Micro-nozzle clogging monitoring system and method based on image processing

Abstract

The present invention discloses a micro-nozzle clogging monitoring system based on image processing, comprising: an image acquisition unit, a data processing unit, a remote control unit, and an intelligent monitoring terminal; wherein the image acquisition unit acquires a real-time image of the micro-nozzle's operating status; the data processing unit receives the real-time image and performs data processing, identifies whether the current micro-nozzle is clogged, and transmits the processing result to the remote control unit, which then sends abnormal data to the intelligent monitoring terminal to alert staff. The present invention also discloses a micro-nozzle clogging monitoring method based on image processing, which effectively solves the problem of real-time monitoring of the entire process of micro-nozzle clogging from initial clogging to severe clogging in actual operation; the present invention has high working efficiency and can realize real-time feedback of monitored image data, so that staff can intervene in the working conditions of the micro-nozzle in a timely manner to prevent the micro-nozzle from being completely clogged, thereby avoiding significant damage to people's production and life.

Description

Technical Field

[0001] The present invention relates to the field of intelligent monitoring technology, and in particular to a micro-nozzle clogging monitoring system and method based on image processing. Background Art

[0002] Nozzles are common mechanical components, widely used in industrial production, agricultural production, environmental protection, cooling, cleaning, sterilization, disinfection, fire prevention, and other fields. With the development of energy-saving and environmental protection standards and technologies, micro-nozzles are playing an increasingly important role in these applications.

[0003] Micronozzles are named for their tiny orifice size, typically ranging from a few microns to tens of microns. They are widely used in numerous industrial precision instruments, such as micro-nanoprocessing equipment, biomedical sensors for spotting biomolecules, and optical devices that spray liquid lens materials. In practical applications, to ensure effective spraying and conserve liquid, micronozzles typically include both liquid and high-pressure gas lines as inputs. Compared to conventional nozzles, micronozzles are more susceptible to orifice clogging due to factors such as liquid impurities, residual dirt, air pressure disturbances, and hydraulic pressure fluctuations. However, the time from initial orifice clogging to complete orifice clogging is much shorter.

[0004] Ordinary nozzles are mostly inspected manually visually. When a nozzle begins to partially clog, workers can clearly detect the abnormality visually, and after a brief intervention, the nozzle can often quickly recover and function normally. However, due to the small nozzle size of micro nozzles, it is difficult for workers to accurately identify abnormalities in the spray area when they begin to clog. On the other hand, the time it takes for micro nozzles to completely clog is short. Once workers detect abnormalities in the spray area, they often have no time to take intervention measures, resulting in a serious blockage of the micro nozzle, causing certain economic losses to industrial production. Summary of the Invention

[0005] In order to overcome the above-mentioned technical problems, the present invention discloses a micro-nozzle clogging monitoring system and method based on image processing. The system and method complete real-time and efficient micro-nozzle clogging monitoring through image acquisition and data processing of the spray area of ​​the micro-nozzle, and realize real-time data feedback of the entire process of micro-nozzle clogging from the beginning to complete clogging, so as to construct a closed-loop intelligent monitoring system for micro-nozzle anti-clogging.

[0006] In one aspect, the present invention provides a micro-nozzle clogging monitoring system based on image processing, comprising:

[0007] The monitoring system includes: an image acquisition unit, a data processing unit, a remote control unit and an intelligent monitoring terminal; wherein,

[0008] The image acquisition unit acquires a real-time image of the working state of the micro-nozzle, transmits the data of the real-time image to the data processing unit, and receives instructions from the remote control unit at the same time;

[0009] The data processing unit receives the data of the real-time image, processes the data through the data processing unit, determines whether the current micro-nozzle is in a blocked state, and transmits the data processing result to the remote control unit;

[0010] On the one hand, the remote control unit sends control instructions to the image acquisition unit; on the other hand, it receives the data processing results sent by the data processing unit, and sends the micro-nozzle blockage status in the data processing results in the form of text to the intelligent monitoring terminal to remind the staff.

