CN116523883B - Intelligent latex viscosity control method based on image processing - Google Patents
Intelligent latex viscosity control method based on image processing Download PDFInfo
- Publication number
- CN116523883B CN116523883B CN202310501453.1A CN202310501453A CN116523883B CN 116523883 B CN116523883 B CN 116523883B CN 202310501453 A CN202310501453 A CN 202310501453A CN 116523883 B CN116523883 B CN 116523883B
- Authority
- CN
- China
- Prior art keywords
- latex
- image
- edge
- viscosity
- frame image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 239000004816 latex Substances 0.000 title claims abstract description 183
- 229920000126 latex Polymers 0.000 title claims abstract description 183
- 238000000034 method Methods 0.000 title claims abstract description 63
- 238000012545 processing Methods 0.000 title claims abstract description 32
- 239000000463 material Substances 0.000 claims abstract description 16
- 238000004458 analytical method Methods 0.000 claims abstract description 10
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims abstract description 9
- 238000003708 edge detection Methods 0.000 claims abstract description 7
- 239000000839 emulsion Substances 0.000 claims description 41
- 238000004364 calculation method Methods 0.000 claims description 8
- 238000005070 sampling Methods 0.000 claims description 7
- 239000002994 raw material Substances 0.000 claims description 6
- OSVFKEXOKOAFNU-UHFFFAOYSA-N NC=B Chemical compound NC=B OSVFKEXOKOAFNU-UHFFFAOYSA-N 0.000 claims description 3
- 230000001105 regulatory effect Effects 0.000 abstract description 2
- 239000000853 adhesive Substances 0.000 abstract 1
- 230000001070 adhesive effect Effects 0.000 abstract 1
- 238000001514 detection method Methods 0.000 description 11
- 238000002360 preparation method Methods 0.000 description 10
- 238000003756 stirring Methods 0.000 description 6
- 239000003292 glue Substances 0.000 description 4
- 238000004519 manufacturing process Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 238000004891 communication Methods 0.000 description 2
- 238000007598 dipping method Methods 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/187—Segmentation; Edge detection involving region growing; involving region merging; involving connected component labelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/06—Recognition of objects for industrial automation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/07—Target detection
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Quality & Reliability (AREA)
- Geometry (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
Abstract
The application relates to an intelligent latex viscosity control method based on image processing, which is characterized in that an image in a latex dripping process is acquired, connected domain analysis, gray level processing and edge detection are carried out to obtain an edge image of a critical frame image, latex dripping flow velocity is obtained according to the change of the bottommost pixel position in all images in the dripping process, the data such as the connected domain length and width in the edge image of the critical frame image, the pixel change slope of the edge of a latex dripping head and the like are analyzed to obtain latex extensibility and latex dripping roundness, latex viscosity is obtained according to the latex extensibility, the latex dripping roundness and the latex dripping flow velocity, whether a material needs to be regulated is judged by using latex adhesive viscosity, if the viscosity is moderate, no regulation is carried out, if the viscosity is thick, water is added, if the viscosity is thin, latex is added based on image processing, intelligent latex viscosity control is realized, and product quality is ensured.
Description
Technical Field
The application relates to the field of artificial intelligence, in particular to an intelligent latex viscosity control method based on image processing.
Background
In the glove processing production process, emulsion is required to be continuously consumed, however, if the emulsion viscosity difference between front and back preparation is large, and the glue fluidity of the semi-dipped glove is small, the new emulsion is added, if the new emulsion is not stirred, layering phenomenon can be generated, and the thickness of the glove glue layer produced by front and back processing is inconsistent, so that the newly prepared emulsion is required to be controlled to be basically consistent with the emulsion viscosity prepared before in advance, in the actual feeding process, the instability exists only by human vision measurement, and great deviation exists, so that the inaccuracy of a detection result can influence the product quality, and the fact that the proper emulsion viscosity is repeatedly detected after feeding is an important premise of all production processes of the dipped glove is required, and the improper emulsion viscosity can influence the effect of a series of subsequent production processes.
