Intelligent filling system and filling method

Abstract

The invention discloses an intelligent filling system and a filling method, and relates to the technical field of intelligent filling, a pressure sensor is adopted to monitor whether a container enters a filling table or not, and a terminal is triggered to generate a detection instruction; a camera cluster is used for collecting images, the position relation between a filling opening and a container opening is extracted, a standard and a to-be-filled planar graph are fitted and verified, and the filling condition is determined; if normal, constructing a panoramic image of the container, determining the capacity and the filling amount, and executing filling; the system comprises a data acquisition module, a data storage module and a data processing module. The pressure sensor and the camera cooperate, the filling process is accurately controlled through the image processing technology, the filling precision and efficiency are improved, the device is suitable for various containers, manual intervention is reduced, and the production automation level is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent filling, and specifically relates to an intelligent filling system and a filling method. Background Art

[0002] With the rapid development of the economy, the demand for automation and intelligence in industrial production is increasing. In the filling industry, traditional filling technology mainly relies on manual operation or simple mechanical control, and has problems such as low efficiency and insufficient precision.

[0003] Although sensor monitoring and image acquisition equipment have been introduced into existing technologies, existing monitoring methods are usually limited to single-point pressure detection, which cannot accurately determine whether the container is correctly positioned, easily causing empty filling or refilling, affecting production efficiency; in the filling volume control link, traditional methods are mostly based on preset parameters and cannot dynamically adjust according to the actual shape of the container, and cannot meet the production needs of special-shaped containers or capacity fluctuations; in addition, the traditional filling system has a low level of intelligence, lacks correlation analysis of data from each link, and cannot form a complete production monitoring closed loop.

[0004] In order to solve the above problems, the present application proposes an intelligent filling system and filling method. Summary of the Invention

[0005] In response to the shortcomings of the existing technology, the present invention provides an intelligent filling system and filling method, which solves the problems of the existing technology that it cannot dynamically adjust according to the actual shape of the container and lacks a complete production monitoring closed loop.

[0006] The purpose of the present invention can be achieved through the following technical solutions: An intelligent filling method, the method comprising the following steps: Step 1: A pressure sensor is installed on the bottom of the filling table to monitor whether there are any containers entering the filling table. If not, continuous monitoring is performed; If yes, a first signal is generated and transmitted to the terminal processor, and the terminal processor confirms the first signal, generates a detection instruction and transmits it to the camera cluster installed in the filling station area for image acquisition; Step 2: Extract the positional relationship between the filling port and the filling table, and fit the images collected by the combined camera cluster into a standard filling plan; Based on the images collected by the camera cluster, the positional relationship between the container mouth and the filling table is extracted, and the plane map to be filled is fitted; Verify the to-be-filled plan with the standard filling plan, and determine abnormal filling conditions and normal filling conditions based on the verification results; If the result of the verification process is an abnormal filling situation, skip the current filling operation and repeat steps 1 and 2; If the result of the verification process is normal filling, go to step 3; Step 3: Extract the images collected by the camera cluster again, construct a panoramic image of the current container, and determine the container capacity; A filling quantity is further determined based on the determined container capacity, and a filling operation is performed.

[0007] As a further solution of the present invention, in step 1, the specific method of generating the detection instruction is: The initial value of the calibrated pressure sensor is 0; The pressure sensor's value is obtained in real time through the edge processor connected to the pressure sensor ; like , then the built-in counter T of the edge processor is triggered to start timing from 0 seconds, and T increases with time. If the length of the counter T is greater than the time threshold preset by the operator , it is determined that there is a container on the pressure sensor, and the edge processor generates a first signal and transmits it to the terminal processor through the system bus.

[0008] As a further solution of the present invention, in step 1, the specific method of performing image acquisition is: The detection instructions generated by the terminal processor are transmitted to the camera cluster via the system bus; Let the camera cluster be ,in, is a count index, indicating the total number of cameras in the camera cluster; The camera cluster is used to fully cover the entire filling station area and take pictures when there are no containers in the filling station area. images, recorded as the filling station image set ,in, and correspond, is a counting index, ranging from 1 to ; When the camera cluster receives the detection instruction, it will take another photo of the filling station area and send the photo to the container image set , where the container image and the filling station image have a one-to-one correlation with the actual size.

