CN110223299A - A kind of particle partition method based on deposition process - Google Patents
A kind of particle partition method based on deposition process Download PDFInfo
- Publication number
- CN110223299A CN110223299A CN201910504647.0A CN201910504647A CN110223299A CN 110223299 A CN110223299 A CN 110223299A CN 201910504647 A CN201910504647 A CN 201910504647A CN 110223299 A CN110223299 A CN 110223299A
- Authority
- CN
- China
- Prior art keywords
- chain
- abrasive
- abrasive grain
- image
- deposition
- 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.)
- Granted
Links
- 239000002245 particle Substances 0.000 title claims abstract description 133
- 238000000034 method Methods 0.000 title claims abstract description 46
- 238000005137 deposition process Methods 0.000 title claims abstract description 27
- 238000005192 partition Methods 0.000 title abstract 2
- 239000006061 abrasive grain Substances 0.000 claims abstract description 162
- 230000008021 deposition Effects 0.000 claims abstract description 44
- 230000008859 change Effects 0.000 claims abstract description 19
- 238000000227 grinding Methods 0.000 claims abstract description 10
- 230000011218 segmentation Effects 0.000 claims description 28
- 238000012946 outsourcing Methods 0.000 claims description 15
- 230000008569 process Effects 0.000 claims description 14
- 238000010586 diagram Methods 0.000 claims description 7
- 238000001914 filtration Methods 0.000 claims description 7
- 238000003708 edge detection Methods 0.000 claims description 6
- 230000009466 transformation Effects 0.000 claims description 6
- 238000003709 image segmentation Methods 0.000 claims description 4
- 230000003044 adaptive effect Effects 0.000 claims description 3
- 230000000877 morphologic effect Effects 0.000 claims description 3
- 238000013519 translation Methods 0.000 claims description 3
- 238000009499 grossing Methods 0.000 claims 1
- 230000000007 visual effect Effects 0.000 abstract description 4
- 238000006243 chemical reaction Methods 0.000 abstract description 2
- 238000012544 monitoring process Methods 0.000 abstract description 2
- XEEYBQQBJWHFJM-UHFFFAOYSA-N Iron Chemical compound [Fe] XEEYBQQBJWHFJM-UHFFFAOYSA-N 0.000 abstract 2
- 238000010191 image analysis Methods 0.000 abstract 1
- 229910052742 iron Inorganic materials 0.000 abstract 1
- 239000008188 pellet Substances 0.000 abstract 1
- 238000012545 processing Methods 0.000 description 6
- 238000004458 analytical method Methods 0.000 description 2
- 238000000354 decomposition reaction Methods 0.000 description 2
- 238000002372 labelling Methods 0.000 description 2
- 238000009825 accumulation Methods 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 239000003550 marker Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/155—Segmentation; Edge detection involving morphological operators
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/194—Segmentation; Edge detection involving foreground-background segmentation
-
- 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
- G06V10/267—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 by performing operations on regions, e.g. growing, shrinking or watersheds
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Multimedia (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
The present invention relates to system state machine monitoring technical field, in particular to a kind of particle partition method based on deposition process, comprising: wear particle deposition process video obtained is composed to online visual iron and is decomposed, the abrasive grain chain image at each moment is obtained；Abrasive grain chain in the abrasive grain chain deposited image at each moment of acquisition is split, abrasive grain chain segment is obtained；It is matched using abrasive grain chain segment of the nearest-neighbor method to abrasive grain each moment, obtains the deposition change procedure of abrasive grain chain；Changed according to the length of wear particle deposition chain to remove the abrasive grain segment of redundancy；Opposite grinding pellet chain segment takes each column pixel value to be separately summed, and subtracts each other to the sum of the pixel value of the abrasive grain chain segment of adjacent moment, matches adjacent abrasive grains chain segment same section；Label is generated with range conversion to newly deposited abrasive grain, is split with label watershed, obtains the segmented image of abrasive grain；The present invention can accurately be partitioned into the abrasive grain in wear particle deposition chain, be of great significance to the intelligence and automation of realizing on-line ferrograph image analysis technology.
Description
Technical Field
The invention relates to the technical field of mechanical system state monitoring, in particular to a method for segmenting abrasive particles based on a deposition process.