Disclosure of Invention
The application provides an intelligent latex viscosity control method based on image processing, which aims to solve the problem of error caused by instability in the existing manual detection of latex viscosity, and adopts the following technical scheme:
step one: collecting each frame of image in the latex dripping process;
step two: carrying out connected domain analysis on each frame image to obtain a critical frame image, and carrying out gray processing and edge detection on the critical frame image to obtain an edge image of the critical frame image;
step three: according to the length of a long axis of a connected domain in an edge image of the critical frame image, the average width of the connected domain and the time feature value of emulsion breaking, the emulsion extensibility is obtained;
step four: extracting latex drip heads in edge images of the critical frame images as an interested region, and obtaining latex drip roundness according to the slope change rate of edge pixel points of the interested region and the width-to-length ratio of the interested region;
step five: obtaining latex drop flow rate according to the position change of the pixel at the bottommost end of the latex drop in the process from the initial frame image to the critical frame image and the sampling frequency of the camera;
step six: obtaining latex viscosity according to the latex extensibility, the latex drop roundness and the latex drop flow rate;
step seven: and comparing the latex viscosity with a set range, and adjusting the addition amount of the raw materials for preparing the latex if the latex viscosity exceeds the set range.
The latex viscosity calculating method comprises the following steps:
NC＝B 1 B 2 e -v
wherein NC is latex viscosity, B 1 For latex extensibility, B 2 The latex drop roundness, v is the latex drop flow rate, and e is a natural constant.
The latex extensibility calculating method comprises the following steps:
in the method, in the process of the application,for the time characteristic value of emulsion break, i is the critical frame image n i I.e., the i-th frame in the n-frame images is a critical frame image, i=1, 2..n, n is the total number of frame images in the latex dripping process, L is the length of the long axis of the connected domain of the edge image of the critical frame image, and W is the average width of the connected domain of the edge image of the critical frame image.
The critical frame image acquisition step comprises the following steps:
carrying out connected domain analysis on each frame of image by using a seed filling method to obtain connected domains with different labels, and using the maximum label number to represent the number of the connected domains in the image to obtain a connected domain number sequence;
when the difference value of two adjacent data in the connected domain number sequence is 1, the frame image corresponding to the previous data is a critical frame image.
The length of the long axis of the connected domain and the average width of the connected domain are obtained by the steps of:
obtaining a pixel width sequence of G= { d according to the pixel width between each row of edge pixel points in the connected domain of the edge image of the critical frame image 1 ，d 2 ，...，d J J is the number of data of G, and the minimum pixel width of G is d min It was used as the stable width d of the latex m I.e. the pixel width of the m-th row in G;
removing unsatisfied d in G m -δ＜d a ＜d m The data of +delta and the edge pixels of the corresponding row of the data in the edge image of the critical frame image obtain a new pixel width sequence G' = { D 1 ，D 2 ，...，D J′ An image in which the edge image of the critical frame image has been removed from the upper edge pixels, wherein d a The pixel width of the a-th row in G is equal to or more than 1 and less than m, delta is an empirical value, and d is set m /10；
The long axis length of the connected domain in the image obtained by removing the upper edge pixels from the edge image of the critical frame image is used as the long axis length L of the connected domain;
the average width W of the connected domain is calculated by:
wherein D is b B=1, 2, 3..j ', J ' is the number of data for G ', D max Is the maximum pixel width in G'.
The latex drop roundness calculation step is as follows:
the d is satisfied in the removal sequence G m -δ＜D b ＜d m Data of +delta and edge pixels of corresponding rows of the data in an edge image of a critical frame image, resulting in a pixel width sequence G' 1 And the image after the edge image of the critical frame image has removed the upper and middle edge pixels, the image is an emulsion drip head edge image, and is positioned at the lower part of the edge image of the critical frame image, wherein m is more than b and less than or equal to J';
the latex drip edge was used as the region of interest, B (x, y max ) The coordinate of the pixel point at the bottommost end of the edge pixel of the latex drip head has the largest ordinate and A (x) min Y) is the pixel point with the minimum abscissa, C (x) max Y) is the pixel point with the largest abscissa;
from point A, traversing edge pixel points one by one in the direction of increasing the ordinate of the 5X 5 sliding window, and calculating the ordinate most in the sliding windowThe slope between the small pixel point and the pixel point with the largest ordinate is obtained until the traversal to the point B is stopped, and a slope sequence { k } 1 ，k 2 ，...，k p }；
Obtaining a slope change rate K according to a slope sequence:
where c=1, 2,..p, p is the number of slopes in the slope sequence;
emulsion drop roundness B 2 The calculation method of (1) is as follows:
in the method, in the process of the application,e is a natural constant, which is the ratio of the width to the length of the latex drip chamber, and has a value of about 2.7.