[0011] The intelligent monitoring terminal receives and displays the micro-nozzle clogging status information sent by the remote control unit.

[0012] Furthermore, the data processing unit is deployed on a cloud platform, and the data processing unit communicates with the remote control unit in a two-way manner. The image acquisition unit obtains instructions from the remote control unit through the data processing unit, and then transmits the acquired spray area image to the data processing unit on the cloud platform.

[0013] Furthermore, the data processing unit includes image data, a data processing unit, a feature storage area, a data decision unit and data processing results, wherein the image data is the input data information, the data processing unit is responsible for data processing, recognition and feature extraction, and stores the extracted image features in the feature storage area, and then obtains the data processing results through the calculation of the data decision unit.

[0014] Furthermore, the feature storage area stores feature data of N pictures. When the feature data of a picture is to be stored in QN, the data in Q1 is first deleted, and the original picture feature data is moved in sequence, that is, Q2 is moved into Q1,..., QN is moved into QN-1, and then the feature data of the picture is stored in QN.

[0015] In another aspect, the present invention provides a micro-nozzle clogging monitoring method based on image processing, comprising the following steps:

[0016] S1, acquiring an image of the spray area generated by the micro nozzle;

[0017] S2. Processing the spray area image to obtain a data processing result;

[0018] S3. Sending the blockage data of the micro-nozzle in the data processing result to the intelligent monitoring terminal.

[0019] Furthermore, step S2 includes the following steps:

[0020] S21, processing the spray area image by a data processing unit to obtain image feature data;

[0021] S22, storing the image feature data in a feature storage area;

[0022] S23, reading the image feature data in the feature storage area to a data decision unit;

[0023] S24. The data decision unit determines the characteristic data and obtains a data processing result.

[0024] Furthermore, the step S21 includes the following steps:

[0025] S211, removing hue and saturation from the collected image to obtain a grayscale image retaining only brightness;

[0026] S212, using an edge extraction algorithm to extract edge image features of the spray area;

[0027] S213, performing filtering processing on the image in the spray area;

[0028] S214, extracting the size of the fog droplets on the image one by one according to the grayscale threshold, and calculating the area of ​​each fog droplet on the image (in terms of the number of pixels);

[0029] S215: Classify the extracted droplets according to the droplet area threshold, connect the droplets larger than the droplet area threshold and smaller than the droplet area threshold to obtain a first droplet combination pattern and a second droplet combination pattern, and calculate the areas of the first droplet combination pattern and the second droplet combination pattern and the maximum height of the droplet combination pattern respectively;

[0030] The grayscale threshold in step S214 can be determined by the following method:

[0031] The grayscale value that can minimize the intra-class variance of the pixels in the spray area image after thresholding is selected as the grayscale threshold, and the grayscale threshold is a positive integer.

[0032] A "class" is a collection of pixel values ​​with similar characteristics or attributes; "intra-class variance" refers to the degree of variation between pixel values ​​within the same class. If the pixel values ​​within a class vary widely, the pixel values ​​within that class will be more dispersed. Conversely, if the pixel values ​​vary less, the pixel values ​​within that class will be more concentrated.

[0033] Furthermore, the step S211 further includes: expanding the pixels of the grayscale image to the entire grayscale range of [0, 255].

[0034] Furthermore, the step S24 includes the following steps:

[0035] S241, arranging the N feature data sets in descending order based on the maximum height value of the first droplet combination pattern to obtain a sorted new feature data set;

[0036] S242. Calculate the area threshold of the first droplet combination pattern using the new feature data set. The calculation formula is as follows:

[0037]

[0038] Among them, S1 threshold is the calculated area threshold of the first droplet combination pattern, S1, S2, ..., SN are the area data of 1-N first droplet combination patterns in the new feature data set arranged in descending order according to the maximum height value of the first droplet combination pattern, and N is a positive integer greater than 3.