Background
Due to the adoption of the magnetic field deposition principle, chain and accumulation phenomena are inevitably generated in the ferrograph image of the abrasive particles, the accuracy of subsequent abrasive particle characteristic analysis is greatly influenced, and the over-segmentation or under-segmentation phenomenon can be generated when the chain or accumulated abrasive particles are segmented. At present, many automatic segmentation methods based on mathematical morphology mainly include a segmentation method based on edge detection, a segmentation method based on watershed segmentation, and a segmentation method based on a separation search point algorithm. These algorithms are difficult to apply to abrasive grain segmentation in chains or piles, irregular shapes, rough edges, wide size distributions.
Disclosure of Invention
In order to realize the segmentation of an online visible ferrographic abrasive particle chain, the invention provides an abrasive particle segmentation method based on a deposition process, which comprises the following steps of:
s1, decomposing the abrasive particle deposition process video obtained by the online visible ferrography to obtain an abrasive particle chain image at each moment;
s2, segmenting the abrasive grain chain in the obtained abrasive grain chain deposition image at each moment to obtain an abrasive grain chain image block;
s3, matching the abrasive particle chain diagram blocks at two adjacent moments by using a nearest neighbor method, and finding out the deposition change process of the abrasive particles of the same abrasive particle chain;
s4, removing redundant abrasive grain blocks according to the length change of the abrasive grain deposition chain;
s5, matching the abrasive particle chain diagram blocks at two adjacent moments, and dividing the abrasive particles newly deposited;
and S6, generating a mark by using distance transformation on the newly deposited abrasive particles, and segmenting by using a mark watershed to obtain a segmented image of the abrasive particles.
Further, the process of acquiring the abrasive grain chain image block comprises the following steps: filtering each obtained abrasive particle deposition picture to remove irrelevant noise; and graying, converting into a binary image, detecting the edge of the abrasive particle by using edge detection, finding out an outsourcing rectangle of the outline, extracting an abrasive particle chain area contained in the outsourcing rectangle, and storing coordinates and geometric information of the outsourcing rectangle, wherein the geometric information is the length of the outsourcing rectangle and the height of the outsourcing rectangle.
Further, the matching of the abrasive particle chain image block at each moment of the abrasive particles by using the nearest neighbor method includes searching for an abrasive particle chain with the shortest distance to the kth abrasive particle chain in the abrasive particle chain deposition image at the t +1 th moment, where the distance between the two abrasive particle chains is represented as:
dk′，t+1＝|Xk,t-Xk′,t+1|+|Yk,t-Yk′,t+1|；
wherein d isk′，t+1Representing the distance between the kth abrasive grain chain (k, t) in the abrasive grain chain deposition image at the t moment and the kth abrasive grain chain' in the abrasive grain chain deposition image at the t +1 moment; (X)k,t,Yk,t) The vertex coordinates of the kth abrasive grain chain in the abrasive grain chain deposition image at the tth moment; (Y)k′,t+1,Yk′,t+1) The vertex coordinates of the k' th abrasive grain chain in the abrasive grain chain deposition image at the t +1 th moment are shown.
Further, the removing redundant abrasive grain blocks according to the length change of the abrasive grain deposition chain comprises: taking the size of the newly added abrasive particles as the length change of the abrasive particle deposition chain, and taking the absolute value of w when the length of the same abrasive particle chain is not changed or is changed very little at two adjacent momentst+1-wt|<P; no new abrasive particles are deemed to be deposited into the chain of abrasive particles; wherein, wtThe length of the abrasive grain chain block at time t is shown, and P is an abrasive grain chain change threshold value.
Further, the matching of the abrasive particle chain pattern blocks at two adjacent time instants includes: and (3) translating the abrasive grain chains, and when the pixel difference of the two abrasive grain chains is minimum, matching is successful, so that the pixel difference of the two abrasive grain chains is expressed as:
wherein,representing the x-th in the block of the chain of grinding grains at time ttThe sum of the pixel values of the columns,representing the x-th in the block representing the chain of grinding grains at time tt+1Sum of pixel values of columns, wtIndicates the length of the abrasive grain chain block at time t; s represents the distance of translation of the abrasive grain chain.
Further, the process of performing watershed segmentation comprises:
s61, taking the content of the image of the last frame of the abrasive particle deposition process video obtained by the online visible ferrography as a basic image of the watershed mark;
s62, performing mean filtering on the abrasive grain chain image to be marked to remove irrelevant noise, and performing binarization on the abrasive grain chain image by using an adaptive threshold value Otsu method after graying;
s63, performing morphological opening operation on the binary image to obtain a background area of the abrasive chain image;
and S64, respectively using distance transformation to the areas of the newly added abrasive grains to form watershed mark points, and carrying out watershed segmentation on all the mark points and the basic image to obtain a segmentation image of the abrasive grain chain.