The calculating step of the latex drop flow rate is as follows:
counting the ordinate value y of the pixel point with the largest ordinate in the latex edge in the process from the initial frame image to the critical frame image g Is changed to obtain the sequence { y } g1 ，y g2 ，...，y gi }；
The calculation method for obtaining the latex flow velocity v comprises the following steps:
where d=1, 2,..i-1, f is the sampling frequency of the camera.
The method for adjusting the addition amount of the raw materials for preparing the latex comprises the following steps:
obtaining the viscosity NC of emulsion drops after interval time h If NC 0 -t≤NC h ≤NC 0 +t, the latex viscosity is moderate, the systemStopping the adjustment of the amount of latex material;
if NC h ＜NC 0 T, the latex viscosity is thinner, the material quantity of the latex is increased, and the previous step is repeated after the latex is uniformly stirred;
if NC h ＞NC 0 +t, the latex viscosity is thicker, the material quantity of water is increased, and after being uniformly stirred, the previous step is repeated;
wherein NC (numerical control) 0 The standard value of viscosity and t is the adjustment value.
The beneficial effects of the application are as follows: according to the change condition in the latex dripping process, a method for detecting latex viscosity by comprehensively analyzing three factors of latex dripping flow rate, latex dripping roundness and latex extensibility is provided, corresponding adjustment is made according to detection results, the glue viscosity used by the latex dipping glove is judged based on image processing, intelligent control of the latex viscosity is realized, errors caused by manual detection are avoided, and the quality of glove dipping production is ensured.
Drawings
In order to more clearly illustrate the embodiments of the application or the technical solutions of the prior art, the drawings which are used in the description of the embodiments or the prior art will be briefly described, it being obvious that the drawings in the description below are only some embodiments of the application, and that other drawings can be obtained according to these drawings without inventive faculty for a person skilled in the art.
FIG. 1 is a schematic flow chart of an intelligent latex viscosity control method based on image processing;
FIG. 2a is a schematic diagram of a latex dripping state with thinner viscosity in an intelligent latex viscosity control method based on image processing according to the application;
FIG. 2b is a schematic diagram of a dripping state of a latex with a thicker viscosity in an intelligent latex viscosity control method based on image processing according to the present application;
FIG. 3 is a schematic view of the width between pixels of each row in an edge image of a critical frame image in an intelligent latex viscosity control method based on image processing according to the present application;
FIG. 4 is a schematic view of the pixel coordinates of the edge of the latex drip head in the intelligent latex viscosity control method based on image processing;
FIG. 5 is a schematic diagram of an embodiment of an intelligent latex viscosity control system based on image processing according to the present application;
Detailed Description
The following description of the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present application, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
An embodiment of an intelligent latex viscosity control method based on image processing of the present application, as shown in fig. 1, includes:
step one: collecting each frame of image in the latex dripping process;
the method aims at acquiring image data and carrying out DNN network segmentation to extract frame image data corresponding to latex drops.
Firstly, stopping stirring in a latex preparation tank, firstly stopping stirring in the latex preparation tank, extending a detection rod into the latex preparation tank, and continuously collecting images of latex dripping after lifting, wherein the images are shown in fig. 2a and 2b, the images in fig. 2a are in a latex dripping state with thinner viscosity, and the images in fig. 2b are in a latex dripping state with thicker viscosity.
Then, since the latex has various colors and the shape of latex drops is not completely the same, in order to adapt the system to various situations and enhance the generalization capability, the application adopts a DNN semantic segmentation mode to identify targets in segmented images.
The relevant description of the DNN network is as follows:
(1) The data set used is an image data set of the latex dripping process, and the latex dripping column is in various forms.
(2) Pixels that need to be segmented are divided into four categories: namely, the corresponding label marking process of the training set is that the single-channel semantic label, the label of the corresponding position pixel belonging to the background class is 0, the label belonging to the plane of the emulsion preparation groove is 1, the label belonging to the detection rod of the preparation groove is 2, and the label belonging to the dripping emulsion is 3.
(3) The task of the network is classification, so the loss function used is a cross entropy loss function.
It should be noted that, in the system, the image is not required to be acquired in real time, but the latex is required to be picked up by the detection rod to drip after the material amount is adjusted, a plurality of frames of images are continuously acquired, and the operation is repeated until the system considers that the viscosity of the latex is suitable at the moment.
Step two: carrying out connected domain analysis on each frame image to obtain a critical frame image, and carrying out gray processing and edge detection on the critical frame image to obtain an edge image of the critical frame image;
the purpose of the step is to perform connected domain analysis and edge detection on the latex drop image extracted in the step one to obtain an edge image of the critical frame image.