[0009] As a further solution of the present invention, in step 2, the specific method of determining abnormal filling conditions and normal filling conditions based on the result of the verification process is: from Extract the filling station image from the top view directly above the filling station and mark it as ; by The first pixel in the lower left corner is the origin. The left border is the vertical axis, The lower boundary is constructed as the horizontal axis The associated two-dimensional coordinate system; Then obtain the two-dimensional coordinates of the center point of the filling port in the two-dimensional coordinate system , extract the filling port radius , in two-dimensional coordinates As the center of the circle, the radius of the filling port Draw a circle with radius, and you get a circle ; Save Circle At The associated two-dimensional coordinate system is recorded as the standard filling plane. ; Similarly, from Extract the container image from the top view directly above the filling station and mark it as , build The associated two-dimensional coordinate system; Sure The two-dimensional coordinates of the center point of the container mouth in the two-dimensional coordinate system ; Get the container opening radius ,by As the center of the circle, the radius of the container mouth As the radius of the circle, draw a circle and get the circle ; Save Circle At The associated two-dimensional coordinate system is recorded as the plane diagram to be filled ; By aligning the coordinates and Superimpose them on the same plane to obtain a plane diagram; At this time, the circle ,round Placed in the same plane, the circle ,round Perform verification processing: like , then the result of the verification process is determined to be a normal filling situation; like , then the result of the verification process is determined to be an abnormal filling situation.

[0010] As a further solution of the present invention, in step 2, the circle ,round Perform verification processing. If it is determined that the result of the verification processing is an abnormal filling situation, the filling operation of the container is skipped; If the result of the verification process is determined to be a normal filling situation, the capacity of the container and the filling amount are further determined, and the filling operation is performed.

[0011] As a further solution of the present invention, the specific method of determining the capacity and filling amount of the container and performing the filling operation is as follows: From the Filling Station Image Collection Extract any filling station image , and then from the container image collection Extract any container image ; Determine the filling station image and container images The number of pixels in , recorded as m; Will The first filling station pixel in the upper left corner is recorded as , the last filling station pixel in the lower right corner is recorded as ; Extract the filling station image in order from left to right and from top to bottom The m filling station pixels in the filling station pixel sequence are obtained. ; Similarly, the container pixel sequence is obtained: ; from Extract any pixel point of a filling station ; from Extract any pixel point corresponding to the filling station The associated container pixel ; extract Pixel value as well as Pixel value , where n is the counting index, ranging from 1 to m; Sure as well as The difference between pixel values , the difference rate The difference rate threshold preset by the operator Make a comparison; If the difference rate , then the container pixel Considered as foreground pixels; If the difference rate , then the container pixel Considered as background pixels; Again Perform the above steps for all container pixels in the container to determine the container image The foreground pixels and background pixels in the image are removed, and the background pixels are removed to obtain the foreground container image. ; Again as well as Repeat the above steps to obtain the foreground container image set: ; right Fitting is performed to obtain a panoramic image associated with the container, and the container volume is determined based on the panoramic image of the container and the correlation between the panoramic image of the container and the actual size. ; Extract the filling percentage preset by the operator , get the filling volume associated with the container , perform filling operation.

[0012] As a further solution of the present invention, an intelligent filling system is provided, which includes: A data acquisition module, including a pressure sensor, a camera cluster, and an edge processor associated with the camera cluster; A data storage module, including a memory, for storing the analysis and calculation results of any of the steps described above and saving an intelligent filling method; The data processing module includes a terminal processor, which reads an intelligent filling method stored in the memory and runs it to implement any of the steps described above.

[0013] As a further solution of the present invention, the system further includes the following: The pressure sensor is used to monitor whether the container enters the filling station; The camera cluster is used to capture images of the filling station area.