The invention has the following beneficial effects:
1. the automatic segmentation of the abrasive particle chain of the online ferrographic image is realized, and a foundation is provided for the feature extraction of the abrasive particles;
2. the problems of over-segmentation and under-segmentation of the abrasive particle chain are solved, and the segmentation precision is improved;
3. the proposed abrasive particle segmentation based on the deposition process gives full play to the resources and advantages of the online ferrographic image technology.
Drawings
FIG. 1 is a flow chart of a method for image segmentation of a chain of abrasive particles based on a deposition process according to the present invention;
FIG. 2 is an outsourcing rectangle of a chain of abrasive particles according to the invention obtained by edge detection;
FIG. 3 is a schematic diagram of coordinate information and geometric information of one of the blocks of the No. 3 abrasive grain chain in FIG. 2 according to the present invention;
FIG. 4 is a schematic illustration of the deposition process of the 3 rd abrasive grain chain of FIG. 2 in accordance with the present invention;
FIG. 5 is a diagram illustrating the difference between the sums of pixel values of each row of two adjacent frames according to the present invention;
FIG. 6 is a schematic view of a watershed marker of the present invention;
FIG. 7 shows the final result of the mark watershed segmentation performed by the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The invention provides an abrasive particle segmentation method based on a deposition process, which specifically comprises the following steps:
s1, decomposing the abrasive particle deposition process video obtained by the online visible ferrography to obtain an abrasive particle chain image at each moment;
s2, segmenting the abrasive grain chain in the obtained abrasive grain chain deposition image at each moment to obtain an abrasive grain chain image block;
s3, matching the abrasive particle chain diagram blocks of the abrasive particles at each moment by using a recent field method to obtain a deposition change process of the abrasive particle chain;
s4, removing redundant abrasive grain blocks according to the length change of the abrasive grain deposition chain;
s5, adding pixel values of each row of the abrasive grain chain image blocks respectively, subtracting the sum of the pixel values of the abrasive grain chain image blocks at adjacent moments, matching the same parts of the adjacent abrasive grain chain image blocks, wherein the different parts are the newly deposited abrasive grains;
and S6, generating a mark by using distance transformation on the newly deposited abrasive particles, and segmenting by using a mark watershed to obtain a segmented image of the abrasive particles.
As shown in fig. 1, the present invention mainly includes three parts: video processing, image processing and watershed segmentation detailed analysis is performed on the three parts:
(1) video processing
1.1) video decomposition
In this embodiment, the abrasive particle deposition process video decomposition is to decompose abrasive particle deposition information in the time period, which is acquired by an online visual ferrography (OLVF), into a frame-by-frame picture, where different abrasive particle chains in each frame include different abrasive particle information, and by comparing abrasive particle chain pictures of adjacent frames, a position relationship between a newly deposited abrasive particle and an abrasive particle chain and geometric information of the abrasive particle can be obtained; because the deposition speed of the abrasive particles is low, in order to improve the operation rate of the abrasive particle chain image segmentation algorithm, one abrasive particle deposition image can be stored at intervals of 10 frames or 20 frames, and the abrasive particle chain image segmentation algorithm is stored every 20 frames, so that the information of abrasive particle deposition is not lost, and the integrity of the deposition process of the abrasive particle chain is ensured.
Filtering each obtained abrasive particle deposition picture to remove irrelevant noise; performing graying, converting into a binary image, detecting the edge of the abrasive particle by using edge detection, finding out an outsourcing rectangle of the outline, extracting an abrasive particle chain area contained in the outsourcing rectangle, and storing coordinates (X, Y) and geometric information (w, h) of the outsourcing rectangle, wherein X is an abscissa of a two-dimensional coordinate system, and Y is an ordinate of the two-dimensional coordinate system as shown in FIG. 3; the geometric information includes the length of the outer-wrapping rectangle and the height of the outer-wrapping rectangle, wherein fig. 2 is an abrasive grain chain image, one abrasive grain chain image may include a plurality of abrasive grain chain pattern blocks, and the portions of the outer-wrapping rectangles 1-8 in fig. 2 are abrasive grain chain pattern blocks.
As shown in fig. 2, the abrasive grain chains in each frame of picture are respectively subjected to edge detection, and an outline rectangle (a box in the figure) is drawn through edge information, and the top left vertex of the picture is selected in this description. For simplification, the description is given only for the segmentation process of a certain frame abrasive grain chain in the video, and the pictures of the rest frames are similar.
1.2) chain matching of abrasive grains
The abrasive grain chain image comprises a plurality of abrasive grain chain image blocks, the change process of each abrasive grain chain image block needs to be extracted, and how to judge that two abrasive grain chain image blocks in two frames of images are the same abrasive grain chain is the matching process of the abrasive grain chain.
Matching a plurality of abrasive grain pattern blocks at different moments by adopting a nearest neighbor method to obtain a deposition change process of the same abrasive grain chain, wherein the process comprises the following steps of matching the abrasive grain chain pattern blocks at each moment of abrasive grains by utilizing the nearest neighbor method, searching an abrasive grain chain with the shortest distance to a kth abrasive grain chain in an abrasive grain chain deposition image at the t +1 th moment, and assuming that the abrasive grain chain with the shortest distance is the kth' abrasive grain chain in the abrasive grain chain deposition image at the t +1 th moment, expressing the distance between the two abrasive grain chains as follows:
dk′，t+1＝|Xk,t-Xk′,t+1|+|Yk,t-Yk′,t+1|；
wherein d isk′，t+1Representing the distance between the kth abrasive grain chain (k, t) in the abrasive grain chain deposition image at the t moment and the kth abrasive grain chain' in the abrasive grain chain deposition image at the t +1 moment; (X)k,t,Yk,t) The vertex coordinates of the kth abrasive grain chain in the abrasive grain chain deposition image at the tth moment; (Y)k′,t+1,Yk′,t+1) The vertex coordinates of the k' th abrasive grain chain in the abrasive grain chain deposition image at the t +1 th moment are shown.
(2) Image processing
The image processing process is as shown in fig. 4, after the abrasive grain chain of each frame is extracted, graying is carried out, whether new abrasive grains are added or not is judged according to whether the lengths of the abrasive grain chains of two adjacent frames are changed or not through the geometric information (w, h) of the outsourcing rectangle of the abrasive grain chain, and then the deposition change process of a certain abrasive grain chain is obtained. The video identification result is 8 abrasive grain chains in total, each abrasive grain chain is subjected to the redundant operation of abrasive grain chain removal, and for the sake of simplicity, only the deposition process of the 3 rd abrasive grain chain in fig. 2 is described.
2.1) removing redundancies
According to the size change of the same abrasive particle deposition chain, redundant abrasive particle image blocks are removed, the abrasive particle chain is wrapped by rectangular geometric information (w, h), the length change of the abrasive particle chain is the size of the newly added abrasive particles, and when the length w of the same abrasive particle chain of two adjacent frames is not changed or is changed slightly, the new abrasive particles are not deposited on the abrasive particle chain, namely: | wt+1-wt|<P；
Wherein, wtThe length of the abrasive grain chain block at time t is shown, and P is an abrasive grain chain change threshold value.
When the above formula is satisfied, w is removedt+1The abrasive particle chain image blocks only keep the abrasive particle chains with new abrasive particle deposition, and the operation can improve the running speed of the algorithm.
2.2) graying
And filtering the extracted mean value of the matched abrasive particle chain image block with redundancy removed to remove noise points, and then carrying out graying to keep the integrity of the pixel values of the abrasive particle chain image block. The range of the number of pixels (x, y) of the grayed grit chain picture is [0,255 ].
2.3) adding new abrasive grains
The abrasive particles are precipitated as shown in fig. 4, and new abrasive particles are deposited in two adjacent abrasive particle chains in three cases: newly adding abrasive particles from the front end of the abrasive particle chain; newly adding abrasive particles from the rear end of the abrasive particle chain; the abrasive grains are added from the front end and the rear end of the abrasive grain chain at the same time, for example, in fig. 4, the second abrasive grain chain from top to bottom is added with the abrasive grains from the front end and the rear end of the abrasive grain chain on the basis of the first abrasive grain chain, the third abrasive grain chain is added with the abrasive grains from the rear end of the second abrasive grain chain, and the fifth abrasive grain chain is added with the abrasive grains from the front end of the fourth abrasive grain chain.
Making the same abrasive grain chain in two adjacent frames of pictures as the difference value of the sum of the pixel values of each row, matching the same parts of the two abrasive grain chains to find out the newly added abrasive grain region, wherein fig. 5 is an example of newly adding abrasive grains from the front end of the abrasive grain chain, and the length of the abrasive grain chain at the t-th moment is wtThe middle aligned part being the same part, i.e. wtTwo ends are newly deposited abrasive grains, and the length of an abrasive grain chain at the t +1 th moment is wt+1The length of the abrasive particle newly deposited at the front end is marked as S in the figure, and the length of the abrasive particle chain at the t-th moment is wtThe minimum value of the difference between the pixels of the two abrasive grain chains can be obtained after the distance S is translated.
Therefore, the deposition position and the size of the newly added abrasive grains need to be obtained, the same abrasive grain chain area is obtained by matching the previous frame of abrasive grain chain with the next frame of abrasive grain chain, and the remaining abrasive grain chain area is the newly added abrasive grains.
The gray-scaled pixel (x, y) of the shot chain picture has a value range of [0,255%]To, forProjecting each row in the horizontal direction, namely adding the sum of pixel values of pixel points of each row, wherein the length and width of a previous frame abrasive grain chain image block is (w)t,ht)，Is a point (x)t,yt) The value of (a) of (b),denotes the x thtThe sum of the pixel values of the columns, expressed as:
the geometric characteristic of the abrasive grain chain image block in the next frame is (w)t+1,ht+1)，Is a point (x)t+1,yt+1)The value of (a) of (b),denotes the x tht+1The sum of the pixel values of the columns, expressed as:
the abrasive grain chains are translated one by one, the difference between pixel values of the two abrasive grain chains is calculated, when the difference between the pixel values is minimum, the two abrasive grain chains are judged to be the same part of the two abrasive grain chains, and the difference between the image blocks of the two abrasive grain chains is represented as:
wherein,representing the x-th in the block of the chain of grinding grains at time ttThe sum of the pixel values of the columns,representing the x-th in the block representing the chain of grinding grains at time tt+1Sum of pixel values of columns, wtIndicates the length of the abrasive grain chain block at time t; s represents the distance of the abrasive grain chain translation, and when the difference between the pixel values of the two abrasive grain chains is the minimum, S at this time is the length of the abrasive grain deposited at the front end of the abrasive grain chain.
2.4) performing distance conversion
Distance transformation is a common technique in watershed algorithms and is not described here.
(3) Watershed segmentation
3.1) complete abrasive grain chain
Because the image of the last frame of the abrasive particle deposition process video obtained by the online visual ferrography includes all the abrasive particle information, the content of the image of the last frame of the abrasive particle deposition process video obtained by the online visual ferrography is used as a base image of the watershed mark, namely an image of a complete abrasive particle chain.
3.2) watershed markers
Before segmentation, a foreground object and a background object in an image need to be labeled and distinguished, namely watershed labeling, so as to ensure a watershed segmentation method, wherein the labeling process comprises the steps of carrying out primary processing on an image of a grinding particle chain, carrying out mean filtering on the image of the grinding particle chain, removing irrelevant noise, and carrying out binarization on the image of the grinding particle chain by using an adaptive threshold Otsu method (OTSU) after graying; and performing morphological opening operation on the binary image to obtain most of foreground areas. Using the distance transform, watershed mark points are formed, and the obtained watershed mark is as shown in fig. 6.
3.3) segmentation results
Performing watershed segmentation on all the mark points and the basic image to obtain a segmentation image of the abrasive chain; this example results in a final segmented image as shown in fig. 7, and the profile of each abrasive grain on the abrasive grain chain can be obtained by the present invention.
Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.
Claims (7)
1. An abrasive particle chain image segmentation method based on a deposition process is characterized by comprising the following steps:
s1, decomposing the abrasive particle deposition process video obtained by the online visible ferrography to obtain an abrasive particle chain image at each moment;
s2, segmenting the abrasive grain chain in the obtained abrasive grain chain deposition image at each moment to obtain an abrasive grain chain image block;
s3, matching the abrasive particle chain diagram blocks at two adjacent moments by using a nearest neighbor method, and finding out the deposition change process of the abrasive particles of the same abrasive particle chain;
s4, removing redundant abrasive particle chain blocks according to the length change of the abrasive particle deposition chain;
s5, matching the abrasive particle chain diagram blocks at two adjacent moments, and dividing the abrasive particles newly deposited;
and S6, generating a mark by applying distance transformation to the newly deposited abrasive particles, and segmenting by applying a watershed algorithm according to the mark to obtain a segmented image of the abrasive particles.
2. The method for segmenting the abrasive particles based on the deposition process, as claimed in claim 1, wherein the process of obtaining the abrasive particle chain image block comprises: filtering each obtained abrasive particle deposition picture to remove irrelevant noise; and graying, converting into a binary image, detecting the edge of the abrasive particle by using edge detection, finding out an outsourcing rectangle of the outline, extracting an abrasive particle chain area contained in the outsourcing rectangle, and storing coordinates and geometric information of the outsourcing rectangle, wherein the geometric information is the length of the outsourcing rectangle and the height of the outsourcing rectangle.
3. The method for segmenting abrasive particles based on the deposition process according to claim 1, wherein the matching of the abrasive particle chain image block at each moment of the abrasive particles by using the nearest neighbor method comprises searching for one abrasive particle chain with the shortest distance to the kth abrasive particle chain in the abrasive particle chain deposition image at the t +1 th moment, and the distance between the two abrasive particle chains is represented as:
dk′，t+1＝|Xk,t-Xk′,t+1|+|Yk,t-Yk′,t+1|；
wherein d isk′，t+1Representing the distance between the kth abrasive grain chain (k, t) in the abrasive grain chain deposition image at the t moment and the kth abrasive grain chain' in the abrasive grain chain deposition image at the t +1 moment; (X)k,t,Yk,t) The vertex coordinates of the kth abrasive grain chain in the abrasive grain chain deposition image at the tth moment; (Y)k′,t+1,Yk′,t+1) For the first in the deposited image of abrasive grain chain at time t +1Vertex coordinates of k' abrasive grain chains.
4. The method for segmenting abrasive particles based on the deposition process according to claim 1, wherein the removing redundant abrasive particle blocks according to the length change of the abrasive particle deposition chain comprises the following steps: taking the size of the newly added abrasive particles as the length change of the abrasive particle deposition chain, and taking the absolute value of w when the length of the same abrasive particle chain is not changed or is changed very little at two adjacent momentst+1-wt|<P; no new abrasive particles are deemed to be deposited into the chain of abrasive particles; wherein, wtThe length of the abrasive grain chain block at time t is shown, and P is an abrasive grain chain change threshold value.
5. The method for segmenting abrasive particles based on the deposition process according to claim 1, wherein the matching of the abrasive particle chain pattern blocks at two adjacent time instants comprises: and (3) translating the abrasive grain chains, and when the pixel difference of the two abrasive grain chains is minimum, matching is successful, so that the pixel difference of the two abrasive grain chains is expressed as:
wherein,representing the x-th in the block of the chain of grinding grains at time ttThe sum of the pixel values of the columns,representing the x-th in the block representing the chain of grinding grains at time tt+1Sum of pixel values of columns, wtIndicates the length of the abrasive grain chain block at time t; s represents the distance of translation of the abrasive grain chain.
6. The method of claim 5, wherein the abrasive grain segmentation based on a deposition process comprises,xth of abrasive grain chain block in t timetSum of pixel values of columnsExpressed as:
wherein h istRepresenting the height of the abrasive chain pattern block at time t;representing a pixel (x)t,yt) The gray value of (a).
7. The method for segmenting abrasive particles based on the deposition process according to claim 1, wherein the step S6 includes:
s61, taking the content of the image of the last frame of the abrasive particle deposition process video obtained by the online visible ferrography as a basic image of the watershed mark;
s62, performing mean filtering, smoothing color details and graying on the abrasive grain chain image to be marked, and performing binarization on all the abrasive grain chain images subjected to graying by using an adaptive threshold Otsu method;
s63, performing morphological opening operation on the binary image to obtain a foreground region of the abrasive chain image;
and S64, respectively using distance transformation to the areas of the newly added abrasive grains to form watershed mark points, and carrying out watershed segmentation on all the mark points and the basic image to obtain a segmentation image of the abrasive grain chain.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910504647.0A CN110223299B (en) | 2019-06-12 | 2019-06-12 | Abrasive particle segmentation method based on deposition process |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910504647.0A CN110223299B (en) | 2019-06-12 | 2019-06-12 | Abrasive particle segmentation method based on deposition process |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110223299A true CN110223299A (en) | 2019-09-10 |
| CN110223299B CN110223299B (en) | 2021-06-18 |
Family
ID=67816567
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910504647.0A Active CN110223299B (en) | 2019-06-12 | 2019-06-12 | Abrasive particle segmentation method based on deposition process |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110223299B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115965624A (en) * | 2023-03-16 | 2023-04-14 | 山东宇驰新材料科技有限公司 | Detection method for anti-wear hydraulic oil pollution particles |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2008099349A1 (en) * | 2007-02-13 | 2008-08-21 | Element Six Ltd | Abrasive segments |
| CN103886579A (en) * | 2013-12-11 | 2014-06-25 | 西安交通大学 | Abrasive particle chain self-adaptive segmentation method orienting online ferrographic image automatic identification |
| CN107703142A (en) * | 2016-08-09 | 2018-02-16 | 象限空间（天津）科技有限公司 | A kind of iron spectral slice image analysis method and system |
| CN109584253A (en) * | 2018-12-20 | 2019-04-05 | 重庆切克威科技有限公司 | Oil liquid abrasive grain image partition method |
-
2019
- 2019-06-12 CN CN201910504647.0A patent/CN110223299B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2008099349A1 (en) * | 2007-02-13 | 2008-08-21 | Element Six Ltd | Abrasive segments |
| CN103886579A (en) * | 2013-12-11 | 2014-06-25 | 西安交通大学 | Abrasive particle chain self-adaptive segmentation method orienting online ferrographic image automatic identification |
| CN107703142A (en) * | 2016-08-09 | 2018-02-16 | 象限空间（天津）科技有限公司 | A kind of iron spectral slice image analysis method and system |
| CN109584253A (en) * | 2018-12-20 | 2019-04-05 | 重庆切克威科技有限公司 | Oil liquid abrasive grain image partition method |
Non-Patent Citations (2)
| Title |
|---|
| SONG FENG ETAL.: "A Wear Debris Segmentation Method for Direct Reflection Online Visual Ferrography", 《SENSORS》 * |
| 姚盼盼: "基于改进灰色聚类的铁谱图像分割", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115965624A (en) * | 2023-03-16 | 2023-04-14 | 山东宇驰新材料科技有限公司 | Detection method for anti-wear hydraulic oil pollution particles |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110223299B (en) | 2021-06-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111814711B (en) | Image feature quick matching method and system applied to mine machine vision | |
| CN107862667B (en) | Urban shadow detection and removal method based on high-resolution remote sensing image | |
| CN106934795B (en) | A kind of automatic testing method and prediction technique of glue into concrete beam cracks | |
| CN107045634B (en) | Text positioning method based on maximum stable extremum region and stroke width | |
| CN108846844B (en) | Sea surface target detection method based on sea antenna | |
| CN105160355B (en) | A kind of method for detecting change of remote sensing image based on region correlation and vision word | |
| CN106709524B (en) | Element symbol detection and identification method in electrical engineering drawing | |
| CN109523479A (en) | A kind of bridge pier surface gaps visible detection method | |
| CN102842037A (en) | Method for removing vehicle shadow based on multi-feature fusion | |
| CN106127735B (en) | A kind of facilities vegetable edge clear class blade face scab dividing method and device | |
| CN112508913B (en) | Cable section edge detection method based on image detection | |
| CN111489337A (en) | Method and system for removing false defects through automatic optical detection | |
| CN109558877B (en) | KCF-based offshore target tracking algorithm | |
| CN110223299B (en) | Abrasive particle segmentation method based on deposition process | |
| CN113610052A (en) | Tunnel water leakage automatic identification method based on deep learning | |
| CN113538500A (en) | Image segmentation method and device, electronic equipment and storage medium | |
| CN117876984A (en) | Road sign line detection and identification method based on MLS point cloud | |
| CN117765287A (en) | Image target extraction method combining LWR and density clustering | |
| CN109145875B (en) | Method and device for removing black frame glasses in face image | |
| CN107704864A (en) | Well-marked target detection method based on image object Semantic detection | |
| Tehsin et al. | Survey of region-based text extraction techniques for efficient indexing of image/video retrieval | |
| Zhu | Moving Objects Detection and Segmentation Based on Background Subtraction and Image Over-Segmentation. | |
| CN116416200B (en) | Forest fire smoke detection method based on region matching | |
| CN113379765B (en) | Road surface repair area extraction and repair type judgment method | |
| Kumari et al. | A Review on Comparative Study of Different Edge Detection Techniques |
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 |