The specific steps of critical frame and image processing are as follows:
(1) Performing connected domain analysis on the obtained multi-frame latex dripping image by using Seed Filling method to obtain connected domains with different labels (label), and obtaining the maximum value x of the label number, wherein the value x is the number of connected domains in the image, thereby obtaining the connected domain number sequence { x ] in each frame image 1 ，x 2 ，...，x i ...，x n }；
(2) After the gray level of the connected domain image of each frame of image is processed, the Canny operator is used for edge detection, and the obtained gradient edge is the glue drop edge pixel;
(3) Calculating the difference between adjacent data in the sequence, if the difference between a certain data and the adjacent data is 1, the image corresponding to the data is a critical frame, namely c i ＝|x i -x i+1 I, when c i When=1, x i Corresponding frame image n i Namely, the critical frame image is obtained,x i represents the number of connected domains in the ith frame image, n i The i-th frame in the n-frame image is represented, and n is the total number of frame images.
Step three: according to the length of a long axis of a connected domain in an edge image of the critical frame image, the average width of the connected domain and the time feature value of emulsion breaking, the emulsion extensibility is obtained;
the purpose of this step is to obtain the overall extensibility of the emulsion by the width change information of the emulsion pixels in the edge image of the critical frame image, as shown in fig. 3, the area corresponding to the reference numeral 1 represents the upper part of the emulsion drop, the area corresponding to the reference numeral 2 represents the middle part of the emulsion drop, the area corresponding to the reference numeral 3 represents the lower part of the emulsion drop, and the area representing the head part of the emulsion drop, because the starting position of the emulsion drop on the upper part of the emulsion drop is the end of the detection rod, the larger emulsion width can be formed here, but the emulsion itself can not be formed, and the width information needs to be removed in order to ensure the accuracy of the result. The emulsion then forms a stable width in the middle, which is also the smallest width in the emulsion drop column.
Wherein the latex extensibility B 1 The calculation steps are as follows:
(1) Recording the pixel width between each row of edge pixel points in the connected domain of the edge image of the critical frame image to obtain a pixel width sequence of G = { d 1 ，d 2 ，...，d I J is the number of data of G, and the minimum pixel width of G is d min It was used as the stable width d of the latex m I.e. the pixel width of the m-th row in G;
(2) Removing unsatisfied d in G m -δ＜d a ＜d m The data of +delta and the edge pixels of the corresponding row of the data in the edge image of the critical frame image obtain a new pixel width sequence G' = { D 1 ，D 2 ，...，D J′ An image in which the edge image of the critical frame image has been removed from the upper edge pixels, wherein d a The pixel width of the a-th row in G is equal to or more than 1 and less than m, delta is an empirical value, and d is set m The number of data of J 'is G';
(3) The long axis length of the connected domain in the image obtained by removing the upper edge pixels from the edge image of the critical frame image is used as the long axis length L of the connected domain ；
(4) The average width W of the connected domain is calculated by:
wherein D is b B=1, 2, 3..j ', D for the pixel width of row b in G' max Is the maximum pixel width in G'.
(5) The latex extensibility can be expressed as:
in the method, in the process of the application,the time characteristic value of emulsion breaking indicates that the earlier and later the breaking of the drop column, the lower the extensibility of the drop column, the lower the emulsion viscosity, and the later the breaking, the higher the extensibility of the drop column, the higher the emulsion viscosity, i is the critical frame image n i I=1, 2..n, n being the total number of frame images during latex drip.
