which mainly includes:

[0068] The preprocessed image and the primary preprocessing model determine the target area information of the combination combined with the monitoring area;

[0069] The preprocessed image and the secondary preprocessing model determine the reference coordinate points of the monitored area, and determine the reference coordinate line of the monitored area according to the reference coordinate points and the area information of the target;

[0070] The preprocessed image, the reference coordinate line and the three-dimensional preprocessing model determine the reference coordinate information on the reference coordinate line, and determine the positioning information of the monitored area according to the reference coordinate information and the target area information;

[0071] The relative speed between the monitored area and the assembly is determined according to a plurality of positioning information of the monitored area in a plurality of processed images, time information corresponding to each processed image and a simulation model.

[0072] Example 2, reference Figure 2 , a first flow chart of the device and its use method provided by the present invention, the device and its use method include receiving an image sent by a front-end device; pre-processing the image includes a monitored area and a combination combined with the monitored area.

[0073] Furthermore, in order to monitor the working process of the cleaning gun and the assembly, it is necessary to take photos or videos of the working process of the monitoring equipment and tools. The front-end device can take photos or videos of the monitoring equipment and the assembly to obtain corresponding monitoring video data or photo data, that is, the image to be processed. After obtaining the image corresponding to the monitoring equipment and the assembly, the front-end device can store the image or directly send the image to the workstation for storage.

[0074] Furthermore, in order to ensure the stable operation of the front-end device and the cleaning of image capture, a self-cleaning device is integrated in the front-end device. According to the cleanliness signal feedback of the image capture, it is compared with the reserved cleaning image, and then the self-cleaning device is started to clean the image capturer of the front-end device to ensure the cleanliness of the transmitted image.

[0075] Furthermore, the workstation receives image information sent by the front-end device, and the image information can be directly obtained after the workstation communicates with the front-end device, or the image is automatically sent to the workstation.

[0076] Furthermore, each preprocessed image includes a single image for video data capture or a dynamic image for video data segmentation; the preprocessed image may be an enlarged or reduced image, an image with contrast adjusted, etc., or may be an original image.

[0077] Further, the equipment to be monitored includes cleaning guns or cleaning nozzles.

[0078] Furthermore, the assembly includes a tank wall and tank wall sludge.

[0079] Further, according to the preprocessed image and the first preprocessing model, target area information to be combined with the monitoring device is determined.

[0080] Furthermore, the primary preprocessing model refers to a model for identifying target area information combined with the monitored area. This model can be obtained through deep learning. After inputting the preprocessing image, the primary preprocessing model can determine the composite image from the preprocessing image, and determine the target area information of the composite combined with the monitoring device in the composite image.

[0081] Further, taking the cleaning process as an example, considering that the actual moving speed of the cleaning gun or nozzle during the working process determines the cleaning effect, it is necessary to determine the residence time of each part of the cleaning gun or cleaning nozzle relative to the tank wall to determine the cleaning effect on each part of the tank wall. Therefore, by determining the target area information combined with the monitored area, the moving speed of the monitored area can be determined according to the target area information. Taking the cleaning gun or cleaning nozzle as an example, the relative displacement speed of the cleaning gun or cleaning nozzle and the tank wall can be determined; at this point, the target edge refers to the edge of the tank wall connected to the cleaning gun or nozzle, that is, the part of the tank wall that is not connected to the cleaning gun or nozzle, and the connecting part of the tank wall connected to the cleaning gun or nozzle.

[0082] Furthermore, before the step of determining the target area information combined with the monitoring device through the preprocessed image and the primary preprocessing model, it also includes: acquiring reference video data; the reference video data includes reference area information combined with a reference tool and a reference device; obtaining a primary processing model, and determining the initial area information in the reference video data based on the first initial capture module and the reference video data; and correcting the primary processing model based on the initial area information and the reference area information to determine a primary preprocessing model.

[0083] Further, before using the primary preprocessing model to identify the edge of the target, it is necessary to train the primary processing model to obtain the first preprocessing mode after deep learning, so as to accurately determine the edge of the target combined with the combination and monitoring device. That is, by using the reference video data to train and compare the initial area information determined by the reference area information and the primary processing model, when there is a difference between the initial area information and the reference area information, the first initial processing mode is modified to obtain the primary preprocessing model.

[0084] Furthermore, in order to make the preprocessing model have higher precision and accuracy for the target area information, the image can be enlarged. This step includes:

[0085] Step A100: based on the preprocessed image and the primary preprocessing model, amplify the portion of the image that is combined with the assembly and the monitoring device, and determine the area information of the target that is combined with the assembly and the monitoring device from the assembly.

[0086] Step A200: After obtaining the preprocessed image, when it is necessary to use a preprocessing model to determine target area information combined with the monitoring device, the preprocessed image can be processed;

[0087] Furthermore, by using a primary preprocessing model to identify the target and magnifying the portion where the tool is combined with the monitored device, target area information of the tool combined with the monitored device can be accurately found to determine the relative speed between the monitored device and the tool.