[0039] S243. If the area of ​​the first droplet combination pattern stored in the feature storage area QN is smaller than the S1 threshold, the monitoring system determines that the micro-nozzle is in a blocked state; otherwise, the monitoring system determines that the micro-nozzle is in a normal state.

[0040] In the step S21, the following method may also be used:

[0041] Construct an image unit grid and mark the grayscale value inside the unit grid. Define the starting position of the nozzle outlet in the micro-nozzle spray area image as the coordinate origin (0, 0). The range direction of the spray area image is the X-axis coordinate, and the direction perpendicular to the X axis in the same plane is the Y-axis coordinate. In the X and Y directions, the unit grid position coordinates in the spray area image are the center point P(i, j) (Xi, Yj) where the unit grid intersects in the X and Y directions, where i and j are the unit grid ordinal numbers of the marked position relative to the coordinate origin.

[0042] Using the image grayscale value recognition algorithm, the grayscale values ​​inside the different unit grids of the micro-nozzle spray area image are marked and recorded as g(i, j). The grayscale value of its Y-direction neighbor is g(i, j+1), and so on.

[0043] Calculate the cumulative change rate α of the grayscale values ​​of all adjacent pixels in the unit grid:

[0044]

[0045] Where n is the total number of pixels in the unit grid;

[0046] Obtain the cumulative change rate α of the unit grid in the Y direction and calculate the congestion change rate β:

[0047]

[0048] Where m is the total number of all selected Y-axis unit grids; is the cumulative change rate of the i-th unit grid;

[0049] The β value is used to determine the clogging degree of the micro-nozzle so as to better monitor the clogging trend. At the same time, it enhances the stability of the image feature data and avoids the influence of extreme values ​​on data calculation.

[0050] The threshold is set according to the industry-level experience value of micro-nozzle monitoring. When the blockage change rate β is greater than the threshold, it indicates that the grayscale change rate of the unit grid in the Y direction of the image is greater than the normal grayscale change rate, indicating that the micro-nozzle is blocked and requires timely intervention. The larger the β value, the more serious the blockage.

[0051] Compared with the prior art, the present invention has the following beneficial effects:

[0052] The present invention proposes a micro-nozzle blockage monitoring system and method based on image processing, which solves the current problem of real-time monitoring of the entire process of micro-nozzles from the beginning of blockage to complete blockage in actual work. The system has high working efficiency and can provide real-time feedback of monitored image data so that staff can intervene in the working conditions of the micro-nozzles in time to prevent the micro-nozzles from being completely blocked, thereby avoiding possible serious damage to people's production and life.

[0053] The present invention proposes to collect and process images of the micro-nozzle spray area, and judge whether the micro-nozzle is in the initial blockage state or the serious blockage state through the dynamic changes of the characteristic data of the continuous images, thereby improving the accuracy of micro-nozzle blockage monitoring.

[0054] In the image processing of the micro-nozzle spray area, the present invention adopts the method of grading and classifying the droplet sizes in the identified spray area, predicting the working status of the micro-nozzle according to the areas where the droplets of different categories are located, and quantifying the blockage of the micro-nozzle that cannot be judged in real time, which is conducive to accurately monitoring the working condition of the micro-nozzle.

[0055] In the data processing process of graded droplets, the present invention takes the brightness data of clustered pixels as the core and determines the droplet size by expanding to the periphery of the clustered pixels. This method can accurately determine the droplet size while consuming less processor resources and effectively eliminate background noise introduced when the camera captures images in different weather and environments.

[0056] The present invention uses the central pixel of a single droplet after classification as the endpoint for graphic drawing to obtain the shapes of droplets of different classifications. According to the dynamic change trend of the shapes of droplets of different classifications, the diagnosis and prediction of the working status of the micro-nozzle are realized, thereby further improving the accuracy of micro-nozzle blockage monitoring.

[0057] The present invention also proposes to use the areas where the shapes of different classified droplets are located, the areas of the graphics of different classified droplets and the corresponding area thresholds as the basis for judging whether the micro-nozzle is blocked. This can not only effectively monitor the working status of the micro-nozzle, but also more quickly predict the blockage status of the micro-nozzle.