Step four: extracting latex drip heads in edge images of the critical frame images as an interested region, and obtaining latex drip roundness according to the slope change rate of edge pixel points of the interested region and the width-to-length ratio of the interested region;
the purpose of this step is to analyze the edge pixel variation of the latex drip head information in the critical frame to obtain the latex drip roundness B 2 。
Wherein, the emulsion drop roundness B 2 The acquisition steps are as follows:
(1) The area 2 is removed on the basis of the area 1 which is removed as shown in fig. 3, so as to obtain the area 3, namely, the upper and middle areas in the edge image of the critical frame image of the latex drip are removed, and only the lower latex drip head image is left, specifically, the method comprises the following steps:
the d is satisfied in the removal sequence G m -δ＜D b ＜d m Data of +delta and edge pixels of corresponding rows of the data in an edge image of a critical frame image, resulting in a pixel width sequence G' 1 And the image after the edge image of the critical frame image has removed the upper and middle edge pixels, the image is an emulsion drip head edge image, and is positioned at the lower part of the edge image of the critical frame image, wherein m is more than b and less than or equal to J';
(2) The latex drip edge was used as the region of interest, B (x, y max ) Is the bottom pixel point coordinate of the edge pixel of the latex drip head, A (x min Y) is the pixel point with the minimum abscissa, C (x) max Y) is the pixel point with the largest abscissa;
(3) As shown in fig. 4, from point a, the edge pixels are traversed one by one in the direction of increasing the ordinate of the 5×5 sliding window, the slope between the pixel with the smallest ordinate and the pixel with the largest ordinate in the sliding window is calculated, and the slope sequence { k ] is obtained until the traversal stops at point B 1 ，k 2 ，...，k p }；
Obtaining a slope change rate K according to a slope sequence:
where c=1, 2,..p, p is the number of slopes in the slope sequence;
(4) Roundness of latex drop B 2 The calculation method of (1) is as follows:
in the method, in the process of the application,is the ratio of the width to the length of the latex drip head,e is a natural constant and has a value equal to about 2.7.
Step five: obtaining latex drop flow rate according to the position change of the pixel at the bottommost end of the latex drop in the process from the initial frame image to the critical frame image and the sampling frequency of the camera;
the purpose of this step is to analyze the latex for flow rate from the beginning of the drop to the pixel position change during the off process.
The method comprises the following steps of:
(1) Counting the ordinate value y of the pixel point with the largest ordinate in the latex edge in the process from the initial frame image to the critical frame image g Is changed to obtain the sequence { y } g1 ，y g2 ，...，y gi }；
Since the velocity calculated from the images of the individual adjacent frames varies unevenly due to the influence of the air resistance, it is necessary to calculate the velocity of movement of the latex droplet head from the image between the start frame and the critical frame.
(2) According to the sampling frequency f, i is the critical frame image n i I=1, 2..n, n is the total number of frame images in the latex drip process, indicating that an image from the initial frame to the critical frame is acquired, the latex flow velocity v can be expressed as:
where d=1, 2,..i-1, f is the sampling frequency of the camera.
Step six: obtaining latex viscosity according to the latex extensibility, the latex drop roundness and the latex drop flow rate;
the purpose of this step is to synthesize latex extensibility B 1 Emulsion drop roundness B 2 And latex drop flow rate v to give latex viscosity NC as a criterion for detecting whether or not the latex needs to be adjusted for material analysis.
Wherein, the latex viscosity NC can be expressed as:
NC＝B 1 B 2 e -v
wherein e is a natural constant, and e has a value of about 2.7;
step seven: and comparing the latex viscosity with a set range, and adjusting the addition amount of the raw materials for preparing the latex if the latex viscosity exceeds the set range.
The purpose of this step is to adjust the material according to the latex viscosity obtained in step five, in this example, the first set latex viscosity NC h Taking the viscosity as a reference viscosity; setting an adjustment value t, representing the allowable range of latex viscosity, wherein the empirical value of t is 0.2;
the method for adjusting the addition amount of the raw materials for preparing the latex comprises the following steps:
(1) Judging the NC obtained above h Satisfy NC 0 -t≤NC h ≤NC 0 If the latex viscosity is satisfied, the latex viscosity is considered to be moderate, the system stops adjusting the latex material amount, otherwise, the next judgment is carried out, NC h Indicating the latex viscosity obtained after the lambda-th adjustment;
(2) If NC is less than NC h T, regarding that the latex viscosity is thinner, increasing the material amount of the latex, and repeating the step 1 after uniformly stirring;
(3) If NC > NC h And +t, regarding that the latex is thick, increasing the water material amount, and repeating the step 1 after uniformly stirring.