[0088] Further, when training the one-time processing model, the first initial capture module can first identify and magnify the part where the target is combined with the monitored area, and then determine the initial area information in the reference video data. This allows the one-time processing model to predetermine the part of the tool and the degree of magnification required during training. Therefore, when the one-time preprocessing model is used for processing in the future, the target area information combining the target with the monitored area can be accurately found.

[0089] Furthermore, if the image magnification is too large or too small, the recognition accuracy may be reduced. Therefore, when training the primary processing model, the required image magnification can be determined to make the target area information determined by the primary preprocessing model and combined with the monitoring device more accurate.

[0090] Further, the step of determining target area information of the assembly combined with the monitoring system according to the preprocessed image and the primary preprocessing model includes:

[0091] Step B100, determining an image of the assembly according to the preprocessed image and the first segmentation model;

[0092] Step B200, searching for a preprocessing monitoring device located outside the assembly based on the image of the assembly and the preprocessing image; when searching for a monitored area outside the assembly, determining the boundary information between the assembly and the monitored area;

[0093] Step B300: Determine target area information combining the assembly and the monitored area according to the image of the assembly and the boundary information of the assembly and the monitored area.

[0094] Furthermore, when a preprocessing model is used to determine the target area information of the combination and the monitored area, the image of the combination can be determined first, and then the monitored device can be found from the outer area of ​​the image of the combination. After finding the monitored area, the boundary area between the combination and the monitored area can be determined. In order to accurately locate the combination and the monitored device, the image information of the boundary area is amplified to determine the accurate information of the target area of ​​the combination and the monitoring device.

[0095] Furthermore, the reference coordinate points of the monitored area are determined according to the preprocessed image and the secondary preprocessing model, and the reference coordinate lines of the monitored area are determined according to the reference coordinate points and the target area information.

[0096] Further, when all preprocessed images include target area information combined with the combination and monitoring device, the secondary preprocessing model processes all preprocessed images. When some of the preprocessed images include target area information combined with the combination or monitoring device, the secondary preprocessing model processes the preprocessed images including target edges combined with the combination / monitoring device to avoid processing unprocessed images without target edges and improve efficiency.

[0097] Furthermore, when the monitored device is combined with the target, it is necessary to determine the reference coordinate point of the monitored device according to the application, and at the same time determine the reference coordinate line of the monitored device by the reference coordinate point and the edge of the target, so as to accurately monitor the moving distance of the monitored device and determine the moving speed of the monitored device.

[0098] Furthermore, the reference coordinate point of the monitored area may be a point on the central axis of the monitored area, and the corresponding reference coordinate line is the central axis of the monitored area. Taking the cleaning gun or cleaning nozzle as an example, there may be a curved part in the cleaning gun or cleaning nozzle. According to the length of the central axis of the cleaning gun or cleaning nozzle, the actual moving distance of the cleaning gun or cleaning nozzle in the tank wall is more accurate. Therefore, it is necessary to first determine the central axis of the cleaning gun or cleaning nozzle, and then determine the central axis of the cleaning gun or cleaning nozzle, so that the actual moving distance of the cleaning gun or cleaning nozzle can be accurately determined.

[0099] Example 3, reference Figure 3 , a second flow chart of the device provided by the present invention and its method of use, the device and its method of use include receiving an image sent by a front-end device; the pre-processed image includes a cleaning gun or a cleaning nozzle and a tank wall where the cleaning gun or the cleaning nozzle is located.

[0100] Furthermore, the detection process of the tank wall is captured by a camera device to obtain a pre-processed image, so that the pre-processed image includes the cleaning gun or the cleaning nozzle and the tank wall where the cleaning gun or the cleaning nozzle is located.

[0101] Furthermore, according to the preprocessed image and the primary preprocessing model, the area information on the tank wall combined with the cleaning gun or the nozzle is determined.

[0102] Further, after obtaining the pre-processed image, the image of the location of the tank wall can be captured first. Then, the edge of the tank wall can be determined based on the portion of the tank wall that is combined with the cleaning gun or nozzle. For example, if the cleaning gun or nozzle usually enters from the top access hole, the top access hole edge can be determined.

[0103] Furthermore, the central axis of the cleaning gun or nozzle is determined according to the preprocessed image and the secondary preprocessed model, and the central axis of the cleaning gun or nozzle is determined according to the central axis of the cleaning gun or nozzle and the edge of the tank wall.

[0104] Specifically, after determining the tank wall area information combined with the cleaning gun or cleaning nozzle, the pre-processed image is captured to obtain an image of the cleaning gun or cleaning nozzle. Due to the width of the cleaning gun or cleaning nozzle, the central axis point of the cleaning gun or cleaning nozzle is determined, and then the central axis point of the cleaning gun or cleaning nozzle is connected to the tank wall area information to determine the central axis of the cleaning gun or flushing nozzle.