[0014] Beneficial effects of the present invention: (1) The present invention accurately monitors the entry and position status of the container by configuring a pressure sensor and a camera cluster on the bottom surface of the filling table. When the container enters the filling table, the system can quickly generate a signal and trigger the camera cluster to collect images, extract the positional relationship between the filling port and the container port in real time, and fit the standard filling plan and the plan to be filled for verification. This design can effectively avoid filling failure or material waste caused by abnormal container position, ensuring the accuracy of the filling operation. Secondly, the system can automatically determine the filling volume based on the container capacity and execute filling, realizing intelligent filling process management, solving the problem of low adaptability of traditional filling lines, and realizing automation and intelligence of the entire filling process. (2) The present invention constructs a high-precision and high-reliability container detection system through the collaborative mechanism of pressure sensors and edge processors, combined with an intelligent image acquisition system; its core advantage is the use of a double verification mechanism to ensure detection accuracy: the pressure sensor eliminates environmental interference through initial calibration, and effectively avoids instantaneous false triggering in combination with a time threshold counter, and the signal retransmission function of the edge processor enhances the robustness of the system; at the same time, the camera cluster significantly improves the accuracy of container positioning and morphological recognition through pixel-level comparison of the container-free reference image and the real-time container image, under the guarantee of one-to-one physical mapping and unified image parameters; the two sets of system processes form a closed-loop control through bus collaboration, which solves the problem of misjudgment caused by environmental interference in traditional filling equipment, and ensures data consistency through standardized image acquisition processes, achieving a dual technical breakthrough of improved detection efficiency and reduced false detection rate; (3) The present invention realizes automatic verification of the spatial position of the filling port and the container port by constructing a precise mapping between a standardized filling plan and a real-time container image coordinate system. Secondly, the present invention realizes the determination of the overlapping relationship between the filling port and the container through the coordinate system alignment and superposition mechanism, identifies abnormal situations such as container offset and size discrepancy, and realizes production continuity through the abnormal situation handling mechanism, thereby significantly optimizing the production line efficiency while improving the stability of filling quality. (4) This method determines the container capacity and filling volume by analyzing and processing image pixels, thus achieving accurate filling operations. The core advantage lies in the use of serialized pixel comparison technology to dynamically separate foreground and background pixels through a difference rate threshold to obtain a set of effective foreground container images, and then fit a panoramic image of the container. This makes the determination of container volume more accurate and can accurately calculate the filling volume associated with the container based on the filling percentage preset by the operator, thereby ensuring the consistency and accuracy of filling. This image analysis-based method enhances the flexibility and adaptability of the filling process and can be applied to containers of different shapes and sizes, providing an efficient, accurate and broadly applicable solution for the field of intelligent filling. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The present invention will be further described below with reference to the accompanying drawings.

[0016] Figure 1 It is a structural diagram of the system of the present invention; Figure 2 Schematic diagram of the process of the method described in Example 2 of the present invention; Figure 3 Schematic diagram of the process of the method described in Example 3 of the present invention; Figure 4 Schematic diagram of the process of the method described in Example 4 of the present invention. DETAILED DESCRIPTION

[0017] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0018] Example 1 An intelligent filling system, such as Figure 1 As shown, the system includes the following: A data acquisition module, which is interconnected with the physical object and includes a pressure sensor for real-time monitoring of whether a container is present at the filling station. When the value of the pressure sensor changes and remains stable for a period of time, it is determined that a container is present at the filling station. The edge processor equipped with the pressure sensor generates a first signal and transmits it to the terminal processor in the data processing module; The pressure sensor value is not fixedly calibrated here because when different containers are used for filling operations, the weight of each container will vary. Therefore, the presence of a container on the filling table cannot be detected based on the fixed weight value. Secondly, if the operator is required to manually enter the fixed weight value before filling, the workload will increase to a certain extent. Since the assembly line operates 24 hours a day, in contrast, using a pressure sensor value that has been stable for a period of time can effectively improve the efficiency of monitoring the presence of containers at the filling table. Secondly, the data acquisition module also includes a camera cluster, which includes multiple cameras and is arranged in different directions of the filling station area, so as to meet the requirements of all-round acquisition of images in the filling station area.

[0019] The data storage module includes a memory for storing the calculation method and calculation results in any step of this solution, or the analysis method and analysis results.

[0020] Data processing module, which serves as the data processing terminal of the system and includes a terminal processor for processing and real-time calculation steps and analysis steps described in this solution; Used to issue detection instructions to camera clusters for image acquisition; Used to transmit filling operation instructions to the filling station to perform filling operations.

[0021] Example 2 This embodiment discloses a method for acquiring images through a camera cluster. Figure 2 As shown, the specific steps include: Data preparation is required before filling operations, including numerical calibration of the pressure sensor and determination of lighting conditions for the camera cluster; The numerical calibration operation of the pressure sensor is to adjust the initial value of the pressure sensor to 0, which means there is no container on the filling table, and to calibrate the real-time pressure value of the pressure sensor (the pressure value is calibrated as ) is transmitted to the edge processor connected to the pressure sensor for processing. If the edge processor monitors the pressure value If the pressure value is greater than 0, the built-in counter T of the edge processor is triggered. The initial value of the counter T is also 0 (0 seconds or 0 milliseconds, the specific value is determined by the operator based on the actual situation). When it is greater than 0, at this moment, the counter T starts timing, and T increases with time. If it is detected that the length of the counter T timing is greater than the time threshold preset by the operator, it means that the value of the pressure sensor is in a stable state, and it is further determined that there is a container on the pressure sensor, that is, there is a container at the filling station; At this point, the edge processor connected to the pressure sensor generates a first signal, and transmits the first signal generated by the edge processor to the terminal processor in the data processing module through the system bus; If the terminal processor in the data processing module receives the first signal transmitted by the edge processor, it generates a detection instruction and transmits it to the camera cluster; If the terminal processor in the data processing module confirms that the reception of the first signal transmitted by the edge processor fails, the confirmed failure information is fed back to the edge processor connected to the pressure sensor, and requests retransmission of the first signal.

[0022] The lighting conditions of the camera cluster need to be determined before the camera cluster takes pictures and no longer adjusted afterwards. This is because if the lighting changes during the camera cluster's image capture process, the images captured will be significantly different, leading to misjudgments during subsequent operations. Before the camera cluster receives the detection command transmitted by the terminal processor and captures images in the filling station area, it must first capture an image of the filling station area without containers as a reference for subsequent image data processing (if the shooting conditions (such as lighting, number of cameras) change, the image without containers needs to be updated); Determine the total number of cameras in the camera cluster, denoted as , and the camera cluster is represented as: ,in, Represents the first to cameras, and is a count index, indicating the total number of cameras in the camera cluster; Through camera clusters The whole filling station area is photographed in all directions to obtain the coverage of the whole filling station area. images, by camera cluster The order of the cameras in The images are recorded as the filling station image set, which is expressed as: , where, determine any filling station image The associated camera is ,and is a counting index, The value range is 1 to ; Filling station image collection Save; When the camera cluster receives the detection instruction (there is a container on the filling table), it uses the camera cluster The filling station area where the container exists is photographed, and the photographed filling station area where the container exists is photographed. images by camera cluster The sorting order of the cameras in is recorded as the container image set, which is expressed as: ; The container images and filling station images in the container image set and the filling station image set are all correlated with the actual size in a one-to-one ratio, so that the corresponding size and position in real space can be accurately determined directly based on information such as the position and size of each part of the container or filling station in the image.

[0023] The container image and the filling station image are both captured by the same type of camera, with the same number of pixels and the same image resolution (the number of pixels in each image is also the same, m), to avoid different measurement errors caused by differences in image quality.

[0024] The purpose of this embodiment is to provide accurate preliminary data support for the filling operation, ensure that the pressure sensor can accurately detect the presence of the container, and the camera cluster can capture images that are accurately correlated with the actual size, thereby ensuring the accuracy of the filling operation and the reliability of subsequent image data processing.

[0025] Example 3 This embodiment further discloses a method for judging abnormal filling conditions and normal filling conditions based on embodiment 2, such as Figure 3 As shown, the specific steps include: Based on the steps described in Example 2, a filling station image set can be obtained: And a collection of container images: ; Next, from the filling station image collection: Extract the filling station image from the top view directly above the filling station and mark it as This step can be achieved by determining a camera specifically for shooting the top view directly above the filling table when setting up the camera, and then directly extracting the image captured by the camera.

[0026] Filling station image Construct a two-dimensional coordinate system, and each coordinate point in the two-dimensional coordinate system corresponds to a pixel point. The method of constructing a two-dimensional coordinate system is as follows: Filling station image The pixel point in the lower left corner is used as the origin of the two-dimensional coordinate system. The left boundary is used as the vertical axis of the origin of the two-dimensional coordinate system, and the filling station image The lower boundary is used as the origin of the two-dimensional coordinate system to construct the filling station image. The associated two-dimensional coordinate system; Then obtain the two-dimensional coordinates of the center point of the filling port in the two-dimensional coordinate system, recorded as In general, the center point of the filling port corresponds to the center point of the filling table. If there are special circumstances, the operator shall determine it based on the actual situation. Then extract the radius of the filling port and mark it as , in two-dimensional coordinates As the center of the circle, the radius of the filling port As the radius of the circle, we can draw a circle , wherein the circle This is the liquid outlet area of the filling port.

[0027] The circle being measured Save it in the two-dimensional coordinate system obtained in the above steps, and record the two-dimensional coordinate system at this time as: standard filling plane ; For container image collections: Repeat the above steps to Extract the container image from the top view directly above the filling station and label it as: , from the container image Get the two-dimensional coordinates of the center point of the container mouth in the two-dimensional coordinate system, recorded as ; Synchronously build container images The associated two-dimensional coordinate system; Then, all container images taken from the front and side of the filling station area are extracted from the container image set. The image of the container mouth area can be obtained through image separation technology, and the actual size, that is, the width of the container mouth, is obtained. The width of the container mouth is the diameter of the container mouth, and half of the diameter is taken to obtain the container mouth diameter. , in two-dimensional coordinates As the center of the circle, the diameter of the container mouth As the radius of the circle, we can draw a circle and get the circle , and the circle Save in container image The associated two-dimensional coordinate system is recorded as: the plane to be filled .

[0028] At this point, you can get the standard filling plan And the floor plan to be filled , the standard filling plan This is an image taken without a container, showing a standard filling plan. The filling outlet area and the plan view of the filling port are shown in the figure. This is an image taken when there is a container and the plane is waiting to be filled. The vessel mouth area is shown in the figure; Plan of the floor plan to be filled With standard filling floor plan Fusion is performed to obtain a plane map, that is, the two coordinate systems are aligned and superimposed on the same plane; At this time, the circle and circle Placed in the same plane, judge the circle ,round If the positional relationship between , then the result of the verification process is determined to be normal filling; if , then the result of the verification process is determined to be an abnormal filling situation; If the final result of the verification process is determined to be an abnormal filling situation, the filling operation of the current container will be skipped and the next container will be transported to the filling station via the conveyor belt on the filling station (the container with the abnormal filling situation will be sent for inspection and handled by the operator); If the result of the verification process is determined to be a normal filling situation, the capacity of the container and the filling amount are determined, and the filling operation is performed.

[0029] This embodiment extracts a top-down image directly above the filling station and constructs a two-dimensional coordinate system to determine the center point and radius of the filling port to form a standard filling plane. Similarly, information related to the container port is obtained from the container image to construct a plane to be filled. After fusing and aligning the two, the positional relationship between circles is used to determine whether the filling is normal. If the current container is abnormal, filling is skipped; if normal, the capacity and filling volume are determined and filling is executed. This embodiment aims to use image recognition and coordinate positioning technology to achieve automated and precise detection and control of the filling process, thereby ensuring the accuracy and reliability of the filling operation.

[0030] Example 4 This embodiment further discloses a method for determining the capacity and filling amount of a container and performing a filling operation based on the embodiment 2 and the embodiment 3, such as Figure 4 As shown, the specific steps include: When the result of the verification process in Example 3 is a normal filling situation, the filling station image set is extracted using the method described in Example 2: And a collection of container images: ; From the identified container image collection: Extract any container image , then from the filling station image collection: Extract any filling station image ; Filling station image The first pixel in the upper left corner is recorded as the first pixel , the last pixel in the lower right corner is recorded as , and extract the filling station image in the order from left to right and from top to bottom m pixels in the filling station to obtain the pixel sequence: ; Extract the container image again The first pixel in the upper left corner And the last pixel , and extract the container image as described above m pixels in , and get the container pixel sequence: ; From the determined filling station image sequence and container image sequences Extract any two pixels: filling station pixel points And the filling station pixel The associated container pixel , and then extract the pixel points of the filling station The pixel value is recorded as , and then extract the container pixels The pixel value is recorded as , where n is the counting index, ranging from 1 to m; For the determined pixel value and pixel values The pixel value difference rate is calculated using the formula: ; Confirm the pixel points of the filling station and container pixels The difference between the pixel values ; Then the determined difference rate The difference rate threshold is preset based on the actual situation of the operator Compare, if the difference rate Greater than the difference rate threshold , then the container pixel Considered as foreground pixels, if the difference rate Less than or equal to the difference rate threshold , then the container pixel Considered as background pixels; So far, the pixel Processing completed, container image The other pixels in the same way as the processing pixels The container image can be finally determined by All foreground pixels and background pixels in the image are removed, and all background pixels are removed (the background part is removed, and the pixel part of the container itself is retained). The foreground container image containing only the foreground pixels associated with the container is obtained and recorded as ; At this point, the container image After processing, collect the filling station images: And a collection of container images: Repeat the above steps to get a collection of container images: The foreground container images associated with all container images in are recorded as the foreground container image set according to the sorting order of the container image set, which is expressed as ; For the foreground container image set: Perform image fitting to obtain the panoramic image associated with the container itself, and further determine the container volume based on the correlation between the determined panoramic image of the container and the actual size, which is recorded as ; Then extract the filling percentage preset by the operator , combined container volume Determine the fill volume associated with the current container , and operate the filling port to perform the filling operation, the filling volume is .

[0031] This example extracts a filling station image set and a container image set from Example 2. One image is taken from each, and pixels are extracted in a specific order to form a sequence. Next, the pixel value difference ratio between the filling station pixels and the container pixels is calculated. Based on the comparison of this difference ratio with a preset threshold, foreground and background pixels are distinguished, and the foreground container pixels are retained to generate a foreground container image, forming a foreground container image set. Image fitting is then performed on the foreground container image set to obtain a panoramic image of the container. The container volume is then determined based on the actual size correlation. The specific filling amount is then determined based on the operator's preset filling percentage, and the filling port is operated to perform the filling operation. Through image processing and pixel analysis, container volume is determined, providing a key basis for subsequent filling operations, thereby improving the automation and intelligence level of the entire filling process.

[0032] Some of the data in the formulas described above are dimensionless and numerically calculated. Meanwhile, the contents not described in detail in this specification belong to the prior art known to those skilled in the art.

[0033] The above contents are merely examples and explanations of the present invention. Those skilled in the art may make various modifications or additions to the described specific embodiments or replace them in similar ways. As long as they do not deviate from the invention or exceed the scope defined by the claims, they should all fall within the scope of protection of the present invention.

[0034] It is important to note that all user data collected in this application is collected with the user's consent and authorization. Furthermore, the use of user data is legal and compliant, and the use and processing of user data complies with the relevant laws, regulations, and standards of the relevant regions.

Claims

1. An intelligent filling method, characterized in that: This method comprises the following steps: Step 1: A pressure sensor is installed on the bottom of the filling table to monitor whether there are any containers entering the filling table. If not, continuous monitoring is performed; If yes, a first signal is generated and transmitted to the terminal processor, and the terminal processor confirms the first signal, generates a detection instruction and transmits it to the camera cluster installed in the filling station area for image acquisition; Step 2: Extract the positional relationship between the filling port and the filling table, and fit the images collected by the combined camera cluster into a standard filling plan; Based on the images collected by the camera cluster, the positional relationship between the container mouth and the filling table is extracted, and the plane map to be filled is fitted; Verify the to-be-filled plan with the standard filling plan, and determine abnormal filling conditions and normal filling conditions based on the verification results; If the result of the verification process is an abnormal filling situation, skip the current filling operation and repeat steps 1 and 2; If the result of the verification process is normal filling, go to step 3; Step 3: Extract the images collected by the camera cluster again, construct a panoramic image of the current container, and determine the container capacity; A filling quantity is further determined based on the determined container capacity, and a filling operation is performed.

2. The intelligent filling method according to claim 1, characterized in that: In step 1, the specific method of generating the detection instruction is: The initial value of the calibrated pressure sensor is 0; The pressure sensor's value is obtained in real time through the edge processor connected to the pressure sensor ; like 0, the built-in counter T of the edge processor is triggered to start timing from 0 seconds, and T increases with time. If the length of the counter T is greater than the time threshold preset by the operator , it is determined that there is a container on the pressure sensor, and the edge processor generates a first signal and transmits it to the terminal processor through the system bus.

3. The intelligent filling method according to claim 1, characterized in that: In step 1, the specific method of performing image acquisition is: The detection instructions generated by the terminal processor are transmitted to the camera cluster via the system bus; Let the camera cluster be ,in, is a count index, indicating the total number of cameras in the camera cluster; The camera cluster is used to fully cover the entire filling station area and take pictures when there are no containers in the filling station area. images, recorded as the filling station image set ,in, and correspond, is a counting index, ranging from 1 to ; When the camera cluster receives the detection instruction, it will take another photo of the filling station area and send the photo to the container image set , where the container image and the filling station image have a one-to-one correlation with the actual size.

4. The intelligent filling method according to claim 3, characterized in that: In step 2, the specific method of determining abnormal filling conditions and normal filling conditions based on the result of the verification process is as follows: from Extract the filling station image from the top view directly above the filling station and mark it as ; by The first pixel in the lower left corner is the origin. The left border is the vertical axis, The lower boundary is constructed as the horizontal axis The associated two-dimensional coordinate system; Then obtain the two-dimensional coordinates of the center point of the filling port in the two-dimensional coordinate system , extract the filling port radius , in two-dimensional coordinates As the center of the circle, the radius of the filling port Draw a circle with radius, and you get a circle ; Save Circle At The associated two-dimensional coordinate system is recorded as the standard filling plane. ; Similarly, from Extract the container image from the top view directly above the filling station and mark it as , build The associated two-dimensional coordinate system; Sure The two-dimensional coordinates of the center point of the container mouth in the two-dimensional coordinate system ; Get the container opening radius ,by As the center of the circle, the radius of the container mouth As the radius of the circle, draw a circle and get the circle ; Save Circle At The associated two-dimensional coordinate system is recorded as the plane diagram to be filled ; By aligning the coordinates and Superimpose them on the same plane to obtain a plane diagram; At this time, the circle ,round Placed in the same plane, the circle ,round Perform verification processing: like , then the result of the verification process is determined to be a normal filling situation; like , then the result of the verification process is determined to be an abnormal filling situation.

5. The intelligent filling method according to claim 4, characterized in that: In the step 2, the circle ,round Perform verification processing. If it is determined that the result of the verification processing is an abnormal filling situation, the filling operation of the container is skipped; If the result of the verification process is determined to be a normal filling situation, the capacity of the container and the filling amount are further determined, and the filling operation is performed.

6. The intelligent filling method according to claim 5, characterized in that: The specific methods for determining the capacity of the container and the filling amount and performing the filling operation are as follows: From the Filling Station Image Collection Extract any filling station image , and then from the container image collection Extract any container image ; Determine the filling station image and container images The number of pixels in , recorded as m; Will The first filling station pixel in the upper left corner is recorded as , the last filling station pixel in the lower right corner is recorded as ; Extract the filling station image in order from left to right and from top to bottom The m filling station pixels in the filling station pixel sequence are obtained. ; Similarly, the container pixel sequence is obtained: ; from Extract any pixel point of a filling station ; Extract any pixel point corresponding to the filling station The associated container pixel ; extract Pixel value as well as Pixel value , where n is the counting index, ranging from 1 to m; Sure as well as The difference between pixel values , the difference rate The difference rate threshold preset by the operator Make a comparison; like , then see is the foreground pixel; like , then see is the background pixel; Again Perform the above steps for all container pixels in the container to determine the container image The foreground pixels and background pixels in the image are removed, and the background pixels are removed to obtain the foreground container image. ; Again as well as Repeat the above steps to obtain the foreground container image set: ; right Fitting is performed to obtain a panoramic image associated with the container, and the container volume is determined based on the panoramic image of the container and the correlation between the panoramic image of the container and the actual size. ; Extract the filling percentage preset by the operator , get the filling volume associated with the container , perform filling operation.

7. An intelligent filling system, characterized in that: This system includes: A data acquisition module, including a pressure sensor, a camera cluster, and an edge processor associated with the camera cluster; A data storage module, comprising a memory for analyzing and calculating the results of the steps described in any one of claims 1 to 6 and storing an intelligent filling method; The data processing module includes a terminal processor, which reads an intelligent filling method stored in a memory and runs it to implement the steps described in any one of claims 1-6.

8. The intelligent filling system according to claim 7, characterized in that: The system also includes the following: The pressure sensor is used to monitor whether the container enters the filling station; The camera cluster is used to capture images of the filling station area.

Images