An intelligent latex viscosity control system based on image processing is provided, as shown in fig. 5, the intelligent latex viscosity control system comprises a latex preparation tank, a liftable detection rod, a camera and a control system, wherein a feed inlet and a water inlet are formed in the preparation tank, a rotating device is arranged in the preparation tank and used for stirring materials, the control system comprises an S100 image processing unit, an S101 calculating unit and an S103 control unit, firstly, the S103 control unit sends a signal to control rotation equipment to stop rotating, the detection rod descends to stretch into the preparation tank to dip latex materials, then ascends to restore the original position, the camera is controlled to collect each frame of image in the latex dripping process, all the collected images are sent to the S100 image processing unit, communication domain analysis, gray level processing and edge detection are carried out to obtain critical frame images and edge pixels of each frame of image, data such as the critical frame pixel width sequence, the communication domain length and width are sent to the S101 calculating unit, the whole extensibility, the latex drip roundness and the latex flow velocity are calculated, the latex viscosity is obtained according to the integral extensibility, the latex viscosity is compared with a set range, and if the viscosity is consistent, and if the viscosity is not consistent with the standard, the material is not regulated; if the viscosity is high, a water adding signal is sent to an S103 control unit, the S103 control unit controls the water inlet to add water, if the viscosity is low, a feeding signal is sent to the S103 control unit, and the S103 control unit controls the feeding port to add latex; after the water/latex is added, the rotating device is controlled to stir, and the detection steps are continuously repeated until the viscosity reaches the standard, so that the intelligent control on the latex viscosity adjustment is realized.
The foregoing description of the preferred embodiments of the application is not intended to be limiting, but rather is intended to cover all modifications, equivalents, alternatives, and improvements that fall within the spirit and scope of the application.
Claims (7)
1. An intelligent latex viscosity control method based on image processing is characterized by comprising the following steps:
collecting each frame of image in the latex dripping process;
carrying out connected domain analysis on each frame image to obtain a critical frame image, and carrying out gray processing and edge detection on the critical frame image to obtain an edge image of the critical frame image;
according to the length of a long axis of a connected domain in an edge image of the critical frame image, the average width of the connected domain and the time feature value of emulsion breaking, the emulsion extensibility is obtained;
extracting latex drip heads in edge images of the critical frame images as an interested region, and obtaining latex drip roundness according to the slope change rate of edge pixel points of the interested region and the width-to-length ratio of the interested region;
obtaining latex drop flow rate according to the position change of the pixel at the bottommost end of the latex drop in the process from the initial frame image to the critical frame image and the sampling frequency of the camera;
obtaining latex viscosity according to the latex extensibility, the latex drop roundness and the latex drop flow rate;
comparing the latex viscosity with a set range, and adjusting the addition amount of the raw materials for preparing the latex if the latex viscosity exceeds the set range;
the critical frame image acquisition step comprises the following steps:
carrying out connected domain analysis on each frame of image by using a seed filling method to obtain connected domains with different labels, and using the maximum label number to represent the number of the connected domains in the image to obtain a connected domain number sequence;
when the difference value of two adjacent data in the connected domain number sequence is 1, the frame image corresponding to the previous data is a critical frame image.
2. The intelligent latex viscosity control method based on image processing according to claim 1, wherein the latex viscosity calculation method is as follows:
NC＝B 1 B 2 e -v
wherein NC is latex viscosity, B 1 For latex extensibility, B 2 The latex drop roundness, v is the latex drop flow rate, and e is a natural constant.
3. The intelligent latex viscosity control method based on image processing according to claim 2, wherein the latex extensibility calculating method is as follows:
in the method, in the process of the application,for the time characteristic value of emulsion break, i is the critical frame image n i The ith frame in the n frame images is a critical frame image, i=1, 2 … n, n is latex dropThe total number of frame images in the process is L, the length of the long axis of the connected domain of the edge image of the critical frame image, and W, the average width of the connected domain of the edge image of the critical frame image.
4. The intelligent latex viscosity control method based on image processing according to claim 3, wherein the step of obtaining the length of the long axis of the connected domain and the average width of the connected domain is as follows:
obtaining a pixel width sequence of G= { d according to the pixel width between each row of edge pixel points in the connected domain of the edge image of the critical frame image 1 ,d 2 ,...,d J J is the number of data of G, and the minimum pixel width of G is d min It was used as the stable width d of the latex m I.e. the pixel width of the m-th row in G;
removing unsatisfied d in G m -δ<d a <d m The data of +delta and the edge pixels of the corresponding row of the data in the edge image of the critical frame image obtain a new pixel width sequence G' = { D 1 ,D 2 ,...,D J′ An image in which the edge image of the critical frame image has been removed from the upper edge pixels, wherein d a The pixel width of the a-th row in G is 1-a<m, delta is an empirical value, d m /10；
The long axis length of the connected domain in the image obtained by removing the upper edge pixels from the edge image of the critical frame image is used as the long axis length L of the connected domain;
the average width W of the connected domain is calculated by:
wherein D is b B=1, 2, 3..j ', J ' is the number of data for G ', D max Is the maximum pixel width in G'.