[0058] In addition, the advantages of additional aspects of the present invention will be partially given in the following description or obtained through the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] The accompanying drawings, which constitute 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 improper limitations on the present invention.

[0060] Figure 1 Block diagram of the micro-nozzle clogging monitoring system based on image processing in the embodiment;

[0061] Figure 2 System block diagram of a micro-nozzle clogging monitoring system based on image processing in an embodiment in which the data processing unit is placed in the cloud;

[0062] Figure 3 Flowchart of the data processing method in the micro-nozzle clogging monitoring system based on image processing in the embodiment;

[0063] Figure 4 Schematic diagram of the working of the data processing unit in the embodiment;

[0064] Figure 5 One of the implementation and installation methods of the image acquisition unit hardware in the embodiment;

[0065] Figure 6 The second implementation and installation method of the hardware of the image acquisition unit in the embodiment;

[0066] Figure 7 The third implementation and installation method of the image acquisition unit hardware in the embodiment.

[0067] In the figure: 1. Image acquisition unit; 2. Orifice unit; 3. Nozzle unit. DETAILED DESCRIPTION

[0068] It should be noted that the following detailed descriptions are exemplary and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present invention belongs.

[0069] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0070] In the absence of conflict, the embodiments of the present invention and the features thereof may be combined with each other.

[0071] Example 1

[0072] This embodiment discloses a micro nozzle clogging monitoring system based on image processing. Figure 1 The monitoring system includes: an image acquisition unit, a data processing unit, a remote control unit and an intelligent monitoring terminal.

[0073] The image acquisition unit acquires a real-time image of the working state of the micro-nozzle, transmits the data of the real-time image to the data processing unit, and receives instructions from the remote control unit at the same time;

[0074] The data processing unit receives the data of the real-time image, processes the data through the data processing unit, determines whether the current micro-nozzle is in a blocked state, and transmits the data processing result to the remote control unit;

[0075] On the one hand, the remote control unit sends control instructions to the image acquisition unit; on the other hand, it receives the data processing results sent by the data processing unit, and sends the micro-nozzle blockage information in the data processing results in text form to the intelligent monitoring terminal to remind the staff.

[0076] The intelligent monitoring terminal receives and displays the micro-nozzle clogging status information sent by the remote control unit.

[0077] Specifically, there may be one or more intelligent monitoring terminals, and the terminal device may be a computer, a mobile phone, a tablet, etc.

[0078] Optionally, the data processing unit is deployed on a cloud platform, such as Figure 2 As shown. Figure 2It can be seen that the data processing unit located on the cloud platform communicates with the remote control unit in a two-way manner, and the image acquisition unit obtains the remote control unit instructions through the data processing unit, and then transmits the acquired spray area image directly to the data processing unit on the cloud platform. At this time, the hardware implementation of the image acquisition unit can be designed to be more compact and can be installed at the micro-nozzle or in a position perpendicular to the forward direction of the spray area. In order to clearly explain the positional relationship between the hardware of the image acquisition unit and the micro-nozzle, nozzle, and spray area, Figure 5-7 The relative positional relationships of the image acquisition unit 1, the nozzle unit 2, and the nozzle unit 3 are given as examples. It can be clearly seen from the direction of movement of the liquid entering the nozzle unit 2 that Figure 5 It is the vertical direction of the spray area generated by the nozzle part 2 when the image acquisition part 1 is installed. Figure 6 The image acquisition unit 1 is mounted on the outer edge of the nozzle unit 3. Figure 7 The image acquisition part 1 is installed through the opening, and the opening is located between the nozzle part 3 and the nozzle hole part 2.

[0079] Back to Figure 1 and Figure 2 The detailed composition of the data processing unit is as follows Figure 4 As shown, it specifically includes image data, a data processing unit, a feature storage area, a data decision unit and a data processing result, wherein the image data is the input data information; the data processing unit is responsible for data processing, recognition and feature extraction, and stores the extracted image features in the feature storage area; then the data processing result is obtained through the calculation of the data decision unit; the data processing result will be sent to the remote control unit.

[0080] It is worth mentioning that the feature storage area stores feature data for N consecutive images. When feature data for a new image is to be stored in QN, the data in Q1 is first deleted. The feature data of the original image is then shifted sequentially, i.e., Q2 is shifted into Q1, ..., QN is shifted into QN-1, and then the feature data of the new image is stored in QN. N is a positive integer greater than 3.

[0081] Example 2

[0082] This embodiment discloses a micro-nozzle clogging monitoring method based on image processing, comprising the following steps:

[0083] S1, acquiring an image of the spray area generated by the micro nozzle;

[0084] S2. Processing the spray area image to obtain a data processing result;

[0085] S3. Sending the blockage data of the micro-nozzle in the data processing result to the intelligent monitoring terminal.

[0086] Furthermore, Figure 3 As shown, step S2 further includes the following processing steps:

[0087] S21, processing the spray area image by a data processing unit to obtain image feature data;

[0088] S22, storing the image feature data in a feature storage area;

[0089] S23, reading the image feature data in the feature storage area to a data decision unit;

[0090] S24. The data decision unit determines the characteristic data and obtains a data processing result.

[0091] Preferably, step S2 is completed in a data processing unit, and the processing flow of the data processing unit is as follows: Figure 4 shown.

[0092] It is worth mentioning that the data processing unit in step S21 processes the collected image data and extracts data features, and can also preferably adopt the following steps:

[0093] S211, removing hue and saturation from the collected image to obtain a grayscale image retaining only brightness;

[0094] S212, using an edge extraction algorithm to extract edge image features of the spray area;

[0095] S213, performing filtering processing on the image in the spray area;

[0096] S214, extracting the size of the fog droplets on the image one by one according to the grayscale threshold, and calculating the area of ​​each fog droplet on the image (in terms of the number of pixels);

[0097] S215. The extracted droplets are sorted according to a droplet area threshold, and droplets larger than and smaller than the droplet area threshold are connected to form a first droplet combination pattern and a second droplet combination pattern, respectively. The areas of the first droplet combination pattern and the second droplet combination pattern, as well as the maximum heights of the droplet combination patterns, are calculated. The droplet area threshold is provided by the micronozzle manufacturer when calibrating the spray area shape characteristics of such nozzles. It is typically determined by taking the liquid droplet area at the boundary between the liquid droplet and the mist droplet as the moving liquid exits the micronozzle.

[0098] Furthermore, between step S211 and step S212, the method further includes: expanding the pixels of the grayscale image to the entire grayscale range of [0, 255] to enhance the contrast of the grayscale image, so that step S213 can more accurately eliminate background noise in the spray area.

[0099] The grayscale threshold in step S214 is determined according to one of the following two methods:

[0100] (1) The grayscale value that can minimize the intra-class variance of the pixels in the spray area image after thresholding is selected as the grayscale threshold, and the grayscale threshold is a positive integer.

[0101] A "class" is a collection of pixel values ​​with similar characteristics or attributes; "intra-class variance" refers to the degree of variation between pixel values ​​within the same class. If the pixel values ​​within a class vary widely, the pixel values ​​within that class will be more dispersed. Conversely, if the pixel values ​​vary less, the pixel values ​​within that class will be more concentrated.

[0102] (2) Extracting the pixel value range of the area other than the spray area in the grayscale image; then using the edge extraction algorithm to extract the pixel average value of the edge of a single droplet in the spray area as the pixel value set of the droplet edge; finally, taking the pixel grayscale value that can distinguish the pixel value range and the pixel value set as the grayscale threshold.

[0103] For example, in the above method (2), when the pixel value range and the pixel value set do not overlap, the pixel grayscale value at the maximum sum of the absolute values ​​of the differences between the pixel grayscale value and the pixel value range is taken as the grayscale threshold; when the pixel value range and the pixel value set overlap, the pixel grayscale value at the minimum variance between the pixel grayscale value and the pixel value set is taken as the grayscale threshold. The values ​​of the pixel value set and the pixel value range are both positive integers, and the pixel grayscale value is a positive integer.