[0105] Further, the coordinate value on the central axis is determined according to the preprocessed image, the central axis and the cubic preprocessing model, and the positioning information of the cleaning gun or the nozzle is determined according to the coordinate value and the area information.

[0106] Furthermore, after determining the central axis of the cleaning gun or the cleaning nozzle, the image of the cleaning gun or the cleaning nozzle is captured, and the coordinate values ​​on the cleaning gun or the cleaning nozzle are found. According to the coordinate values ​​on the cleaning gun or the cleaning nozzle, the coordinate values ​​closest to the edge of the tank wall are found and the positioning information of the cleaning gun or the cleaning nozzle is determined.

[0107] Furthermore, the relative speed between the cleaning gun or the cleaning nozzle and the tank wall is determined according to the positioning information of the cleaning gun or the cleaning nozzle in the preprocessed images, the monitoring information corresponding to each preprocessed image and the simulation model.

[0108] Furthermore, after determining several positioning information of the cleaning gun or the cleaning nozzle, the cleaning speed of the cleaning gun and the cleaning nozzle can be determined through time learning, positioning learning and simulation model to determine the cleaning effect and take corresponding measures.

[0109] Example 4, reference Figure 4 , with the tank wall 41 as a combination and the cleaning gun or cleaning nozzle 42 as a monitoring system. After determining the target area information 411 of the tank wall 41 combined with the cleaning gun or cleaning nozzle 42 through the processed image and the secondary preprocessing model, the reference coordinate point 421 of the cleaning gun or cleaning nozzle 42 is determined according to the second capture mode. Then, based on the reference coordinate point 421 and the target area information 411, finally, the reference coordinate line 422 of the cleaning gun or nozzle is determined.

[0110] Furthermore, the method for determining the reference coordinate point includes using a geometric modeling transverse diameter measurement method to determine the reference coordinate point and obtaining the reference coordinate point through an algorithm.

[0111] Further, a secondary preprocessing model may be used to capture an image of a cleaning gun or a cleaning nozzle from a preprocessed image, and the image may be processed to obtain a contour map thereof, and then reference coordinate points may be obtained to obtain reference coordinate lines.

[0112] Furthermore, before the preprocessed image and the secondary preprocessed model determine the reference coordinate points of the monitored area, and before determining the reference coordinate lines of the system to be monitored according to the reference coordinate points and the target area information, further steps include:

[0113] Step C100, obtaining a secondary processing model;

[0114] Step C200, obtaining primary reference image data, wherein the primary reference image includes reference region information in the reference video data combined with a reference assembly and a reference device;

[0115] Step C300, the primary reference image data includes reference coordinate points and reference coordinate lines;

[0116] Step C400, determining the initial coordinate point of the reference device according to the primary reference image data and the secondary processing model, and determining the initial coordinate line of the reference device according to the initial coordinate point and the reference area information;

[0117] Step C500: modify the secondary processing model according to the initial coordinate points, the initial coordinate lines, the reference coordinate points and the reference coordinate lines to determine the secondary preprocessing model.

[0118] Furthermore, before using the secondary preprocessing model to process the preprocessed image, it is necessary to first train the secondary processing model. In order to improve the training efficiency, the primary reference image data is directly trained by combining the reference area information with the reference combination and the reference device. By using the primary reference image data and the secondary processing model, the initial coordinate points and the initial coordinate lines are obtained. Then, the initial coordinate points, the initial coordinate lines, the reference coordinate points and the reference coordinate lines are compared, and the second initial segmentation model is corrected to obtain the trained secondary preprocessing model.

[0119] Furthermore, the steps of determining the reference coordinate points of the monitored area according to the preprocessed image and the secondary preprocessing model, and determining the reference coordinate line of the monitored system according to the reference coordinate points and the target area information include: determining the two sides of the monitored area according to the preprocessed image and the secondary preprocessing model; determining the parallel parts between the two sides of the monitored area according to the two sides of the monitored area and the secondary preprocessing model; determining the reference coordinate points of the monitored area according to each parallel part and the secondary preprocessing model; and determining the reference coordinate line of the monitored area according to the reference coordinate points and area information of the target.

[0120] Further, based on the preprocessed image information and the secondary preprocessed model, two side edges of the cleaning gun or cleaning nozzle 42 can be determined. A number of parallel parts, i.e., a number of parallel components, can be determined based on the two side edges of the cleaning gun or cleaning nozzle 42 and the secondary preprocessed model. Then, the reference coordinate point 421 of the cleaning gun or cleaning nozzle is determined. For example, if Figure 4If the cleaning gun or cleaning nozzle 42 in the monitoring area has three parallel parts, several reference coordinate points can be obtained by processing these three parallel parts. In this way, the reference coordinate line of the monitored area can be determined according to this area information.

[0121] Further, the step of determining the reference coordinate points of the monitored area according to each parallel component and the secondary preprocessing model includes: determining the vertical line of each parallel component according to each parallel part and the second segmented model; determining the midpoint of each vertical line according to the vertical line of the parallel part and the secondary preprocessing model; and determining the reference coordinate points of the monitored area according to the midpoint of each vertical line.

[0122] Further, the method for determining the reference coordinate point includes determining a perpendicular line of the parallel parts on both sides, and determining the midpoint of the perpendicular line through a secondary preprocessing model. The midpoint of the perpendicular line is used as the reference coordinate point, and then the reference coordinate point is connected and connected to the target area information to determine the reference coordinate line.

[0123] Furthermore, reference coordinate information on the reference coordinate line is determined according to the preprocessed image, the reference coordinate line and the third preprocessing model, and positioning information of the monitored device is determined according to the reference coordinate information and the target edge.

[0124] Further, after determining the reference coordinate line, before determining the location of the monitored device, it is necessary to determine the reference coordinate information to determine the moving distance of the monitored area. At this time, the three-time preprocessing model is used to determine the reference coordinate information, thereby determining the location information of the monitored area.

[0125] Furthermore, before determining the reference coordinate information on the reference coordinate line according to the preprocessed image, the reference coordinate line and the three-dimensional preprocessing model, and determining the positioning information of the monitored device according to the reference coordinate information and the target area information, further steps include:

[0126] Step D100, obtaining the three-time processing model;

[0127] Step D200, obtaining secondary reference image data including reference coordinate points and reference coordinate lines in the primary reference image data;

[0128] Step D300, the secondary reference image data includes reference positioning information of the reference device;

[0129] Step D400, determining initial coordinate information of the reference coordinate line according to the secondary reference image data, the reference coordinate line and the third-order processing model, and determining initial positioning information of the reference device according to the initial coordinate information and the reference area information;

[0130] Step D500: modify the tertiary processing model according to the initial positioning information and the reference positioning information, and determine the tertiary preprocessing model.

[0131] Furthermore, before using the three-dimensional preprocessing model to process the preprocessed image, the three-dimensional processing model is trained to obtain the third capture mode of deep learning. If the coordinate information on the monitoring device is difficult to see, the area can be enlarged to a certain extent for easy viewing, so that the trained three-dimensional preprocessing model can better process the three-dimensional processing model and obtain accurate reference coordinate information.

[0132] Further, the steps of determining reference coordinate information on the reference coordinate line according to the preprocessed image, the reference coordinate line and the three-dimensional preprocessing model, and determining positioning information of the monitored device according to the reference coordinate information and the target area information include:

[0133] Step E100, determining reference coordinate information on the reference coordinate line according to the preprocessed image, the reference coordinate line and the three-dimensional preprocessing model;

[0134] Step E200: when there are a plurality of reference coordinate information on the reference coordinate line, determine the distance between each reference coordinate information and the target area information according to the target area information;

[0135] Step E300: Select the reference coordinate information closest to the target edge from each reference coordinate information, and determine the positioning information of the monitored device according to the reference coordinate values ​​closest to the target edge and the target edge.

[0136] Further, after the target area information 411 and the reference coordinate line 422 of the cleaning gun or the cleaning nozzle 42 are determined by the first and second preprocessing models, the preprocessed image is captured by the third preprocessing model to capture the image of the cleaning gun or the cleaning nozzle 42, and the enlargement process is performed to determine the reference coordinate information on the reference coordinate line 422 (such as Figure 3 Then, the cubic preprocessing model is used to determine the distance between each reference coordinate information and the target area information. Figure 3 In the example, the distance between the reference coordinate information A and the target area information 411 is C, and the distance between the reference coordinate coefficient B and the target edge component 411 is D. Since C is less than D, the reference coordinate information closest to the target edge component can be determined as A, and the positioning information of the monitored device can be determined based on the reference coordinate information and the target edge component. For example, the reference coordinate information A of the cleaning gun or nozzle at this time is determined, and the distance C from the target edge is recorded, and the above data is recorded.

[0137] Furthermore, the relative speed between the monitored area and the assembly is determined according to a plurality of positioning information of the monitored area in the plurality of pre-processed images, time information corresponding to each pre-processed image, and a simulation model.

[0138] Further, the simulation model may convert the values ​​determined by the primary preprocessing model, the secondary preprocessing model, and the tertiary preprocessing model into text, thereby facilitating determination of the relative speed between the monitored area and the assembly.

[0139] Furthermore, the reference area information, reference coordinate points, reference coordinate lines and reference positioning information in the reference video data may be calibrated by the tank wall.

[0140] In one embodiment, the time information corresponding to each pre-processed image may be obtained from the pre-processed image or from data corresponding to the video data by a simulation model.

[0141] Furthermore, based on a number of positioning information of the monitored area in a number of processed images and time information corresponding to each processed image, the simulation model includes: determining adjacent positioning information of the monitored area based on a number of positioning information of the monitored area in a number of processed images; determining the relative speed between the monitored area and the assembly based on the adjacent positioning information of the monitored area, the time information of each processed image and the simulation model, including the adjacent positioning information in the processed image.

[0142] Further, Figure 3 For example, assuming that A is 300mm and B is 400mm, there will be two situations:

[0143] In one case, if the cleaning speed between the 300mm and 400mm scale values ​​has been determined, it is necessary to determine the cleaning speed of the cleaning gun or the cleaning nozzle based on the screen with the 200mm scale value. Specifically, the cleaning speed of the cleaning gun or the cleaning nozzle can be determined by using the positioning information of the cleaning gun or the cleaning nozzle at the 300mm scale value, the positioning information of the adjacent 200mm scale, and the time information corresponding to the two positioning information and the simulation model. In this case, there are two more cases: one is that when the 300mm scale value is set, the positioning information of the cleaning gun or the cleaning nozzle is the same as the positioning information of the 200mm scale value, that is, the distance between the 300mm scale value and the target area information is C, and the distance between the 200mm scale value and the target area information is also C. If the time difference between the two is 15 seconds, it can be determined that the average cleaning speed during the period from the 300mm scale value to the 200mm scale value is 400mm / min.

[0144] In one embodiment, the positioning information of the cleaning gun or nozzle at the 300mm scale value is different from the positioning information of the 200mm scale value, that is, the distance between the 300mm scale value and the target area information is C, while the distance between the 200mm scale value and the target area information is not C. The time difference between the two occurrences is 15 seconds. At this time, it is necessary to search for other scale values, and determine the two scale values ​​of the same distance through a large number of scale values, and then determine the average cleaning speed according to the time.

[0145] In another embodiment, the cleaning speed between the 300 mm and 400 mm scale values ​​has not been determined, and there are not two separate screens for the 300 mm and 400 mm scale values. Therefore, the cleaning speed between the 300 mm and 200 mm scale values, the cleaning speed between the 400 mm and 200 mm scale values, and time can be used to determine the cleaning speed between the 300 mm and 400 mm scale values.

[0146] According to the above embodiments, it can be seen that after obtaining the preprocessed images of the monitored device and the combination, the present invention determines the relative speed of the monitored device and the combination through the primary preprocessing model, the secondary preprocessing model, the tertiary preprocessing model and the simulation model. Therefore, the speed of the monitored area can be monitored. Taking a cleaning gun or a cleaning nozzle as an example, the cleaning speed of the cleaning gun or the cleaning nozzle can be monitored, and the operator can be reminded or supervised according to the cleaning speed to improve the effectiveness of the cleaning gun or nozzle detection.

[0147] Furthermore, for example, when monitoring the scale value from 300 mm to 200 mm, if the cleaning speed of the cleaning gun or the cleaning nozzle is faster, the operator can be notified through a prompt to reduce the cleaning speed in time to ensure the effectiveness of the cleaning gun or the cleaning nozzle.

[0148] The present invention provides a method for using a device, which obtains a preprocessed image of a monitored area when the device is working. Through the preprocessed image and the primary preprocessing model, the secondary preprocessing model and the tertiary preprocessing model, the target area information of the combination and the monitored area, and the reference coordinate line of the monitored area, the positioning information of the monitored area can be used to determine the relative speed between the monitored area and the combination through the device positioning information and the simulation model at different times. That is to say, during the cleaning process, the cleaning speed of the cleaning gun or the cleaning nozzle can be determined, and the cleaning speed of the cleaning gun can be determined by the relative displacement of the cleaning gun or the cleaning nozzle, and the cleaning speed of the cleaning gun or the cleaning nozzle can still be accurately determined, thereby realizing the monitoring of the cleaning speed of the cleaning gun or the cleaning nozzle during the cleaning process.

[0149] Example 5 Reference Figure 5 The structural diagram of the equipment monitoring system provided by the present invention includes the following modules:

[0150] The image acquisition unit 51 is mainly used to receive the pre-processed image including the monitored area and the combination combined with the monitored area sent by the front-end device;

[0151] The first target monitoring area 52 is mainly used to determine the target area information of the combination combined with the monitored area according to the preprocessed image and the primary preprocessing model;

[0152] The second target monitoring area 53 is mainly used to determine the reference coordinate points of the monitored area according to the preprocessed image and the secondary preprocessing model, and determine the reference coordinate line of the device to be monitored according to the reference coordinate points and the target area information;

[0153] The third target monitoring area 54 is mainly used to determine the reference coordinate information on the reference coordinate line according to the preprocessed image, the reference coordinate line and the three-dimensional preprocessing model, and determine the positioning information of the monitored area according to the reference coordinate information and the target area information;

[0154] The analysis and judgment unit 55 is mainly used to determine the relative speed between the monitored area and the assembly according to a plurality of positioning information of the monitored area in a plurality of processed images, time information corresponding to each processed image and a simulation model.

[0155] Furthermore, the equipment monitoring device also includes a first training module and a first training module for acquiring reference video data; the reference video data includes reference area information combined with a reference tool and a reference device; a processing model is acquired, and the initial area information in the reference video data is determined based on the first initial capture module and the reference video data; based on the initial area information and the reference area information, a processing model is corrected to determine a preprocessing model.

[0156] Furthermore, the equipment monitoring device also includes a second training module for obtaining a secondary processing model; obtaining primary reference image data, the primary reference image including reference area information in the reference video data combined with the reference assembly and the reference device; the primary reference image data including reference coordinate points and reference coordinate lines; determining the initial coordinate points of the reference device based on the primary reference image data and the secondary processing model, and determining the initial coordinate lines of the reference device based on the initial coordinate points and the reference area information; and correcting the secondary processing model based on the initial coordinate points, the initial coordinate lines, the reference coordinate points and the reference coordinate lines to determine the secondary preprocessing model.

[0157] Furthermore, the equipment monitoring device also includes a third training module for obtaining a tertiary processing model; obtaining secondary reference image data including reference coordinate points and reference coordinate lines in the primary reference image data; the secondary reference image data including reference positioning information of a reference device; determining initial coordinate information of the reference coordinate line based on the secondary reference image data, the reference coordinate line and the tertiary processing model, and determining initial positioning information of the reference device based on the initial coordinate information and the reference area information; and correcting the tertiary processing model based on the initial positioning information and the reference positioning information to determine a tertiary preprocessing model.

[0158] Furthermore, the first target monitoring area 52 is used to determine the image of the assembly based on the preprocessed image and the primary preprocessing model; search for the preprocessing monitoring equipment located outside the assembly based on the image of the assembly and the preprocessed image; determine the boundary between the assembly and the monitored area when searching for the monitored area outside the assembly; determine the target area information combining the assembly and the monitored area based on the image of the assembly and the boundary between the assembly and the monitored area.

[0159] Furthermore, the second target monitoring area 53 is used to determine the two sides of the monitored area based on the preprocessed image and the secondary preprocessing model; determine the parallel parts between the two sides of the monitored area based on the two sides of the monitored area and the secondary preprocessing model; determine the reference coordinate points of the monitored area based on each parallel component and the secondary preprocessing model; determine the reference coordinate line of the monitored area based on the reference coordinate points and area information of the target.

[0160] Furthermore, the second target monitoring area 53 is used to determine the vertical lines of the parallel parts according to each parallel part and the secondary preprocessing model; determine the midpoint of each vertical line according to the vertical lines of the parallel parts and the secondary preprocessing model; and determine the reference coordinate point of the monitored area according to the midpoint of each vertical line.

[0161] Furthermore, the third target monitoring unit 54 is used to determine the reference coordinate information on the reference coordinate line based on the preprocessed image, the reference coordinate line and the three-dimensional preprocessing model; when there are several reference coordinate information on the reference coordinate line, determine the distance between each reference coordinate information and the target area information according to the target area information; select the reference coordinate information closest to the target edge from each reference coordinate information, and determine the positioning information of the monitored area according to the reference coordinate values ​​closest to the target edge and the target edge.

[0162] Furthermore, the identification and analysis unit 55 is used to determine the adjacent positioning information of the monitored area by processing several positioning information in multiple images of the monitored area; and determine the relative speed between the monitored area and the combination according to the adjacent positioning information of the monitored area, the time information of each processed image containing the adjacent positioning information in the processed image, and the simulation model.

[0163] Example 6, reference Figure 6 The workstation may include a radio frequency module 61, a memory 62 including one or more computer-readable storage media, an input unit 63, a display unit 64, a sensor 65, an audio module 66, a wireless module 67, a processor 68 including one or more processing cores, and a power supply 69, as well as other components, wherein:

[0164] The radio frequency module 61 can be used to receive and send signals in the process of sending and receiving information or calls, especially to receive downlink information of the base station and hand it over to one or more processors 68 for processing; further, to send data related to the uplink to the base station. The memory 62 can be used to store software programs and modules, and the processor 68 executes various functional applications and data processing by running the software programs and modules stored in the memory 62. The input unit 63 can be used to receive input digital or character information and generate signal input related to user settings and function control.

[0165] The display unit 64 may be used to display information input by a user or provided to a user, as well as various graphical user interfaces of the workstation, which may be composed of graphics, images, texts, icons, videos, and any combination thereof.

[0166] The workstation may also include at least one sensor 65, such as a light sensor, a motion sensor, and other sensors. An audio module 66 includes a speaker that provides an audio interface between the user and the workstation.

[0167] The workstation also includes a power supply 69 for supplying power. Preferably, the power supply can be logically connected to the processor δ8 through a power management system, so as to implement functions such as charging, discharging and power consumption management through the power management system.

[0168] Furthermore, the workstation system may also include a camera, a Bluetooth module, etc., which will not be described in detail here. In the present invention, the processor δ8 in the workstation will load the corresponding executable file into the memory 62 according to the following instructions, and the processor 68 runs the application stored in the memory 62 to achieve the following functions:

[0169] Receiving a preprocessed image, and determining target area information of a combination combined with a monitoring device according to the preprocessed image and a primary preprocessing model;

[0170] Determine the reference coordinate points of the monitored area according to the preprocessed image and the secondary preprocessing model, and determine the reference coordinate line of the equipment to be monitored according to the reference coordinate points and the area information of the target;

[0171] Determine reference coordinate information on the reference coordinate line according to the preprocessed image, the reference coordinate line and the three-dimensional preprocessing model, and determine the positioning information of the monitored area according to the reference coordinate information and the target area information;

[0172] The relative speed between the monitored area and the assembly is determined according to a plurality of positioning information of the monitored area in a plurality of processed images, time information corresponding to each processed image and a simulation model.

[0173] Furthermore, the present invention provides a computer-readable storage medium, which includes any medium that can record and store information, etc., and stores a number of instructions that can be loaded by a processor to implement the following functions:

[0174] Receiving an image sent by a front-end device;

[0175] Determine the target area information of the assembly combined with the monitoring equipment according to the preprocessed image and the primary preprocessing model; determine the reference coordinate points of the monitored area according to the preprocessed image and the secondary preprocessing model, and determine the reference coordinate line of the equipment to be monitored according to the reference coordinate points and the target area information;

[0176] Determine reference coordinate information on the reference coordinate line according to the preprocessed image, the reference coordinate line and the three-dimensional preprocessing model, and determine the positioning information of the monitored area according to the reference coordinate information and the target area information;

[0177] The relative speed between the monitored area and the assembly is determined according to a plurality of positioning information of the monitored area in a plurality of processed images, time information corresponding to each processed image and a simulation model.

[0178] Due to the instructions stored in the storage medium, the steps in any method provided by the present invention can be executed, thereby achieving the beneficial effects that can be achieved by any method described in the present invention.

[0179] The above describes in detail a monitoring system for a tank cleaning process and a cleaning gun and a method for using the same provided by the present invention. The embodiment introduces its principle and implementation method through specific examples.

[0180] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A visual monitoring method for a tank cleaning process and a cleaning gun, characterized in that: include: Receiving a pre-processed image sent by a front-end device, including a monitored area and a combination combined with the monitored area; Determine target area information of a combination combined with the monitoring device according to the preprocessed image and the primary preprocessing model; Determine the reference coordinate points of the monitored area based on the preprocessed image and the secondary preprocessing model, and determine the reference coordinate lines of the equipment to be monitored based on the reference coordinate points and the target area information; According to the preprocessed image, the reference coordinate line and the three-dimensional preprocessing model, the reference coordinate information on the reference coordinate line is determined, and according to the reference coordinate information and the target area information, the positioning information of the monitored area is determined; The relative displacement speed between the monitored area and the assembly is determined based on a plurality of positioning information in a plurality of pre-processed images of the monitored area and the corresponding time information and model.

2. A visual monitoring method for a tank cleaning process and a cleaning gun according to claim 1, characterized in that: Before the preprocessed image and the primary preprocessed model determine the target area information of the combination combined with the monitoring device, the method further includes: Acquire reference image data; the reference image data includes reference target and reference region information combined with the reference target, the reference target includes a preprocessed image and a preprocessed model; Acquire a first initial capture processing model, the first initial capture processing model being the regional structure and range formed within the cleaning range, and determine initial region information in the reference video data according to the first initial capture module and the reference image data; According to the initial region information and the reference region information, a processing model is modified to determine a preprocessing model.

3. A visual monitoring method for a tank cleaning process and a cleaning gun according to claim 2, characterized in that: Before the preprocessed image and the secondary preprocessed model determine the reference coordinate points of the monitored area, and before the reference coordinate lines of the monitored system are determined according to the reference coordinate points and the area information of the target, the method further includes: obtaining the secondary processing model; Obtaining first initial captured image data, wherein the first initial captured image includes capture area information of a capture assembly and a capture element combination in the captured video data; the first captured image data includes capture coordinate points and capture coordinate lines; Determine the initial coordinate point of the capture device according to the first captured image data and the second capture processing model, and determine the initial coordinate line of the capture device according to the initial coordinate point and the capture area information; The secondary processing model is corrected according to the initial coordinate points, the initial coordinate lines, the captured coordinate points and the captured coordinate lines to determine the secondary preprocessing model.

4. A visual monitoring method for a tank cleaning process and a cleaning gun according to claim 3, characterized in that: The preprocessed image, the reference coordinate line and the three-dimensional preprocessing model determine the reference coordinate information on the reference coordinate line, and before determining the positioning information of the monitored area according to the reference coordinate information and the target area information, the method further includes: obtaining the three-times processing model; Obtaining secondary reference image data of reference coordinate points and reference coordinate lines in the primary reference image data; the secondary reference image data includes reference positioning information of the reference device; Determine initial coordinate information of the reference coordinate line according to the secondary reference image data, the reference coordinate line and the three-dimensional processing model, and determine initial positioning information of the reference target according to the initial coordinate information and the reference area information; According to the initial positioning information and the reference positioning information, the tertiary processing model is modified to determine the tertiary preprocessing model.

5. The visual monitoring method of a tank cleaning process and a cleaning gun according to claim 1, characterized in that: The steps of determining the target area information of the assembly combined with the monitoring device by using the preprocessed image and the primary preprocessing model include: Step A1, determining an image of the assembly according to the preprocessed image and the primary preprocessing model; Step A2: searching for a preprocessing monitoring device located outside the assembly according to the image of the assembly and the preprocessing image; Step A3, when searching for the monitored area outside the assembly, determining the boundary between the assembly and the monitored area; Step A4: determining target area information combining the assembly and the monitored area based on the image of the assembly and the boundary between the assembly and the monitored area.

6. A visual monitoring method for a tank cleaning process and a cleaning gun according to claim 1, characterized in that: The steps of determining the reference coordinate points of the monitored area using the preprocessed image and the secondary preprocessed model, and determining the reference coordinate line of the system to be monitored according to the reference coordinate points and the target area information include: Step B1, determining the two sides of the monitored area according to the preprocessed image and the secondary preprocessing model; Step B2, determining the parallel portion between the two sides of the monitored area according to the two sides of the monitored area and the secondary preprocessing model; Step B3, determining the reference coordinate points of the monitored area according to each parallel component and the secondary preprocessing model; Step B4: Determine the reference coordinate line of the monitored area according to the reference coordinate point and area information of the target.

7. A visual monitoring method for a tank cleaning process and a cleaning gun according to claim 6, characterized in that: The steps of each parallel component and the secondary preprocessing model determining the reference coordinate points of the monitored area include: Step C1, determining a vertical line of each parallel part according to each parallel part and the secondary preprocessing model; Step C2, determining the midpoint of each vertical line according to the vertical lines of the parallel parts and the secondary preprocessing model; Step C3: Determine the reference coordinate point of the area to be monitored according to the midpoint of each vertical line.

8. The visual monitoring method of a tank cleaning process and a cleaning gun according to claim 1, characterized in that: The steps of determining reference coordinate information on the reference coordinate line using the preprocessed image, the reference coordinate line and the three-dimensional preprocessing model, and determining the location information of the monitored area according to the reference coordinate information and the target area information include: Step D1, determining reference coordinate information on the reference coordinate line according to the preprocessed image, the reference coordinate line and the three-dimensional preprocessing model; Step D2: when there are several reference coordinate information on the reference coordinate line, determine the distance between each reference coordinate information and the target area information according to the target area information; Step D3: Select the reference coordinate information closest to the target edge from each reference coordinate information, and determine the positioning information of the area to be monitored according to the reference coordinate values ​​closest to the target edge and the target edge.

9. A visual monitoring method for a tank cleaning process and a cleaning gun according to claim 1, characterized in that: The step of determining the relative displacement speed between the monitored area and the assembly based on the plurality of positioning information of the monitored area in the plurality of pre-processed images, the corresponding time of each pre-processed image information and the simulation model thereof comprises: Step E1, determining adjacent positioning information of the monitored device according to a plurality of positioning information of the monitored device in a plurality of pre-processed images; Step E2: determining the relative displacement speed between the monitored device and the combination according to the adjacent positioning information of the monitored device, the time information of each processed image of the adjacent positioning information contained in the processed image, and the simulation model.

10. A visual monitoring system for a tank cleaning process and a cleaning gun, based on a visual monitoring method for a tank cleaning process and a cleaning gun according to any one of claims 1 to 9, characterized in that: include: Using the front-end device and the image capture and acquisition unit to receive a pre-processed image of the monitored area, the pre-processed image includes the monitored area and a combination combined with the monitored area; A first monitoring area, for use in preprocessed images and a primary preprocessing model to determine target area information of a combination combined with a monitoring device; The second monitoring area is used to preprocess the image and the secondary preprocessing model to determine the reference coordinate points of the monitored area, and determine the reference coordinate line of the monitored area according to the reference coordinate points and the area information of the target; The third monitoring area is used to preprocess the image, the reference coordinate line and the three-dimensional preprocessing model to determine the reference coordinate information on the reference coordinate line, and determine the positioning information of the monitored area according to the reference coordinate information and the area information of the target; The identification and analysis unit is used to determine the relative displacement speed between the monitored area and the assembly based on the positioning information in the processed images of the monitored area, the corresponding time information of each processed image and the simulation model.

Images