5. The intelligent latex viscosity control method based on image processing according to claim 4, wherein the step of calculating the latex drop roundness is:
the d is satisfied in the removal sequence G m -δ<D b <d m Data of +delta and edge pixels of corresponding rows of the data in an edge image of a critical frame image, resulting in a sequence of pixel widths G 1 The edge image of the' and critical frame images, from which the upper and middle edge pixels have been removed, is a latex drip edge image, which is located below the edge image of the critical frame image, where m<b≤J′；
The latex drip edge was used as the region of interest, B (x, y max ) The coordinate of the pixel point at the bottommost end of the latex drip edge image pixel has the largest ordinate value, A (x min Y) is the pixel point with the minimum abscissa, C (x) max Y) is the pixel point with the largest abscissa;
traversing edge pixel points one by one from a point A in the direction of increasing the ordinate of a 5X 5 sliding window, calculating the slope between the pixel point with the smallest ordinate and the pixel point with the largest ordinate in the sliding window until traversing to a point B to stop, and obtaining a slope sequence { k } 1 ,k 2 ,...,k p }；
Obtaining a slope change rate K according to a slope sequence:
wherein c=1, 2, …, p, p is the number of slopes in the slope sequence;
emulsion drop roundness B 2 The calculation method of (1) is as follows:
in the method, in the process of the application,is the ratio of the width to the length of the latex drip head.
6. The intelligent latex viscosity control method based on image processing according to claim 3, wherein the step of calculating the latex drop flow rate is as follows:
counting the ordinate value y of the pixel point with the largest ordinate in the latex edge in the process from the initial frame image to the critical frame image g Is changed to obtain the sequence { y } g1 ,y g2 ,...,y gi }；
The calculation method for obtaining the latex flow velocity v comprises the following steps:
wherein d=1, 2, …, i-1; f is the sampling frequency of the camera.
7. The intelligent latex viscosity control method based on image processing according to claim 1, wherein the method for adjusting the addition amount of the raw materials for preparing the latex is as follows:
obtaining the viscosity NC of emulsion drops after interval time h If NC 0 -t≤NC h ≤NC 0 +t, the latex viscosity is moderate, and the system stops the adjustment of the latex material amount;
if NC h <NC 0 T, the latex viscosity is thinner, the material quantity of the latex is increased, and the previous step is repeated after the latex is uniformly stirred;
if NC h >NC 0 +t, the latex viscosity is thicker, the material quantity of water is increased, and after being uniformly stirred, the previous step is repeated;
wherein NC (numerical control) 0 The standard value of viscosity and t is the adjustment value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310501453.1A CN116523883B (en) | 2023-05-06 | 2023-05-06 | Intelligent latex viscosity control method based on image processing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310501453.1A CN116523883B (en) | 2023-05-06 | 2023-05-06 | Intelligent latex viscosity control method based on image processing |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116523883A CN116523883A (en) | 2023-08-01 |
| CN116523883B true CN116523883B (en) | 2023-10-13 |
Family
ID=87389938
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310501453.1A Active CN116523883B (en) | 2023-05-06 | 2023-05-06 | Intelligent latex viscosity control method based on image processing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116523883B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106633937A (en) * | 2015-11-02 | 2017-05-10 | 中国石油化工股份有限公司 | Modified asphalt particle having high softening point, preparing method thereof and applications of the particle |
| CN111127418A (en) * | 2019-12-20 | 2020-05-08 | 苏州大学 | Method for detecting curing degree of lacquer coating based on OCT system |
| CN112950558A (en) * | 2021-02-08 | 2021-06-11 | 南宝树脂（佛山）有限公司 | Anti-condensation monitoring method and system for low-temperature non-freezing polyurethane adhesive processing process |
| CN113168891A (en) * | 2018-09-14 | 2021-07-23 | 西北大学 | Data-driven representation and clustering discretization method and system for design optimization and/or performance prediction of material systems and application thereof |
| WO2022224852A1 (en) * | 2021-04-20 | 2022-10-27 | 帝人株式会社 | Method for individual recognition of fiber-reinforced plastic |
| CN115419309A (en) * | 2022-08-31 | 2022-12-02 | 山东百顿减震科技有限公司 | Building vibration isolation method, equipment and medium based on intelligent vibration isolation support |