[0104] Furthermore, the area of ​​the single droplet in step S214 is measured in pixels, which can more accurately monitor the quality of the spray area generated by the micro-nozzle, while also reducing the complexity of data processing and the computing resources required of the processor.

[0105] Returning to step S21 , the image feature data is the area of ​​the first and second mist drop combination patterns and the maximum height value of the mist drop combination pattern in step S215 .

[0106] Furthermore, the feature storage area in step S22 includes N storage areas, where N is a positive integer greater than 3, such as Figure 4 shown.

[0107] In step S23, the image feature data in the feature storage area is read to the data decision unit, specifically, the feature data of N images, QN, QN-1, ..., Q1, in the feature storage area are read to the data decision unit.

[0108] It is worth noting that the event that triggers the execution of step S23 is when the data processing unit generates new image feature data and completes writing the new image feature data into the feature storage area.

[0109] Exemplarily, the decision-making steps of step S24 are described by taking the first mist droplet as an example:

[0110] S241, arranging the N feature data sets in descending order based on the maximum height value of the first droplet combination pattern to obtain a sorted new feature data set;

[0111] S242. Calculate the area threshold of the first droplet combination pattern using the new feature data set. The calculation formula is as follows:

[0112]

[0113] Among them, S1 threshold is the calculated area threshold of the first droplet combination pattern, S1, S2, ..., SN are the area data of 1-N first droplet combination patterns in the new feature data set arranged in descending order according to the maximum height value of the first droplet combination pattern, and N is a positive integer greater than 3.

[0114] S243. If the area of ​​the first droplet combination pattern stored in the feature storage area QN is smaller than the S1 threshold, the monitoring system determines that the micro-nozzle is in a blocked state; otherwise, the monitoring system determines that the micro-nozzle is in a normal state.

[0115] Similarly, the area threshold of the second mist droplet combination pattern can be obtained using the area and maximum height of the second mist droplet combination pattern.

[0116] Preferably, if Figure 6 or Figure 7 The hardware of the image acquisition unit is at an oblique angle relative to the spray area image, which will cause the droplet image in the first droplet combination pattern area to be distorted. The area of ​​the new second droplet combination pattern can be determined based on the area threshold of the second droplet combination pattern.

[0117] Furthermore, when the area of ​​the first droplet combination pattern stored in QN of the feature storage area is between 30% and 50% of the area threshold of the first droplet combination pattern, it is necessary to determine whether the area of ​​the second droplet combination pattern is less than 60% of the area threshold of the second droplet combination pattern; if so, it is determined that the micro-nozzle is in a serious blockage state, and an alarm message is sent to the remote control unit and to the intelligent monitoring terminal to notify the staff to perform manual intervention; if not, it is determined that the micro-nozzle is in a blockage state, and a new image needs to be collected to determine whether it is in a serious blockage state.

[0118] When the area of ​​the first droplet pattern stored in the feature storage area QN is less than or equal to 30% of the area threshold of the first droplet pattern, the micro-nozzle is determined to be severely clogged, and personnel are notified for manual intervention. Because the feature data stored in the feature storage area QN is used to calculate the area threshold of the droplet pattern, the present invention effectively eliminates noise during image extraction while maintaining monitoring accuracy.

[0119] Preferably, to ensure the normal operation of the micronozzle, when the micronozzle is initially clogged, manual intervention is recommended to restore the micronozzle to normal operation. When the micronozzle is severely clogged, personnel should manually intervene to ensure the micronozzle returns to normal operation. The manual intervention method includes, but is not limited to, adjusting the liquid pressure and / or gas pressure at the micronozzle inlet to restore the micronozzle to normal operation. During the manual intervention process, the remote control unit notifies the image acquisition unit, data processing unit, and intelligent monitoring terminal to stop monitoring the micronozzle spray area.

[0120] Optionally, in the step S21, the following method may be preferably used:

[0121] A micronozzle image unit grid is constructed and the grayscale value inside the unit grid is marked. By calculating the cumulative change of the grayscale value of the unit grid and the grayscale values ​​of all its adjacent grids, the grayscale feature change rate is determined to identify whether the micronozzle is blocked.

[0122] Unit grid construction involves dividing the micro-nozzle spray area image into unit grids of equal size. Each unit grid can be a single pixel or a collection of multiple pixels. A loop can be used to traverse the image pixels and assign them to the corresponding unit grid. The size of the collection depends on the application requirements.

[0123] Specifically, the starting position of the nozzle outlet in the micro-nozzle spray area image is defined as the coordinate origin (0, 0), the range direction of the spray area image is the X-axis coordinate, and the direction perpendicular to the X in the same plane is the Y-axis coordinate. In the X and Y directions, the unit grid position coordinates in the spray area image are the center points where the unit grids intersect in the X and Y directions, recorded as P(i, j) (Xi, Yj), where i and j are the unit grid ordinal numbers of the position point where the mark is located relative to the coordinate origin. Correspondingly, the position coordinates of the Y-axis neighborhood are P(i, j+1) (Xi, Yj +1).

[0124] The pixel grayscale values ​​of the micro-nozzle blocked spray area image may have the following characteristics along the vertical Y-axis:

[0125] (1) The grayscale value of the blocked area is higher: Since the blockage will cause the liquid in the spray area to be unable to spray normally, the grayscale value of the pixels in the blocked area may be relatively high because the spray material cannot cover these pixels.

[0126] (2) Grayscale value changes are relatively gentle: For pixels near the blocked area, the grayscale value changes may be relatively gentle due to the blockage of the spray material, because the spray material cannot diffuse normally to these pixels.

[0127] (3) The grayscale values ​​of the blocked area and the non-blocked area are obviously contrasted: There may be an obvious contrast in the grayscale values ​​of the blocked area and the non-blocked area, because the blockage will cause the uneven distribution of the spray material, thus forming obvious regional differences in the image.

[0128] The shape and area of ​​each unit grid can be equal. Preferably, a rectangular unit grid is recommended. Based on the Y-axis pixel grayscale characteristics of the micro-nozzle blocked spray area image and the grayscale processing time, the pixel size of the unit grid is preferably 1×1, 1×3, 1×5, 3×3, 3×5, or 5×5.

[0129] The image grayscale value recognition algorithm is used to mark the grayscale values ​​inside different unit grids of the micro-nozzle spray area image, which are recorded as g(i, j). Correspondingly, the grayscale value of its Y-direction neighborhood is g(i, j+1), and so on.

[0130] The cumulative change rate α of the grayscale values ​​of all adjacent pixels within the unit grid is calculated according to the following formula:

[0131]

[0132] Where n is the total number of pixels in the unit grid;

[0133] Obtain the cumulative change rate α of the unit grid in the Y direction and calculate the congestion change rate β:

[0134]

[0135] Where m is the total number of all selected Y-axis unit grids; is the cumulative change rate of the i-th unit grid;

[0136] The β value is used to determine the clogging degree of the micro-nozzle so as to better monitor the clogging trend. At the same time, it enhances the stability of the image feature data and avoids the influence of extreme values ​​on data calculation.

[0137] The threshold is set according to the industry-level experience value of micro-nozzle monitoring. When the blockage change rate β is greater than the threshold, it indicates that the grayscale change rate of the unit grid in the Y direction of the image is greater than the normal grayscale change rate, indicating that the micro-nozzle is blocked and requires timely intervention. The larger the β value, the more serious the blockage.

[0138] In the above embodiments, “greater than” includes the cases of being greater than or equal to; “less than” includes only the case of being less than but does not include the case of being equal to.

[0139] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A micro nozzle clogging monitoring system based on image processing, characterized in that: The monitoring system includes: an image acquisition unit, a data processing unit, a remote control unit and an intelligent monitoring terminal; wherein, The image acquisition unit acquires a real-time image of the working state of the micro-nozzle, transmits the data of the real-time image to the data processing unit, and receives instructions from the remote control unit at the same time; The data processing unit receives the data of the real-time image, processes the data through the data processing unit, determines whether the current micro-nozzle is in a blocked state, and transmits the data processing result to the remote control unit; The remote control unit sends a control instruction to the image acquisition unit on the one hand; on the other hand, receives the data processing result sent by the data processing unit, and sends the micro-nozzle blockage status in the data processing result in the form of text to the intelligent monitoring terminal; The intelligent monitoring terminal receives and displays the micro-nozzle clogging status information sent by the remote control unit; The data processing unit includes image data, a data processing unit, a feature storage area, a data decision unit, and a data processing result; wherein the image data is the input data information, the data processing unit is responsible for data processing, recognition and feature extraction, and stores the extracted image features in the feature storage area, and then obtains the data processing result through the calculation of the data decision unit; The feature storage area stores feature data of N consecutive images. When feature data of a new image is to be stored in Q N When Q1 is first deleted, the original image feature data is moved in sequence, that is, Q2 is moved into Q1, ..., Q N Move into Q N-1 Then store the feature data of the new image into Q N； The micro nozzle clogging monitoring method based on image processing includes the following steps: S1, acquiring an image of the spray area generated by the micro nozzle; S2. Processing the spray area image to obtain a data processing result; S3, sending the blockage data of the micro-nozzle in the data processing result to the intelligent monitoring terminal; The step S2 comprises the following steps: S21, processing the spray area image by a data processing unit to obtain image feature data; S22, storing the image feature data in a feature storage area; S23, reading the image feature data in the feature storage area to a data decision unit; S24, the data decision unit judges the characteristic data and obtains a data processing result; The step S21 includes the following steps: S211, removing hue and saturation from the collected image to obtain a grayscale image retaining only brightness; S212, using an edge extraction algorithm to extract edge image features of the spray area; S213, performing filtering processing on the image in the spray area; S214, extracting the size of the fog droplets on the image one by one according to the grayscale threshold, and calculating the area of ​​each fog droplet on the image, the area being measured in pixels; S215: Classify the extracted droplets according to the droplet area threshold, connect the droplets larger than the droplet area threshold and smaller than the droplet area threshold to obtain a first droplet combination pattern and a second droplet combination pattern, and calculate the areas of the first droplet combination pattern and the second droplet combination pattern and the maximum height of the droplet combination pattern respectively; The grayscale threshold in step S214 can be determined by the following method: Selecting a grayscale value that can minimize the intra-class variance of pixels in the spray area image after thresholding as the grayscale threshold, wherein the grayscale threshold is a positive integer; The step S211 further includes: expanding the pixels of the grayscale image to the entire grayscale range of [0, 255]; The step S24 includes the following steps: S241, arranging the N feature data sets in descending order based on the maximum height value of the first droplet combination pattern to obtain a sorted new feature data set; S242. Calculate the area threshold of the first droplet combination pattern using the new feature data set. The calculation formula is as follows: S 1阈= ； Among them, S 1阈 is the calculated area threshold of the first droplet combination pattern, S1, S2, ..., S N is the area data of 1-N first mist droplet combination patterns in the new feature data set arranged in descending order of the maximum height value of the first mist droplet combination pattern, where N is a positive integer greater than 3; S243, the feature storage area Q N If the area of ​​the first droplet combination pattern stored in is less than S 1阈 , the monitoring system determines that the micro nozzle is in a blocked state, otherwise it determines that the micro nozzle is in a normal state.

2. The micro-nozzle clogging monitoring system according to claim 1, characterized in that: The data processing unit is deployed on the cloud platform, and the data processing unit communicates with the remote control unit in a two-way manner. The image acquisition unit obtains remote control unit instructions through the data processing unit, and then transmits the acquired spray area image to the data processing unit on the cloud platform.

Images