| CN115508353A (en) * | 2022-09-20 | 2022-12-23 | 东北石油大学 | Quantitative characterization and prediction method for micro-kinetic behavior of liquid drop in crude oil emulsion |
-
2023
- 2023-05-06 CN CN202310501453.1A patent/CN116523883B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106633937A (en) * | 2015-11-02 | 2017-05-10 | 中国石油化工股份有限公司 | Modified asphalt particle having high softening point, preparing method thereof and applications of the particle |
| CN113168891A (en) * | 2018-09-14 | 2021-07-23 | 西北大学 | Data-driven representation and clustering discretization method and system for design optimization and/or performance prediction of material systems and application thereof |
| CN111127418A (en) * | 2019-12-20 | 2020-05-08 | 苏州大学 | Method for detecting curing degree of lacquer coating based on OCT system |
| CN112950558A (en) * | 2021-02-08 | 2021-06-11 | 南宝树脂（佛山）有限公司 | Anti-condensation monitoring method and system for low-temperature non-freezing polyurethane adhesive processing process |
| WO2022224852A1 (en) * | 2021-04-20 | 2022-10-27 | 帝人株式会社 | Method for individual recognition of fiber-reinforced plastic |
| CN115419309A (en) * | 2022-08-31 | 2022-12-02 | 山东百顿减震科技有限公司 | Building vibration isolation method, equipment and medium based on intelligent vibration isolation support |
| CN115508353A (en) * | 2022-09-20 | 2022-12-23 | 东北石油大学 | Quantitative characterization and prediction method for micro-kinetic behavior of liquid drop in crude oil emulsion |
Non-Patent Citations (5)
| Title |
|---|
| Leung B M等.Media additives to promote spheroid circularity and compactness in hanging drop platform.《Biomaterials science》.2015,第3卷(第3期),336-344. * |
| Shang L等.Bioinspired multifunctional spindle‐knotted microfibers from microfluidics.《Small》.2017,第13卷(第4期),1-8. * |
| Zhu J等.All-aqueous printing of viscoelastic droplets in yield-stress fluids.《Acta Biomaterialia》.2022,第165卷60-71. * |
| 任喜伟等.基于连通域标记的原油乳状液粒径检测算法.《计算机应用与软件》.2021,第38卷(第7期),196-201. * |
| 陈海秀等.图像分析在液滴检测技术中的应用.《传感器与微系统》.2016,第35卷(第4期),157-160. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116523883A (en) | 2023-08-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP3971904A1 (en) | Qualitative or quantitative characterization of a coating surface | |
| CN109712162B (en) | Cable character defect detection method and device based on projection histogram difference | |
| CN110148122A (en) | Apple presentation quality stage division based on deep learning | |
| CN110667147B (en) | Character extraction method for film section image | |
| CN106529006A (en) | Depth image-based broiler growth model fitting method and apparatus | |
| CN111368643A (en) | Escherichia coli dynamic growth monitoring method | |
| CN111833313B (en) | Industrial product surface defect detection method and system based on deep active learning | |
| CN116721391B (en) | Method for detecting separation effect of raw oil based on computer vision | |
| CN112949517B (en) | Plant stomata density and opening degree identification method and system based on deep migration learning | |
| CN116523883B (en) | Intelligent latex viscosity control method based on image processing | |
| CN116821630A (en) | Agricultural product maturity prediction method based on fusion data analysis | |
| CN111260658A (en) | Novel depth reinforcement learning algorithm for image segmentation | |
| CN107310133A (en) | BOPP film method for controlling thickness based on image procossing | |
| CN111627059B (en) | Cotton leaf center point positioning method | |
| CN117187943A (en) | Melt detection method and device in crystal pulling process and electronic equipment | |
| CN115641335B (en) | Embryo abnormity multi-cascade intelligent comprehensive analysis system based on time difference incubator | |
| CN106447045A (en) | Evaluation method for ADAS (Advanced Driver Assistant System) based on machine learning | |
| CN114662594A (en) | Target feature recognition analysis system | |
| CN113887395A (en) | Depth separable convolution YOLOv4 model-based filter bag opening position detection method | |
| CN112911164A (en) | Exposure time adjusting method and system for camera for air tightness test | |
| CN112580626A (en) | Method for determining addition amount of covering slag in metallurgical continuous casting process based on artificial intelligence | |
| CN112800929B (en) | Bamboo shoot quantity and high growth rate online monitoring method based on deep learning | |
| CN117282687B (en) | Automatic mark picking and supplementing control system for visual inspection of printed matter | |
| CN117894012B (en) | Machine vision-based mass cake baking stage identification method | |
| CN117784826B (en) | Deviation correction control method for automatic coating machine |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |