CN115115621A - Lubricating oil pollution degree detection method based on image processing - Google Patents
Lubricating oil pollution degree detection method based on image processing Download PDFInfo
- Publication number
- CN115115621A CN115115621A CN202211015805.4A CN202211015805A CN115115621A CN 115115621 A CN115115621 A CN 115115621A CN 202211015805 A CN202211015805 A CN 202211015805A CN 115115621 A CN115115621 A CN 115115621A
- Authority
- CN
- China
- Prior art keywords
- clustering
- value
- cluster
- moment
- oil
- 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
Images
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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/762—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using clustering, e.g. of similar faces in social networks
- G06V10/763—Non-hierarchical techniques, e.g. based on statistics of modelling distributions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- 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
- G06T2207/10004—Still image; Photographic image
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20024—Filtering details
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
Abstract
The invention relates to the technical field of image data processing, in particular to a lubricating oil pollution degree detection method based on image processing, which comprises the following steps: acquiring an oil stain area gray image at each moment in a preset time period, and performing image processing on the oil stain area gray image at each moment to obtain an oil stain clustering gray image corresponding to each preset clustering K value at each moment; determining an oil spot clustering gray level image corresponding to the optimal preset clustering K value at each moment, and further determining the oil spot boundary mixing fuzzy degree corresponding to the optimal preset clustering K value, so as to determine an optimal degree index and obtain the average radius and the average gray level value of a deposition ring in the optimal oil spot clustering gray level image; the pollution degree evaluation value of the lubricating oil to be detected is determined, the lubricating oil to be detected is analyzed based on the pollution degree evaluation value, the problem that the detection accuracy of the pollution degree of the lubricating oil is low is solved, and the accuracy of detecting the pollution degree of the lubricating oil is improved.
Description
Technical Field
The invention relates to the technical field of image data processing, in particular to a lubricating oil pollution degree detection method based on image processing.
Background
In the operation process of mechanical equipment, it is very important to detect the pollution degree of various lubricating oils such as diesel oil, engine oil and the like in real time, the lubricating oil pollution means that the lubricating oil contains wear particles and oil sludge generated by deep oiling, the lubricating oil pollution is one of important reasons of mechanical equipment failure, and the lubricating oil with serious pollution can accelerate the wear of the mechanical equipment, reduce the mechanical life and even cause major accidents.
The pollution degree of the lubricating oil is generally judged by manually observing a spot spectrum of the lubricating oil at present, the standard of the pollution degree detection is established according to manual experience, a large amount of human resources can be wasted by the detection method, the detection process is easily influenced by various external factors, the detection result has certain subjectivity, and the detection accuracy of the pollution degree of the lubricating oil by the detection method is poor. In order to overcome some defects existing in the process of artificially detecting the pollution degree of the lubricating oil, along with the development of an image data processing technology, the prior method provides a method and a device for detecting the pollution degree of the lubricating oil with the publication number of CN 101196510B. The method comprises the steps of carrying out image data analysis processing on image characteristics of a lubricating oil spot map at any detection moment to obtain three measurement parameters capable of reflecting the pollution degree of lubricating oil, and judging the pollution degree of the lubricating oil according to the sizes of the three measurement parameters. For lubricating oil spot maps at different detection moments, the detection moment is too early, the areas of three rings are not formed due to incomplete diffusion and dryness of lubricating oil spots, the detection moment is too late, the edges of the rings are blurred due to interpenetration and diffusion of the three rings of the lubricating oil spots, so that the image characteristics of the lubricating oil spot maps at different detection moments are greatly different, three measurement parameters of the degree of pollution of the lubricating oil obtained from the image characteristics are greatly different, and finally the problems of low detection precision and low accuracy of the degree of pollution of the lubricating oil are caused.
Disclosure of Invention
In order to solve the technical problem of low detection accuracy of the existing lubricating oil pollution degree, the invention aims to provide a lubricating oil pollution degree detection method based on image processing.
The invention provides a method for detecting the pollution degree of lubricating oil based on image processing, which comprises the following steps:
acquiring a lubricating oil spot map image of lubricating oil to be detected at each moment in a preset time period, and performing image preprocessing operation on the lubricating oil spot map image to obtain an oil spot gray image at each moment so as to obtain an oil spot area gray image at each moment;
clustering the oil spot region gray level images at each moment according to the first preset cluster K value, the second preset cluster K value, the third preset cluster K value and the oil spot region gray level images at each moment to obtain oil spot cluster gray level images corresponding to the preset cluster K values at each moment;
determining a clustering effect evaluation value of the oil spot clustering gray level image corresponding to each preset clustering K value at each moment according to the number of pixel points in each clustering cluster in the oil spot clustering gray level image corresponding to each preset clustering K value at each moment, the gray value and the position of each pixel point, the gray value and the position of a clustering center, the position of an image center and the number of clustering clusters, and further determining the oil spot clustering gray level image corresponding to the optimal preset clustering K value at each moment;
determining the oil spot boundary mixing fuzzy degree of the oil spot clustering gray level image corresponding to the optimal preset clustering K value at each moment according to the gradient value and the gradient direction of each pixel point in the oil spot clustering gray level image corresponding to the optimal preset clustering K value at each moment;
determining an optimal degree index corresponding to the oil spot clustering gray level image at each moment according to the oil spot boundary mixing fuzzy degree of the oil spot clustering gray level image corresponding to the time sequence number corresponding to each moment and the optimal preset clustering K value at each moment in a preset time period, and further determining the optimal oil spot clustering gray level image of the lubricating oil to be detected;
the average radius and the average gray value of the deposit ring in the optimal oil spot cluster gray image of the lubricating oil to be detected are obtained, the pollution degree evaluation value of the lubricating oil to be detected is determined according to the average radius and the average gray value of the deposit ring in the optimal oil spot cluster gray image of the lubricating oil to be detected, and then the pollution degree of the lubricating oil to be detected is determined.
Further, the step of determining the clustering effect evaluation value of the oil spot clustering gray level image corresponding to each preset clustering K value at each moment comprises the following steps:
determining the distance from each pixel point of each cluster to the image center and the distance from the cluster center to the image center according to the positions of each pixel point, the cluster center and the image center in each cluster in the oil spot clustering gray level image corresponding to each preset clustering K value at each moment;
determining a clustering distance corresponding to each pixel point in each clustering cluster in the oil spot clustering gray level image corresponding to each preset clustering K value at each moment according to the gray values of each pixel point and the clustering center in each clustering cluster in the oil spot clustering gray level image corresponding to each preset clustering K value at each moment, the distance from each pixel point to the image center and the distance from the clustering center to the image center;
determining the cluster clustering evaluation value of the oil spot clustering gray level image corresponding to each preset clustering K value at each moment according to the number of clustering clusters in the oil spot clustering gray level image corresponding to each preset clustering K value at each moment, the number of pixel points in each clustering cluster and the clustering distance corresponding to each pixel point in each clustering cluster;
calculating the distance between every two adjacent clustering centers in the oil spot clustering gray level image corresponding to every preset clustering K value at every moment, and determining the inter-class clustering evaluation value of the oil spot clustering gray level image corresponding to every preset clustering K value at every moment according to the number of the clustering centers in the oil spot clustering gray level image corresponding to every preset clustering K value at every moment and the distance between every two adjacent clustering centers;
acquiring the area of each cluster in the oil stain clustering gray level image corresponding to each preset clustering K value at each moment, the maximum inscribed circle area and the minimum circumscribed circle area corresponding to the fitting outline of each cluster, wherein the area, the maximum inscribed circle area and the minimum circumscribed circle area of each cluster are the number of pixel points in the region, and determining the circular ring circularity of the oil stain clustering gray level image corresponding to each preset clustering K value at each moment according to the number of clusters in the oil stain clustering gray level image corresponding to each preset clustering K value at each moment, the maximum inscribed circle area and the minimum circumscribed circle area corresponding to the fitting outline of each cluster and the area of each cluster;
and determining the clustering effect evaluation value of the oil spot clustering gray level image corresponding to each preset clustering K value at each moment according to the cluster intra-cluster evaluation value, the cluster inter-cluster evaluation value and the annular circularity of the oil spot clustering gray level image corresponding to each preset clustering K value at each moment.
Further, the calculation formula for determining the annular circularity of the oil spot clustering gray level image corresponding to each preset clustering K value at each moment is as follows:
wherein the content of the first and second substances,Qthe circularity of the ring of the oil spot clustering gray level image corresponding to each preset clustering K value at each moment,Kthe number of clustering clusters in the oil spot clustering gray level image corresponding to each preset clustering K value at each moment,clustering the oil spots in the gray level image corresponding to each preset cluster K value at each momentiThe area of each of the clusters to be clustered,clustering the oil spots in the gray level image corresponding to each preset cluster K value at each momentiThe minimum circumscribed circle area corresponding to the fitted contour of each cluster,clustering the oil spots in the gray level image corresponding to each preset cluster K value at each momenti-the maximum inscribed circle area for the fitted contours of 1 cluster.
Further, the step of determining the oil spot clustering gray level image corresponding to the optimal preset clustering K value at each moment comprises:
and if the clustering effect evaluation value of the oil spot clustering gray level image corresponding to a certain preset clustering K value at any moment is greater than a preset clustering effect evaluation threshold value, and the clustering effect evaluation value of the oil spot clustering gray level image corresponding to the preset clustering K value is the maximum value, judging that the oil spot clustering gray level image corresponding to the preset clustering K value at the moment is the oil spot clustering gray level image corresponding to the optimal preset clustering K value.
Further, the step of determining the oil spot boundary mixing fuzzy degree of the oil spot clustering gray level image corresponding to the optimal preset clustering K value at each moment comprises the following steps:
when the optimal preset clustering K value at the nth moment is the first preset clustering K value, obtaining the oil spot boundary mixing fuzzy degree of the oil spot clustering gray level image corresponding to the first preset clustering K value at the nth moment;
when it comes toWhen the best preset clustering K value at each moment is the second preset clustering K value, the first preset clustering K value is adjustedCarrying out edge detection on the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment to obtain the first preset clustering K valueOil spot edge lines in the oil spot cluster gray level images corresponding to the second preset cluster K value at each moment;
according to the firstDetermining the gray value of each edge pixel point of the oil stain edge line in the oil stain cluster gray image corresponding to the second preset cluster K value at each moment and the number of the edge pixel pointsA first oil spot boundary mixing fuzzy factor of the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment;
get the firstGradient directions of all edge pixel points and eight neighborhood pixel points of an oil spot edge line in the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment are determined according to the gradient directions of the eight neighborhood pixel pointsDetermining the number of edge pixel points of an oil spot edge line in the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment, the gradient direction of each edge pixel point and the gradient direction of eight neighborhood pixel points of each edge pixel point, and determining the first preset clustering K valueA second oil spot boundary mixing fuzzy factor of the oil spot clustering gray level image corresponding to a second preset clustering K value at each moment;
according to the firstDetermining a first oil spot boundary mixed fuzzy factor and a second oil spot boundary mixed fuzzy factor of the oil spot clustering gray level image corresponding to a second preset clustering K value at each momentAnd mixing fuzzy degrees of oil spot boundaries of the oil spot clustering gray level images corresponding to the second preset clustering K value at each moment.
Further, the step of determining the oil spot boundary mixing fuzzy degree of the oil spot clustering gray level image corresponding to the optimal preset clustering K value at each moment further comprises:
when it comes toWhen the best preset cluster K value at each moment is the third preset cluster K value, the first preset cluster K value is selected as the second preset cluster K valueSecond preset cluster K value of each momentCarrying out edge detection on the corresponding oil spot clustering gray level image to obtain the first stepClustering the edge lines of the diffusion ring in the gray level image according to the oil spots corresponding to the third preset clustering K value at each moment, and further obtaining gradient values of all edge pixel points and eight neighborhood pixel points of the edge lines of the diffusion ring;
according to the firstGradient values of all edge pixel points of a diffusion ring edge line in the oil spot clustering gray level image corresponding to a third preset clustering K value at each moment and gradient values of eight neighborhood pixel points of the gradient pixel points of the edge line, entropy values and variances of gradient co-occurrence matrixes corresponding to all edge pixel points of the diffusion ring edge line are determined, and then a first preset clustering K value is determinedDiffusion ring boundary mixed fuzzy factors corresponding to the oil spot clustering gray level images corresponding to the third preset clustering K value at each moment;
according to the firstDetermining gradient values of all edge pixel points of diffusion ring edge lines in the oil spot clustering gray level image corresponding to the third preset clustering K value at each moment and gradient values of eight neighborhood pixel points of all edge pixel points in the oil spot clustering gray level image at each momentThe number of target neighborhood pixels corresponding to each edge pixel point of the diffusion ring edge line in the oil spot clustering gray level image corresponding to the third preset clustering K value at each moment;
according to the firstThe number of target neighborhood pixel points corresponding to each edge pixel point of the diffusion ring edge line in the oil spot clustering gray level image corresponding to the third preset clustering K value at each moment,Determining the number of edge pixel points of the edge line of the diffusion ring and the diffusion ring boundary mixed fuzzy factorAnd mixing fuzzy degrees of oil spot boundaries of the oil spot clustering gray level images corresponding to the third preset clustering K value at each moment.
Further, it is determinedThe calculation formula of the second oil spot boundary mixed fuzzy factor of the oil spot cluster gray level image corresponding to the second preset cluster K value at each moment is as follows:
wherein the content of the first and second substances,Bis as followsA second oil spot boundary mixing fuzzy factor of the oil spot clustering gray level image corresponding to a second preset clustering K value at each moment,uis as followsThe number of edge pixel points of the oil spot edge line in the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment,is as followsThe first of the oil stain edge lines in the oil stain clustering gray level image corresponding to the second preset clustering K value of each momentiThe gradient direction of the pixel points at each edge,is as followsThe first of the oil stain edge lines in the oil stain clustering gray level image corresponding to the second preset clustering K value of each momentiThe first of each edge pixelvGradient direction of each neighborhood pixel.
Further, it is determinedThe calculation formula of the diffusion ring boundary mixed fuzzy factor corresponding to the oil spot clustering gray level image corresponding to the third preset clustering K value at each moment is as follows:
wherein the content of the first and second substances,Cis as followsDiffusion ring boundary mixed fuzzy factors corresponding to the oil spot clustering gray level image corresponding to the third preset clustering K value at each moment,mis as followsThe number of edge pixel points of the edge line of the diffusion ring in the oil spot clustering gray level image corresponding to the third preset clustering K value at each moment,is as followsThe third preset cluster K value at each moment corresponds to the first diffusion ring edge line in the oil spot cluster gray level imageiEntropy values of gradient co-occurrence matrixes corresponding to the edge pixel points,is as followsOil spot aggregation corresponding to third preset clustering K value of each momentDiffusion ring edge line in gray-like imageiThe variance of the gradient co-occurrence matrix corresponding to each edge pixel point,based on a natural constant eIs used as the exponential function of (1).
Further, a calculation formula for determining the preference degree index corresponding to the oil spot cluster gray level image at each moment is as follows:
wherein the content of the first and second substances,is composed oftThe corresponding preference degree index of the oil spot cluster gray level image at the moment,is composed oftThe time sequence number corresponding to the time is,is composed oftThe oil spot boundary mixing fuzzy degree corresponding to the oil spot clustering gray level image at the moment,based on a natural constant eIs used as the exponential function of (1).
Further, the step of determining the contamination level of the lubricating oil to be detected comprises:
and if the evaluation value of the pollution degree of the lubricating oil to be detected is larger than the pollution degree threshold value, judging that the pollution degree of the lubricating oil to be detected is serious, otherwise, judging that the pollution degree of the lubricating oil to be detected is not serious.
The invention has the following beneficial effects:
the invention provides a method for detecting the pollution degree of lubricating oil based on image processing, which comprises the steps of acquiring a lubricating oil spot map image of the lubricating oil to be detected at each moment in a preset time period, carrying out image preprocessing operation on the lubricating oil spot map image in order to facilitate image data analysis on the lubricating oil spot map image of the lubricating oil to be detected, thereby obtaining an oil spot area gray level image at each moment, and taking the oil spot area gray level image at each moment as a reference image, thereby being beneficial to improving the reference value of the determined oil spot clustering gray level image corresponding to the optimal preset clustering K value at each moment; in order to determine the oil stain cluster gray level image corresponding to the optimal preset cluster K value at each moment and also to enable each moment to have the oil stain cluster gray level image with the best cluster effect, the oil stain cluster gray level image at each moment is clustered according to the first preset cluster K value, the second preset cluster K value, the third preset cluster K value and the oil stain area gray level image at each moment to obtain the oil stain cluster gray level image corresponding to each preset cluster K value at each moment, then the cluster effect evaluation value of the oil stain cluster gray level image corresponding to each preset cluster K value at each moment is determined according to the image characteristic information of the oil stain cluster gray level image corresponding to each preset cluster K value at each moment, the oil stain cluster gray level image corresponding to the optimal preset cluster K value at each moment is further determined, and compared with the existing lubricating oil stain image at any detection moment, the lubricating oil stain image at one detection moment is selected for lubricating oil contamination degree detection, the referential performance of the lubricating oil pollution detection is stronger by the oil spot clustering gray level image with the best clustering effect corresponding to each moment based on the clustering effect evaluation value; in order to select an optimal oil stain clustering gray image in a preset time period, determining the oil stain boundary mixing fuzzy degree of the oil stain clustering gray image corresponding to the optimal preset clustering K value at each moment according to the gradient value and the gradient direction of each pixel point in the oil stain clustering gray image corresponding to the optimal preset clustering K value at each moment by using an image data processing technology, and further determining the optimization degree index corresponding to the oil stain clustering gray image at each moment, wherein the oil stain boundary mixing fuzzy degree and the optimization degree index are in a negative correlation relationship, the greater the oil stain boundary mixing fuzzy degree, the smaller the optimization degree index is, and based on the optimization degree index, the more accurate, more real and better clustering effect optimal oil stain clustering gray image can be obtained; according to the average radius and the average gray value of the deposition ring in the optimal oil spot cluster gray level image of the lubricating oil to be detected, the pollution degree evaluation value of the lubricating oil to be detected is determined, the pollution degree of the lubricating oil to be detected is further determined, the pollution degree of the lubricating oil is analyzed by utilizing image data such as image characteristic information of the optimal oil spot cluster gray level image, and the accuracy of the detection of the pollution degree of the lubricating oil can be effectively improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions and advantages of the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
FIG. 1 is a flow chart of a method for detecting contamination level of lubricating oil based on image processing according to the present invention;
fig. 2 is a schematic view of a lubricating oil spot map image in an embodiment of the present invention, in which 201 is a deposition ring, 202 is a diffusion ring, and 203 is an oil ring.
Detailed Description
To further explain the technical means and effects of the present invention adopted to achieve the predetermined objects, the following detailed description of the embodiments, structures, features and effects of the technical solutions according to the present invention will be given with reference to the accompanying drawings and preferred embodiments. In the following description, different references to "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The embodiment provides a method for detecting the pollution degree of lubricating oil based on image processing, as shown in fig. 1, the method includes the following steps:
(1) the method comprises the steps of obtaining a lubricating oil spot map image of lubricating oil to be detected at each moment in a preset time period, carrying out image preprocessing operation on the lubricating oil spot map image to obtain an oil spot gray image at each moment, and further obtaining an oil spot area gray image at each moment.
(1-1) acquiring a lubricating oil spot map image of the lubricating oil to be detected at each moment in a preset time period, and performing image preprocessing operation on the lubricating oil spot map image to obtain an oil spot gray image at each moment.
In this embodiment, the lubricating oil spot spectrum image is used for detecting the contamination degree of the lubricating oil on site, and the step of obtaining the lubricating oil spot spectrum image may be: the method comprises the steps of firstly obtaining a sample liquid of lubricating oil to be detected, dripping the sample liquid of the lubricating oil to be detected on filter paper, forming lubricating oil spots on the filter paper, wherein the lubricating oil spots can also be called as oil spots, then shooting the filter paper with the lubricating oil spots at each moment in a preset time period by using an industrial camera, and obtaining an image shot by the industrial camera as a lubricating oil spot spectrum image. Generally, after oil spots are manufactured, the oil spots are naturally dried for 24 hours, and more real lubricating oil spot map images exist within 1-3 days after 24 hours, so that the preset time period of the embodiment is 72 hours, each time is every hour, an industrial camera can shoot one lubricating oil spot map image within 72 hours after 24 hours of natural drying of the oil spots, and obtaining the lubricating oil spot map image of the lubricating oil to be detected at each time is a precondition for detecting the pollution degree of the lubricating oil.
The schematic diagram of the lubricating oil spot map image is shown in fig. 2, and according to the national standard GB8030-87, the lubricating oil spot map image generally has three circular rings or two circular rings. In fig. 2, the three rings are: the deposition ring 201, in the center of the spot, is light gray and is a large particle insoluble precipitate zone. When the lubricating oil is near to discard, the detergent dispersant disappears, the diameter of the deposit ring is small, and the color is black. The diffuser ring 202, which is a ring with fine carbon particles diffusing out with the detergent dispersant, is light gray to gray outside the deposition ring. The wider the width of the diffuser ring, the better the detergent dispersing ability of the lubricating oil, and the narrower or vanished the width, the poor detergent dispersing ability or the exhausted detergent dispersant. The oil ring 203 is arranged on the outer ring of the diffusion ring and is in a light yellow to brownish red oil immersion area, the oxidation degree of oil is shown, the oil ring of new oil is transparent, and the color of the oil ring of waste oil is changed from yellow to brownish red.
In order to reduce the influence of noise caused by other external factors on the image data, the image preprocessing operation is performed on the lubricating oil speckle pattern image in the embodiment, and the image preprocessing operation may include the following steps: the method comprises the steps of utilizing a weighted average algorithm to conduct graying processing on a lubricating oil spot map image, conducting denoising processing on the lubricating oil spot map image after graying processing by Gaussian filtering in order to reduce the influence of image noise on the collected lubricating oil spot map image, and utilizing a histogram equalization algorithm to conduct image enhancement operation on the lubricating oil spot map image after denoising processing in order to reduce the influence of image shooting environment and the influence of the three rings which are not easy to distinguish. Thus, the present embodiment obtains the lubricant oil spot map image after the image preprocessing at each time, that is, the oil spot gray scale image at each time. The implementation processes of the weighted average algorithm, the gaussian filtering algorithm and the histogram equalization algorithm are all the prior art, are out of the protection scope of the invention, and are not elaborated herein.
And (1-2) obtaining the oil stain area gray level image at each moment according to the oil stain gray level image at each moment.
In this embodiment, in order to facilitate analysis of image features of a ring in an oil spot grayscale image, a grayscale image including an oil spot region is obtained, where the oil spot region grayscale image is a feature image, and the specific steps are as follows: and inputting the oil stain gray level image at each moment into a pre-constructed and trained semantic segmentation neural network, and outputting the oil stain area gray level image at each moment. The architecture of the semantic segmentation neural network is a ResNet neural network, labels are divided into oil spots and a background, pixel points in an oil spot area of an oil spot gray image are marked as 1, pixel points in the background area are marked as 0, and a loss function of the semantic segmentation neural network is a cross entropy loss function. The construction and training process of the semantic segmentation neural network is the prior art and is not within the protection scope of the invention, and the detailed description is not provided herein.
(2) And clustering the oil stain area gray level images at each moment according to the first preset cluster K value, the second preset cluster K value, the third preset cluster K value and the oil stain area gray level images at each moment to obtain the oil stain cluster gray level images corresponding to the preset cluster K values at each moment.
In this embodiment, based on the priori knowledge of the lubricating oil spot map image, the number of rings of the oil spot region gray scale image at each time may be 3, 2, or 1, and since the number of rings has a certain relationship with the cluster K value, when the cluster analysis is performed on the oil spot region gray scale image at each time, the cluster K value may be 3, 2, or 1. For example, if the cluster K value is 3, the oil stain area gray level image is clustered by using a K-means clustering algorithm with the cluster K value of 3, where the cluster of the oil stain area gray level image is 3, that is, the number of the circular rings is 3. In order to obtain an oil stain clustering gray image with the best clustering effect at each moment, that is, to select an oil stain clustering gray image corresponding to a clustering K value with the best clustering effect from oil stain clustering gray images corresponding to different clustering K values, it is necessary to determine an oil stain clustering gray image corresponding to each preset clustering K value at each moment.
And clustering the oil stain area gray level images at each moment by using a K-means clustering algorithm corresponding to each preset clustering K value to obtain oil stain clustering gray level images corresponding to each preset clustering K value at each moment, namely the oil stain clustering gray level images corresponding to 3 preset clustering K values corresponding to each moment. The 3 preset cluster K values are a first preset cluster K value, a second preset cluster K value and a third preset cluster K value, the first preset cluster K value is set to 1, the second preset cluster K value is set to 2, and the third preset cluster K value is set to 3. The implementation process of the K-means clustering algorithm is the prior art, is out of the protection scope of the present invention, and is not described in detail herein.
(3) Determining a clustering effect evaluation value of the oil spot clustering gray level image corresponding to each preset clustering K value at each moment according to the number of pixel points in each clustering cluster, the gray value and the position of each pixel point, the gray value and the position of a clustering center, the position of an image center and the number of the clustering clusters in the oil spot clustering gray level image corresponding to each preset clustering K value at each moment, and further determining the oil spot clustering gray level image corresponding to the optimal preset clustering K value at each moment, wherein the method comprises the following steps of:
(3-1) determining a clustering effect evaluation value of the oil stain clustering gray level image corresponding to each preset clustering K value at each moment according to each pixel point in each clustering cluster in the oil stain clustering gray level image corresponding to each preset clustering K value at each moment, the gray value and the position of each pixel point, the gray value and the position of a clustering center and the position of an image center, wherein the steps comprise:
(3-1-1) determining the distance from each pixel point of each cluster to the image center and the distance from the cluster center to the image center according to the positions of each pixel point, cluster center and image center in each cluster in the oil spot cluster gray level image corresponding to each preset cluster K value at each moment.
It should be noted that each cluster presents a circular ring shape when clustering is performed on the oil spot clustering gray level image, so that clustering can be performed based on the distance from an image pixel point to the center of an image during clustering, and the distance from the image pixel point to the center of the image is the distance from the pixel point to the center of the image and the distance from the clustering center to the center of the image respectively.
In this embodiment, in order to obtain a clustering distance corresponding to each pixel point in each cluster in the subsequent step, based on the positions of each pixel point, a clustering center, and an image center in each cluster in the oil spot clustering grayscale image corresponding to each preset clustering K value at each timeCalculating the distance from each pixel point of each cluster to the center of the image, and recording the distance from the pixel point to the center of the image asCalculating the distance from the cluster center of each cluster to the image center, and recording the distance from the cluster center to the image center as. The process of calculating the position distance between any two image pixels is the prior art, is not within the protection scope of the invention, and is not elaborated herein.
(3-1-2) according to the gray values of all pixel points and cluster centers in each cluster in the oil spot clustering gray level image corresponding to each preset clustering K value at each moment, the distance from each pixel point to the image center and the distance from the cluster centers to the image center, determining the clustering distance corresponding to each pixel point in each cluster in the oil spot clustering gray level image corresponding to each preset clustering K value at each moment.
Firstly, it should be noted that the clustering distance is one of important indexes for evaluating the clustering effect of the clustering clusters, and when determining the clustering distance corresponding to each pixel point in each clustering cluster, not only the distance difference between each pixel point in the clustering cluster but also the gray level difference between each pixel point in the clustering cluster are considered, the distance difference is the difference between the distance from each pixel point to the image center and the distance from the clustering center to the image center, and the gray level difference is the difference between the gray level value of each pixel point and the gray level value of the clustering center.
In this embodiment, the clustering distance corresponding to each pixel point in each cluster can be calculated by the gray values of each pixel point and the cluster center in each cluster in the oil spot clustering gray level image corresponding to each preset clustering K value at each moment, the distance from each pixel point to the image center, and the distance from the cluster center to the image center, and the calculation formula is as follows:
wherein the content of the first and second substances,the first oil spot clustering gray level image corresponding to each preset clustering K value at each momentiThe clustering distance corresponding to each pixel point in each cluster,for the gray value of each pixel point in each cluster in the oil spot cluster gray image corresponding to each preset cluster K value at each moment,for the gray value of the cluster center in each cluster in the oil spot cluster gray image corresponding to each preset cluster K value at each moment,for the distance from each pixel point in each cluster in the oil spot clustering gray level image corresponding to each preset clustering K value at each moment to the center of the image,and the distance from the cluster center in each cluster to the image center in the oil spot cluster gray level image corresponding to each preset cluster K value at each moment.
It should be noted that, the gray value difference of the pixel points in the calculation formula of the clustering distance corresponding to each pixel point in each clustering clusterDistance difference between pixel point and image centerAnd imageClustering distance corresponding to prime pointDThe distance difference between the pixel point of any cluster and the center of the image is positive correlationThe smaller the difference between the gray values of the pixelsThe smaller the cluster is, the corresponding clustering distance of the pixel point of the clustering clusterDThe smaller the cluster size, the better the clustering effect of the cluster.
(3-1-3) determining the cluster clustering evaluation value of the oil spot clustering gray level image corresponding to each preset clustering K value at each moment according to the number of clustering clusters in the oil spot clustering gray level image corresponding to each preset clustering K value at each moment, the number of pixel points in each clustering cluster and the clustering distance corresponding to each pixel point in each clustering cluster.
In order to evaluate the cluster clustering effect of the oil stain clustering gray level images corresponding to the preset clustering K values at each moment from the overall perspective, the clustering distances corresponding to the pixel points in each clustering cluster in the oil stain clustering gray level images corresponding to the preset clustering K values at each moment are accumulated to obtain cluster clustering evaluation values of the oil stain clustering gray level images corresponding to the preset clustering K values at each moment, and the cluster clustering evaluation values of the oil stain clustering gray level images corresponding to different preset clustering K values are different.
In this embodiment, based on the number of clusters in the oil stain clustering gray level image corresponding to each preset clustering K value at each time, the number of pixels in each cluster, and the clustering distance corresponding to each pixel in each cluster, from the perspective of mathematical modeling, the cluster clustering evaluation value of the oil stain clustering gray level image corresponding to each preset clustering K value at each time is calculated, and the calculation formula is as follows:
wherein the content of the first and second substances,Wfor each hourCluster clustering evaluation values of the oil spot clustering gray level images corresponding to the preset cluster K values,Kthe number of clustering clusters in the oil spot clustering gray level image corresponding to each preset clustering K value at each moment,jthe number of pixel points in each cluster in the oil spot clustering gray level image corresponding to each preset clustering K value at each moment,the first oil spot clustering gray level image corresponding to each preset clustering K value at each momentiFirst in a clustercThe cluster distance corresponding to each pixel point,is a hyper-parameter.
It should be noted that the cumulative sum of the cluster distances corresponding to the cluster points of the cluster K values in the cluster K value-preset oil spot cluster gray level image is a negative correlation, and the cumulative sum of the cluster distances corresponding to the cluster points of the cluster K values in the cluster K value-preset oil spot cluster gray level image is a negative correlationThe larger the cluster K value is, the cluster cohesion evaluation value of the oil spot cluster gray level image corresponding to the preset cluster K value isWThe smaller will be.
(3-1-4) calculating the distance between every two adjacent clustering centers in the oil stain clustering gray level image corresponding to every preset clustering K value at every moment, and determining the inter-class clustering evaluation value of the oil stain clustering gray level image corresponding to every preset clustering K value at every moment according to the number of clustering clusters in the oil stain clustering gray level image corresponding to every preset clustering K value at every moment and the distance between every two adjacent clustering centers.
In order to determine the distance difference between each cluster in the oil stain cluster gray level image corresponding to each preset cluster K value, namely, in order to determine the inter-class cluster evaluation value of the oil stain cluster gray level image corresponding to each preset cluster K value, the distance between each adjacent cluster centers in the oil stain cluster gray level image corresponding to each preset cluster K value at each moment is calculated, then the distance between each adjacent cluster centers in the oil stain cluster gray level image is subjected to accumulation calculation, and the accumulation calculation is used as the inter-class cluster evaluation value of the oil stain cluster gray level image corresponding to the corresponding preset cluster K value.
In this embodiment, based on the number of cluster clusters in the oil spot cluster gray-scale image corresponding to each preset cluster K value at each time and the distance between each adjacent cluster centers, the inter-class cluster evaluation value of the oil spot cluster gray-scale image corresponding to each preset cluster K value at each time is calculated, and the calculation formula is as follows:
wherein the content of the first and second substances,Ecluster evaluation values among clusters of the oil spot cluster gray level images corresponding to the preset cluster K values at each moment,Kthe number of clustering clusters in the oil spot clustering gray level image corresponding to each preset clustering K value at each moment,the first oil spot clustering gray level image corresponding to each preset clustering K value at each momentiCluster center and number of individual clusteri-distance between cluster centers of 1 cluster.
It should be noted that the inter-class cluster evaluation value is in positive correlation with the distance between each adjacent cluster center, and the larger the distance between the adjacent cluster centers in the oil stain cluster gray level image corresponding to any one preset cluster K value at each time in the inter-class cluster evaluation value calculation formula is, the larger the inter-class cluster evaluation value of the oil stain cluster gray level image corresponding to the preset cluster K value is.
(3-1-5) acquiring the area of each cluster in the oil spot clustering gray level image corresponding to each preset clustering K value at each moment, the maximum inscribed circle area and the minimum circumscribed circle area corresponding to the fitting outline of each cluster, wherein the areas of the clusters, the maximum inscribed circle area and the minimum circumscribed circle area are the number of pixels in a region, and determining the circular degree of the oil spot clustering gray level image corresponding to each preset clustering K value at each moment according to the number of clusters in the oil spot clustering gray level image corresponding to each preset clustering K value at each moment, the maximum inscribed circle area and the minimum circumscribed circle area corresponding to the fitting outline of each cluster and the area of each cluster.
In this embodiment, each cluster in the oil stain clustering gray level image should be a circular ring, and the clustering effect of the oil stain clustering gray level image can be determined by detecting the degree of conformity of each cluster with the circular ring characteristics, and if the degree of conformity of each cluster in the oil stain clustering gray level image corresponding to a certain preset clustering K value with the circular ring characteristics is higher, it indicates that the clustering effect of the oil stain clustering gray level image corresponding to the preset clustering K value is better. In addition, it is worth to be noted that, in this embodiment, each cluster in the oil stain clustering gray-scale image corresponding to each preset cluster K value is sorted according to the sequence from the center to the edge, for example, the 1 st cluster in the oil stain clustering gray-scale image is marked as the first clusteriThen the 2 nd cluster isi+1, the 3 rd cluster isi+2. Only one cluster exists in the oil spot clustering gray level image corresponding to the first preset cluster K value (numerical value 1), and the cluster is an oil ring region; two cluster clusters exist in the oil spot clustering gray level image corresponding to the second preset cluster K value (numerical value 2), the 1 st cluster is a mixed region of a diffusion ring and a deposition ring, and the 2 nd cluster is an oil ring region; three clusters exist in the oil spot clustering gray level image corresponding to the third preset clustering K value (numerical value 3), the 1 st cluster is a sediment ring region, the 2 nd cluster is a diffusion ring region, and the 3 rd cluster is an oil ring region.
In order to facilitate the subsequent determination of the clustering effect of the oil stain clustering gray level image corresponding to each preset clustering K value, in the embodiment, the area of each clustering cluster in the oil stain clustering gray level image corresponding to each preset clustering K value at each moment, the maximum inscribed circle area and the minimum circumscribed circle area corresponding to the fitting contour of each clustering cluster are obtained first, the areas of the clustering clusters, the maximum inscribed circle area and the minimum circumscribed circle area are the numbers of pixel points in the region, the obtaining process of the areas of the clustering clusters, the maximum inscribed circle area and the minimum circumscribed circle area is the prior art, and is not described in detail here, and then based on the number of the clustering clusters in the oil stain clustering gray level image corresponding to each preset clustering K value at each moment, the maximum inscribed circle area and the minimum circumscribed circle area corresponding to the fitting contour of each cluster and the area of each clustering cluster, calculating the annular circularity of the oil spot clustering gray level image corresponding to each preset clustering K value at each moment by using relevant knowledge of mathematical modeling, wherein the calculation formula is as follows:
wherein, the first and the second end of the pipe are connected with each other,Qthe circularity of the ring of the oil spot clustering gray level image corresponding to each preset clustering K value at each moment,Kthe number of clustering clusters in the oil spot clustering gray level image corresponding to each preset clustering K value at each moment,clustering the oil spots in the gray level image corresponding to each preset cluster K value at each momentiThe area of each cluster is the number of pixel points in the cluster,clustering the oil spots in the gray level image corresponding to each preset cluster K value at each momentiThe minimum circumcircle area corresponding to the contour is fitted to each cluster,clustering the oil spots in the gray level image corresponding to each preset cluster K value at each momenti-1 cluster fitting contour corresponding maximum inscribed circle area whenMaximum inscribed circle areaIs 0.
First, theiMinimum circumscribed circle area corresponding to cluster fitting contourAnd a first step ofi-maximum inscribed circle area corresponding to 1 cluster fitted contourIs a difference ofiThe circle area corresponding to the contour is fitted to each cluster, in order to ensureiThe area of the circular ring corresponding to the fitted contour of each cluster is larger than that of the firstiArea of individual clusters, i.e. in order to ensure circularity of the ringThe value range of the cluster is 0 to 1, the annular circularity of the oil stain clustering gray level image is determined based on the maximum inscribed circle area and the minimum circumscribed circle area corresponding to the clustering cluster fitting contour, and compared with the condition that the fitting contour of each clustering cluster is the maximum inscribed circle area or the minimum circumscribed circle area, the accuracy of the annular circularity of the oil stain clustering gray level image is effectively improved.
It should be noted that the larger the cumulative sum of the ratio of the area of the torus corresponding to each cluster fitting contour in the oil spot cluster gray-scale image corresponding to any one preset cluster K value to the area of the cluster corresponding to the cluster fitting contour is, the larger the circularity of the torus of the oil spot cluster gray-scale image corresponding to the preset cluster K value is, the more similar the torus area corresponding to each cluster fitting contour in the oil spot cluster gray-scale image corresponding to the preset cluster K value is to the cluster area corresponding to the torus, that is, the more similar the cluster in the oil spot cluster gray-scale image corresponding to the preset cluster K value is to the torus characteristic.
(3-1-6) determining the clustering effect evaluation value of the oil spot clustering gray level image corresponding to each preset clustering K value at each moment according to the intra-class clustering evaluation value, the inter-class clustering evaluation value and the annular circularity of the oil spot clustering gray level image corresponding to each preset clustering K value at each moment.
In the embodiment, from 3 aspects, the 3 aspects are intra-class cluster evaluation values, inter-class cluster evaluation values and annular circularity, and the cluster effect of the oil spot cluster gray level image corresponding to each preset cluster K value at each moment is comprehensively analyzed and evaluated. Based on the intra-class cluster evaluation value, the inter-class cluster evaluation value and the relation between the annular circularity and the clustering effect of the oil spot clustering gray level image corresponding to each preset clustering K value at each moment, the clustering effect evaluation value of the oil spot clustering gray level image corresponding to each preset clustering K value at each moment is calculated by using the relevant knowledge analysis of mathematical modeling, and the calculation formula is as follows:
wherein, the first and the second end of the pipe are connected with each other,Pthe cluster effect evaluation value of the oil spot cluster gray level image corresponding to each preset cluster K value at each moment,Qthe circularity of the ring of the oil spot clustering gray level image corresponding to each preset clustering K value at each moment,Ecluster evaluation values among clusters of the oil spot cluster gray level images corresponding to the preset cluster K values at each moment,Wand the cluster clustering evaluation value of the oil spot clustering gray level image corresponding to each preset cluster K value at each moment.
It should be noted that, the larger the annular circularity of the oil stain cluster gray level image corresponding to any one preset cluster K value is, the larger the inter-class cluster evaluation value is, and the smaller the intra-class cluster evaluation value is, the larger the cluster effect evaluation value of the oil stain cluster gray level image corresponding to the preset cluster K value is, which indicates that the better the cluster effect of the oil stain cluster gray level image corresponding to the preset cluster K value is, and each oil stain cluster gray level image at each time has a corresponding cluster effect evaluation value.
And (3-2) determining the oil stain clustering gray level image corresponding to the optimal preset clustering K value at each moment according to the clustering effect evaluation value of the oil stain clustering gray level image corresponding to each preset clustering K value at each moment.
And if the clustering effect evaluation value of the oil spot clustering gray level image corresponding to a certain preset clustering K value at any moment is greater than a preset clustering effect evaluation threshold value, and the clustering effect evaluation value of the oil spot clustering gray level image corresponding to the preset clustering K value is the maximum value, judging that the oil spot clustering gray level image corresponding to the preset clustering K value is the oil spot clustering gray level image corresponding to the optimal preset clustering K value.
In this embodiment, the cluster effect evaluation value of the oil stain cluster gray level image corresponding to each preset cluster K value at each time, that is, the cluster effect evaluation value of each oil stain cluster gray level image at each time, is compared with a preset cluster effect evaluation threshold, and the preset cluster effect evaluation threshold may be set by an implementer according to the actual situation of the scene. And if the clustering effect evaluation value of a certain oil stain clustering gray level image at any moment is greater than the preset clustering effect evaluation threshold value, and the clustering effect evaluation value of the oil stain clustering gray level image is the maximum value in the clustering effect evaluation values of all the oil stain clustering gray level images at the moment, the preset clustering K value corresponding to the clustering effect evaluation value of the oil stain clustering gray level image is the optimal preset clustering K value at the moment.
To this end, the oil stain clustering gray level image corresponding to the optimal preset clustering K value at each time is obtained in this embodiment, and each time has the oil stain clustering gray level image corresponding to the optimal preset clustering K value corresponding thereto.
(4) And determining the oil spot boundary mixing fuzzy degree of the oil spot clustering gray level image corresponding to the optimal preset clustering K value at each moment according to the gradient value and the gradient direction of each pixel point in the oil spot clustering gray level image corresponding to the optimal preset clustering K value at each moment.
It should be noted that each time has an oil spot cluster gray level image corresponding to the optimal preset cluster K value corresponding to each time, where the optimal preset cluster K value may be a first preset cluster K value, a second preset cluster K value, or a third preset cluster K value, and in order to determine an optimization degree index corresponding to the oil spot cluster gray level image at each time later, it is necessary to determine an oil spot boundary mixing blur degree of the oil spot cluster gray level image corresponding to the optimal preset cluster K value at each time, where the oil spot boundary mixing blur degree and the optimization degree index are negative correlations, and the step of determining the oil spot boundary mixing blur degree includes:
and (4-1) when the optimal preset cluster K value at the nth moment is the first preset cluster K value, obtaining the oil spot boundary mixing fuzzy degree of the oil spot cluster gray level image corresponding to the first preset cluster K value at the nth moment.
In this embodiment, when the optimal preset cluster K value at the nth time is the first preset cluster K value, the first preset cluster K value is 1, and only one cluster exists in the oil stain cluster gray scale image corresponding to the nth time, which indicates that the oil ring in the oil stain area gray scale image at the nth time is seriously diffused due to the too late detection time, the oil stain area gray scale image at the nth time loses the reference value of the detection lubricating oil contamination degree, no excessive analysis is performed on the reference value, the oil stain boundary mixing blur degree of the oil stain cluster gray scale image corresponding to the first preset cluster K value at the nth time is assigned to be 1, and the oil stain boundary mixing blur degrees of the oil stain cluster gray scale images corresponding to the subsequent different preset cluster K values are normalized values, and the value range of the oil stain boundary mixing blur degrees is 0-1.
(4-2) when noWhen the best preset clustering K value of each moment is a second preset clustering K value, the first preset clustering K value is obtainedAnd mixing fuzzy degrees of oil spot boundaries of the oil spot clustering gray level images corresponding to the second preset clustering K value at each moment.
In addition, when the second stepWhen the optimal preset cluster K value at each moment is the second preset cluster K value, the second preset cluster K value is 2, and the optimal preset cluster K value of 2 is probably because the width of the diffusion ring is caused by poor capability of the clean dispersant or exhausted clean dispersantThe oil spot clustering gray level image is narrower and narrower or disappears, the diffusion layer and the deposition layer in the oil spot clustering gray level image are overlapped in the same layer, at the moment, the embodiment only needs to analyze whether the edge closest to the center of the image has the phenomena of fuzzy mixing or diffusion mixing and the like, namely, the oil spot boundary mixing fuzzy degree of the oil spot clustering gray level image is analyzed, the edge closest to the center of the image is the oil spot edge, and the oil spot edge can be the edge of a deposition ring, can be the edge of a diffusion ring, and can also be the edge formed by overlapping the deposition ring and the diffusion ring. The lower the mixed fuzzy degree of the oil spot boundary, the more likely the oil spot clustering gray level image at the corresponding moment is to be the optimal oil spot clustering gray level image, so the oil spot boundary mixed fuzzy degree of the oil spot clustering gray level image is taken as one of important indexes for determining the optimal oil spot clustering gray level image, and the first step is determinedThe step of mixing the fuzzy degree of the oil spot boundary of the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment comprises the following steps:
(4-2-1) toCarrying out edge detection on the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment to obtain the first preset clustering K valueAnd oil spot edge lines in the oil spot clustering gray level images corresponding to the second preset clustering K value at each moment.
In this embodiment, Sobel edge detection operator is used to perform the second stepCarrying out edge detection on the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment to obtain each edge in the oil spot clustering gray level image, and selecting the edge closest to the center of the image as an oil spot edge line to obtain the first preset clustering K valueAnd oil spot edge lines in the oil spot clustering gray level images corresponding to the second preset clustering K value at each moment are convenient for analyzing the oil spot edge lines subsequently. The Sobel edge detection operator is the prior art, is out of the protection scope of the invention, and is not explained in detail.
(4-2-2) according to theDetermining the gray value of each edge pixel point of the oil spot edge line in the oil spot clustering gray image corresponding to the second preset clustering K value at each moment and the number of the edge pixel pointsAnd mixing fuzzy factors of the first oil spot boundary of the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment.
It should be noted that, because the gradient values of the edge pixel points of the edge line are all greater than the gradient values of the other pixel points, and the gradient values of the edge pixel points of the edge line are all similar, the edge line is formed, but it cannot be guaranteed that the gray values of the edge pixel points of the edge line are also similar, so the mixed fuzzy degree of the oil stain edge line is determined by analyzing the gray value difference change of the edge pixel points of the oil stain edge line in this embodiment.
In this embodiment, based onGray values of all edge pixel points of oil spot edge lines in the oil spot clustering gray image corresponding to the second preset clustering K value at each moment and the number of the edge pixel points are calculatedAnd taking the mean value of the gray difference values of all adjacent edge pixel points of the oil spot edge lines in the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment as a first oil spot boundary mixed fuzzy factor, wherein the calculation formula is as follows:
wherein the content of the first and second substances,Ais as followsA first oil spot boundary mixing fuzzy factor of the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment,uis as followsThe number of edge pixel points of the oil spot edge line in the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment,is as followsThe first of the oil stain edge lines in the oil stain clustering gray level image corresponding to the second preset clustering K value of each momentiThe gray value of the pixel points at each edge,is as followsThe first of the oil stain edge lines in the oil stain clustering gray level image corresponding to the second preset clustering K value of each momenti+1 gray values of edge pixels.
In addition, the first stepFirst oil spot boundary mixed fuzzy factor of oil spot clustering gray level image corresponding to second preset clustering K value of each momentAThe larger, the description isSecond preset cluster of individual momentsThe fuzzy distribution of the oil spot edge lines of the oil spot clustering gray level image corresponding to the K value is. Reference toAnd determining a first oil spot boundary mixed fuzzy factor of the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment, wherein the first oil spot boundary mixed fuzzy factor of the oil spot clustering gray level image can be obtained when the optimal preset clustering K value at each moment is the second preset clustering K value.
(4-2-3) obtaining theThe gradient directions of all edge pixel points and eight neighborhood pixel points of an oil spot edge line in the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment are determined according to the gradient directions of the eight neighborhood pixel pointsDetermining the number of edge pixel points of an oil spot edge line in the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment, the gradient direction of each edge pixel point and the gradient direction of eight neighborhood pixel points of each edge pixel point, and determining the first preset clustering K valueAnd mixing fuzzy factors of a second oil spot boundary of the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment.
It should be noted that, according to the diffusion rule of the lubricant oil spot, when the diffusion ring is in a diffusion state, that is, the diffusion ring is not diffused to the fusion state, the gradient direction at the edge line of the oil spot is diffused from the center of the oil spot to the periphery of the oil spot, and based on the analysis of the gradient direction at the edge line of the oil spot, the first step can be determinedAnd mixing fuzzy factors of a second oil spot boundary of the oil spot cluster gray level image corresponding to the second preset cluster K value at each moment.
In this embodiment, first, theGet firstThe process of acquiring the eight-neighborhood pixels of the oil spot edge line in the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment is the prior art, and detailed description is not given here. Based onThe number of edge pixel points of an oil spot edge line in the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment, the gradient direction of each edge pixel point and the gradient direction of eight neighborhood pixel points of each edge pixel point are calculated, and the first preset clustering K value is calculatedThe average value of the gradient relation difference between each edge pixel point of the oil stain edge line in the oil stain cluster gray level image corresponding to the second preset cluster K value at each moment and eight neighborhood pixel points thereof is used as the gradient direction difference average value corresponding to each edge pixel point, then the average value of the gradient direction difference average values corresponding to all edge pixel points is used as a second oil stain boundary mixed fuzzy factor, and the calculation formula is as follows:
wherein the content of the first and second substances,Bis as followsA second oil spot boundary mixing fuzzy factor of the oil spot clustering gray level image corresponding to a second preset clustering K value at each moment,uis as followsThe number of edge pixel points of the oil spot edge line in the oil spot cluster gray level image corresponding to the second preset cluster K value at each moment,Is as followsThe first of the oil stain edge lines in the oil stain clustering gray level image corresponding to the second preset clustering K value of each momentiThe gradient direction of the pixel points at each edge,is as followsThe first of the oil stain edge lines in the oil stain clustering gray level image corresponding to the second preset clustering K value of each momentiThe first of each edge pixelvGradient direction of each neighborhood pixel.
In addition, the first stepThe larger the second oil stain boundary mixed fuzzy factor B of the oil stain cluster gray level image corresponding to the second preset cluster K value at each moment is, the more the gradient direction of the oil stain edge line is not accordant with the ideal diffusion direction, and the first timeThe more fuzzy the oil spot edge line of the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment. Reference toAnd determining a second oil spot boundary mixed fuzzy factor of the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment, wherein the second oil spot boundary mixed fuzzy factor of the oil spot clustering gray level image can be obtained when the optimal preset clustering K value at each moment is the second preset clustering K value.
(4-2-4) according to the second aspectDetermining a first oil spot boundary mixed fuzzy factor and a second oil spot boundary mixed fuzzy factor of the oil spot clustering gray level image corresponding to a second preset clustering K value at each momentAnd mixing fuzzy degrees of oil spot boundaries of the oil spot clustering gray level images corresponding to the second preset clustering K value at each moment.
In this embodiment, the oil spot boundary mixing fuzzy degree of the oil spot clustering gray level image corresponding to the second preset clustering K value is comprehensively analyzed from two aspects of the oil spot edge line, which is helpful for improving the accuracy of the oil spot boundary mixing fuzzy degree value. The second step based on step (4-2-2)And (3) determining the oil spot boundary mixing fuzzy degree by referring to the first oil spot boundary mixing fuzzy factor, the second oil spot boundary mixing fuzzy factor and the relation of the oil spot boundary mixing fuzzy degree of the first oil spot boundary mixing fuzzy factor, the second oil spot boundary mixing fuzzy factor and the oil spot boundary mixing fuzzy degree corresponding to the second preset cluster K value of each moment, wherein the calculation formula is as follows:
wherein, the first and the second end of the pipe are connected with each other,Tis as followsThe oil spot boundary mixing fuzzy degree of the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment,Ais as followsA first oil spot boundary mixing fuzzy factor of the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment,Bis as followsAnd mixing fuzzy factors of a second oil spot boundary of the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment.
The first oil spot boundary blending blurring factor, the second oil spot boundary blending blurring factor and the oil spot boundary blending blurring degree are in positive correlation, andfirst oil spot boundary mixed fuzzy factor of oil spot clustering gray level image corresponding to second preset clustering K value of each momentALarger, second oil spot boundary blending blurring factorBThe larger, theOil spot boundary mixing fuzzy degree of oil spot clustering gray level image corresponding to second preset clustering K value of each momentTThe larger, the description isThe more inaccurate the clustering hierarchy of the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment is, the more the excessive diffusion phenomenon may occur. Reference toAnd determining the oil spot boundary mixing fuzzy degree of the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment, so that the oil spot boundary mixing fuzzy degree of the oil spot clustering gray level image can be obtained when the optimal preset clustering K value at each moment is the second preset clustering K value.
(4-3) whenWhen the best preset cluster K value at each moment is a third preset cluster K value, the first preset cluster K value is obtainedAnd mixing fuzzy degrees of oil spot boundaries of the oil spot clustering gray level images corresponding to the third preset clustering K value at each moment.
In addition, the first stepWhen the optimal preset cluster K value at each moment is a third preset cluster K value, the third preset cluster K value is 3, and if the optimal preset cluster K value at each moment is the third preset cluster K valueThe oil spot cluster gray level image at each moment is excessively diffused, the radius of a diffusion ring in the oil spot cluster gray level image is reduced, and the ring characteristics of the diffusion ring are greatly changed, so that the embodiment focuses on the first stepPerforming key analysis on a diffusion ring in the oil spot clustering gray level image at each moment to determine the oil spot boundary mixing fuzzy degree, wherein the steps comprise:
(4-3-1) to the firstPerforming edge detection on the oil spot cluster gray level image corresponding to the second preset cluster K value at each moment to obtain the first preset cluster K valueAnd clustering the edge line of the diffusion ring in the gray image by using the oil spots corresponding to the third preset clustering K value at each moment, and further obtaining gradient values of each edge pixel point of the edge line of the diffusion ring and eight neighborhood pixel points of the edge line of the diffusion ring.
In this embodiment, the Sobel edge detection operator is used to detect the secondThe oil spot clustering gray level image corresponding to the second preset clustering K value at each moment is subjected to edge detection, so that 3 circular ring edges can be obtained, the radius value of the 3 circular ring edges is counted, the circular ring edge with the radius value positioned at the second is a diffusion ring edge line, and thus, the embodiment obtains the secondAnd (4) a diffusion ring edge line in the oil spot clustering gray level image corresponding to the third preset clustering K value at each moment. The edge detection process of the Sobel edge detection operator is the prior art, is out of the protection scope of the invention, and is not elaborated herein. To facilitate the subsequent determination of the diffusion ring boundary blending ambiguity factor, the present embodiment is based onAnd clustering the edge line of the diffusion ring in the gray level image according to the oil spot corresponding to the third preset clustering K value at each moment to obtain gradient values of all edge pixel points and eight neighborhood pixel points of the edge line of the diffusion ring.
(4-3-2) according to theGradient values of all edge pixel points of diffusion ring edge lines in the oil spot clustering gray level image corresponding to the third preset clustering K value at each moment and gradient values of eight neighborhood pixel points of the gradient pixel points, entropy values and variances of gradient co-occurrence matrixes corresponding to all edge pixel points of the diffusion ring edge lines are determined, and then the first preset clustering K value is determinedDiffusion ring boundary mixed fuzzy factors corresponding to the oil spot clustering gray level images corresponding to the third preset clustering K value at each moment comprise the following steps:
(4-3-2-1) according to theAnd determining the entropy value and the variance of a gradient co-occurrence matrix corresponding to each edge pixel point of the edge line of the diffusion ring in the oil spot clustering gray level image corresponding to the third preset clustering K value at each moment and the gradient values of eight neighborhood pixel points of the gradient pixel points of the edge line of the diffusion ring.
In this embodiment, based onAnd establishing a gradient co-occurrence matrix according to gradient values of all edge pixel points and eight neighborhood pixel points of the diffusion ring edge line in the oil spot clustering gray level image corresponding to the third preset clustering K value at each moment, further calculating entropy values and variances of the gradient co-occurrence matrix corresponding to all edge pixel points of the diffusion ring edge line, recording the entropy values of the gradient co-occurrence matrix as ENT, recording the variances of the gradient co-occurrence matrix as CON, and enabling each edge pixel point of the diffusion ring edge line to have a corresponding gradient co-occurrence matrix. The process of constructing the gradient co-occurrence matrix and the process of calculating the entropy and variance of the gradient co-occurrence matrix are all the prior art, are not within the scope of the present invention, and are not elaborated herein.
It should be noted that the entropy value of the gradient co-occurrence matrix is a randomness measure of image texture distribution information, the entropy value may represent a non-uniformity degree or a complexity degree of the gradient co-occurrence matrix, and the entropy value is larger when elements in the gradient co-occurrence matrix are distributed dispersedly. The variance of the gradient co-occurrence matrix is a measure of image texture period information, the variance can represent the texture periodicity of the gradient co-occurrence matrix, and the variance is larger when the texture period in the gradient co-occurrence matrix is larger.
(4-3-2-2) determining the second step according to the number of edge pixel points of the edge line of the diffusion ring and the entropy value and the variance of the gradient co-occurrence matrix corresponding to each edge pixel pointAnd (3) diffusion ring boundary mixed fuzzy factors corresponding to the oil spot clustering gray level images corresponding to the third preset clustering K value at each moment.
In this embodiment, since the entropy and variance of the gradient co-occurrence matrix and the mixed fuzzy factor of the boundary of the diffuser ring are in a positive correlation, the first and second thresholds can be calculated based on the number of edge pixels of the edge line of the diffuser ring and the entropy and variance of the gradient co-occurrence matrix corresponding to each edge pixel, and using the knowledge related to mathematical modelingDiffusion ring boundary corresponding to oil spot clustering gray level image corresponding to third preset clustering K value at each momentThe mixed fuzzy factor is calculated by the following formula:
wherein the content of the first and second substances,Cis a firstDiffusion ring boundary mixed fuzzy factors corresponding to the oil spot clustering gray level image corresponding to the third preset clustering K value at each moment,mis as followsThe number of edge pixel points of the edge line of the diffusion ring in the oil spot clustering gray level image corresponding to the third preset clustering K value at each moment,is as followsThe first diffusion ring edge line in the oil spot cluster gray level image corresponding to the third preset cluster K value of each momentiEntropy values of gradient co-occurrence matrixes corresponding to the edge pixel points,is as followsThe third preset cluster K value at each moment corresponds to the first diffusion ring edge line in the oil spot cluster gray level imageiThe variance of the gradient co-occurrence matrix corresponding to each edge pixel point,based on a natural constant eIs used as the exponential function of (1).
In addition, the first stepThe larger the entropy value and the variance of the gradient co-occurrence matrix corresponding to each edge pixel point of the diffusion ring edge line in the oil spot cluster gray level image corresponding to the third preset cluster K value at each moment are, the larger the diffusion ring boundary mixed fuzzy factor corresponding to the oil spot cluster gray level image is. Reference toAnd in the determining process of the diffusion ring boundary mixed fuzzy factor corresponding to the oil spot clustering gray level image corresponding to the third preset clustering K value at each moment, the diffusion ring boundary mixed fuzzy factor corresponding to the oil spot clustering gray level image when the optimal preset clustering K value at each moment is the third preset clustering K value can be obtained.
(4-3-3) according to theDetermining gradient values of all edge pixel points of diffusion ring edge lines in the oil spot clustering gray level image corresponding to the third preset clustering K value at each moment and gradient values of eight neighborhood pixel points of all edge pixel points in the oil spot clustering gray level image at each momentAnd the number of target neighborhood pixel points corresponding to each edge pixel point of the diffusion ring edge line in the oil spot clustering gray level image corresponding to the third preset clustering K value at each moment.
It should be noted that, when the edge line of the diffuser ring is in a diffusion state, the edge line of the diffuser ring becomes wider and more blurred, so the width of the edge line of the diffuser ring is one of the important indicators for subsequently determining the degree of blurring of the oil spot boundary mixture. And determining the width of the edge line of the diffusion ring by acquiring the number of neighborhood pixels close to the gradient value of each edge pixel.
In this embodiment, based onThird of the momentPresetting gradient values of all edge pixel points of a diffusion ring edge line in the oil spot clustering gray level image corresponding to the clustering K value and gradient values of eight neighborhood pixel points of the diffusion ring edge line, calculating gradient difference values of all edge pixel points and the corresponding eight neighborhood pixel points, and setting a gradient difference threshold value, wherein the gradient difference threshold value can be set by an implementer according to specific actual conditions. Counting the number of neighborhood pixels of which the gradient difference values are smaller than the gradient difference threshold value in the gradient difference values corresponding to the eight neighborhood pixels of each edge pixel, namely counting the number of neighborhood pixels of which the gradient difference values are smaller than the gradient difference threshold value in the gradient difference values corresponding to the eight neighborhood pixels of each edge pixel, taking the neighborhood pixels as target neighborhood pixels of the corresponding edge pixels, wherein at least 3 pixels close to the gradient values of the corresponding edge pixels generally exist in the eight neighborhood, namely each edge pixel at least corresponds to 3 target neighborhood pixels.
(4-3-4) according to theDetermining the number of target neighborhood pixels corresponding to each edge pixel point of the edge line of the diffusion ring in the oil spot clustering gray level image corresponding to the third preset clustering K value at each moment, the number of edge pixels of the edge line of the diffusion ring and a diffusion ring boundary mixed fuzzy factor, and determining the first preset clustering K value at each momentAnd mixing fuzzy degrees of oil spot boundaries of the oil spot cluster gray level images corresponding to the third preset cluster K value at each moment.
In this embodiment, the oil spot boundary mixing fuzzy degree of the oil spot clustering gray level image corresponding to the third preset clustering K value is analyzed from two angles of the width of the edge line of the diffusion ring and the characteristics of the ring, so that the accuracy of the oil spot boundary mixing fuzzy degree is effectively improved. Based onTarget neighborhoods corresponding to edge pixel points of diffusion ring edge lines in oil spot clustering gray level images corresponding to third preset clustering K value at each momentCalculating the number of pixels, the number of edge pixels of a diffusion ring edge line and a diffusion ring boundary mixed fuzzy factor by utilizing relevant knowledge of mathematical modeling, the number of target neighborhood pixels and the relevant relation between the diffusion ring boundary mixed fuzzy factor and the oil spot boundary mixed fuzzy degreeThe oil spot boundary mixing fuzzy degree of the oil spot clustering gray level image corresponding to the third preset clustering K value at each moment is calculated according to the following formula:
wherein the content of the first and second substances,is as followsThe oil spot boundary mixing fuzzy degree of the oil spot clustering gray level image corresponding to the third preset clustering K value at each moment,Cis as followsDiffusion ring boundary mixed fuzzy factors corresponding to the oil spot clustering gray level image corresponding to the third preset clustering K value at each moment,his as followsThe number of edge pixel points of the edge line of the diffusion ring in the oil spot clustering gray level image corresponding to the third preset clustering K value at each moment,is as followsThe third preset cluster K value at each moment corresponds to the first diffusion ring edge line in the oil spot cluster gray level imageiAn edge pixelThe number of the corresponding target neighborhood pixel points,based on a natural constant eIs used as the exponential function of (1).
It should be noted that the number of target neighborhood pixels and the mixed fuzzy degree of the diffusion ring boundary and the mixed fuzzy degree of the oil spot boundary are in a positive correlation relationship, and the larger the number of target neighborhood pixels, the larger the mixed fuzzy factor of the diffusion ring boundary, and the larger the mixed fuzzy degree of the oil spot boundary. Reference to the firstAnd determining the oil spot boundary mixing fuzzy degree of the oil spot clustering gray level image corresponding to the third preset clustering K value at each moment, and obtaining the oil spot boundary mixing fuzzy degree of the oil spot clustering gray level image when the optimal preset clustering K value at each moment is the third preset clustering K value.
To this end, based on the oil spot boundary mixing fuzzy degree of the oil spot clustering gray level image when the optimal preset clustering K value at each moment is the first preset clustering K value, the oil spot boundary mixing fuzzy degree of the oil spot clustering gray level image when the optimal preset clustering K value at each moment is the second preset clustering K value, and the oil spot boundary mixing fuzzy degree of the oil spot clustering gray level image when the optimal preset clustering K value at each moment is the third preset clustering K value, the oil spot boundary mixing fuzzy degree of the oil spot clustering gray level image corresponding to the optimal preset clustering K value at each moment is summarized and analyzed, and the calculation formula can be as follows:
wherein, the first and the second end of the pipe are connected with each other,oil corresponding to the optimal preset cluster K value at each momentThe degree of oil spot boundary mixture blur of the spot cluster gray scale image,the oil spot boundary mixing fuzzy degree of the corresponding oil spot clustering gray level image when the optimal preset clustering K value at any moment is the second preset clustering K value,the oil spot boundary mixing fuzzy degree of the corresponding oil spot clustering gray level image when the optimal preset clustering K value at any moment is the third preset clustering K value,kand the value of the K value of the cluster is optimally preset.
(5) Determining an optimal degree index corresponding to the oil stain clustering gray level image at each moment according to the oil stain boundary mixing fuzzy degree of the oil stain clustering gray level image corresponding to the time sequence number corresponding to each moment and the optimal preset clustering K value at each moment in a preset time period, and further determining the optimal oil stain clustering gray level image, wherein the method comprises the following steps of:
and (5-1) determining an optimization degree index corresponding to the oil stain clustering gray level image at each moment according to the time sequence number corresponding to each moment in the preset time period and the oil stain boundary mixing fuzzy degree of the oil stain clustering gray level image corresponding to the optimal preset clustering K value at each moment.
In this embodiment, when determining the oil spot boundary mixing blur degree of the oil spot cluster gray scale image corresponding to different preset cluster K values by combining the ring feature and the diffusion blur feature corresponding to different preset cluster K values, the method has a self-adaptive feature, which effectively improves the accuracy of the determined preference degree index, and calculates the preference degree index corresponding to the oil spot cluster gray scale image at each time based on the time sequence number corresponding to each time in the preset time period and the oil spot boundary mixing blur degree of the oil spot cluster gray scale image corresponding to the optimal preset cluster K value at each time by using the time sequence number corresponding to each time in the preset time period and the correlation between the oil spot boundary mixing blur degree and the preference degree index, and the calculation formula is as follows:
wherein the content of the first and second substances,is composed oftThe corresponding preference degree index of the oil spot cluster gray level image at the moment,is composed oftThe time sequence number corresponding to the time is,is composed oftThe oil spot boundary mixing fuzzy degree corresponding to the oil spot clustering gray level image at the moment,based on a natural constant eIs used as the exponential function of (1).
It should be noted that, the larger the time sequence number corresponding to any one time is, the longer the time for naturally drying the oil spots of the lubricating oil to be detected at that time is, the more real the diffusion degree of the map image of the oil spots corresponding to that time is, that is, the more sufficient the map image of the oil spots is diffused, the larger the preference degree index corresponding to the cluster gray level image of the oil spots corresponding to that time should be. Therefore, the larger the time sequence number corresponding to any one time is, the smaller the oil stain boundary mixing fuzzy degree corresponding to the oil stain clustering gray level image is, the larger the preference degree index corresponding to the oil stain clustering gray level image corresponding to the time is, and the more accurate the description of the oil stain clustering gray level image corresponding to the time on the lubricating oil pollution condition is.
And (5-2) determining the optimal oil spot cluster gray level image of the lubricating oil to be detected according to the optimization degree index corresponding to the oil spot cluster gray level image at each moment.
In this embodiment, if the preference degree index corresponding to the oil stain cluster gray level image at a certain time is greater than the preference degree threshold, and the preference degree index corresponding to the oil stain cluster gray level image at the certain time is the maximum value, it is determined that the oil stain cluster gray level image at the certain time is the optimal oil stain cluster gray level image of the lubricating oil to be detected. The preferred degree threshold is set to 0.7 in the embodiment, and the preferred degree threshold can be set by the implementer according to the specific practical situation. To this end, in this embodiment, according to the optimal preset cluster K value and the preference degree index at each time within the preset time period, an optimal oil spot cluster gray level image is screened from a plurality of oil spot cluster gray level images within the preset time period.
(6) The method comprises the steps of obtaining the average radius and the average gray value of a deposit ring in the optimal oil spot cluster gray image of the lubricating oil to be detected, determining the pollution degree evaluation value of the lubricating oil to be detected according to the average radius and the average gray value of the deposit ring in the optimal oil spot cluster gray image of the lubricating oil to be detected, and further determining the pollution degree of the lubricating oil to be detected, wherein the steps comprise:
(6-1) acquiring the average radius and the average gray value of the deposit ring in the optimal oil spot cluster gray image of the lubricating oil to be detected, and determining the pollution degree evaluation value of the lubricating oil to be detected according to the average radius and the average gray value of the deposit ring in the optimal oil spot cluster gray image of the lubricating oil to be detected.
In this embodiment, according to the relevant property of the deposit ring corresponding to the lubricating oil, when the lubricating oil is near to scrapping or heavily contaminated, the diameter of the deposit ring is small, and the color of the deposit ring is black, so as to determine the contamination degree of the lubricating oil to be detected. Firstly, obtaining the average radius and the average gray value of a deposition ring in the best oil spot clustering gray image of the lubricating oil to be detected, wherein the average radius is the average value of the distances from the center of the image to all boundary pixel points of the deposition ring, and the average gray value is the average value of the gray values of all the pixel points in the region of the deposition ring. Then, based on the average radius and the average gray value of the deposit ring in the best oil spot cluster gray level image of the lubricating oil to be detected, calculating the pollution degree evaluation value of the lubricating oil to be detected, wherein the calculation formula is as follows:
wherein the content of the first and second substances,Zas an evaluation value of the contamination degree of the lubricating oil to be detected,lthe average radius of the deposition ring in the optimal oil spot cluster gray level image of the lubricating oil to be detected,gthe average gray value of the deposition rings in the optimal oil spot cluster gray image of the lubricating oil to be detected,is based on a natural constant eIs determined by the exponential function of (a),based on a natural constant eIs used as an exponential function of (c).
It should be noted that the average radius of the deposit ring in the optimal oil spot cluster gray level image of the lubricating oil to be detectedlLarger, average gray scale valuegThe smaller the contamination degree evaluation value of the lubricating oil to be detectedZThe smaller the contamination, the less serious the contamination of the lubricating oil to be detected; average radius of deposition ring in optimal oil spot clustering gray level image of lubricating oil to be detectedlSmaller, average gray valuegThe greater the evaluation value of the contamination degree of the lubricating oil to be detectedZThe larger the contamination, the more serious the contamination of the lubricating oil to be detected.
And (6-2) determining the pollution degree of the lubricating oil to be detected according to the pollution degree evaluation value of the lubricating oil to be detected.
In this embodiment, according to a comparison result between the evaluation value of the contamination degree of the lubricating oil to be detected and the contamination degree threshold, whether the contamination degree of the lubricating oil to be detected is serious can be determined, if the evaluation value of the contamination degree of the lubricating oil to be detected is greater than the contamination degree threshold, the contamination degree of the lubricating oil to be detected is determined to be serious, otherwise, the contamination degree of the lubricating oil to be detected is determined to be not serious, and the magnitude of the contamination degree threshold can be set by an implementer according to actual specific conditions. Therefore, the embodiment realizes accurate detection of the pollution degree of the lubricating oil to be detected.
In the embodiment, the optimal oil spot clustering gray level image of the lubricating oil to be detected is obtained by performing image data processing on the lubricating oil spot atlas image of the lubricating oil to be detected, and then the pollution degree of the lubricating oil to be detected is determined based on the image characteristic information of the optimal oil spot clustering gray level image, so that the accuracy of detecting the pollution degree of the lubricating oil is effectively improved.
The above-mentioned embodiments are only used for illustrating the technical solutions of the present application, and not for limiting the same; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; the modifications or substitutions do not make the essence of the corresponding technical solutions deviate from the technical solutions of the embodiments of the present application, and are included in the protection scope of the present application.
Claims (10)
1. A lubricating oil pollution degree detection method based on image processing is characterized by comprising the following steps:
acquiring a lubricating oil spot map image of lubricating oil to be detected at each moment in a preset time period, and performing image preprocessing operation on the lubricating oil spot map image to obtain an oil spot gray image at each moment so as to obtain an oil spot area gray image at each moment;
clustering the oil stain area gray level images at each moment according to the first preset cluster K value, the second preset cluster K value, the third preset cluster K value and the oil stain area gray level images at each moment to obtain oil stain cluster gray level images corresponding to the preset cluster K values at each moment;
determining a clustering effect evaluation value of the oil spot clustering gray level image corresponding to each preset clustering K value at each moment according to the number of pixel points in each clustering cluster in the oil spot clustering gray level image corresponding to each preset clustering K value at each moment, the gray value and the position of each pixel point, the gray value and the position of a clustering center, the position of an image center and the number of clustering clusters, and further determining the oil spot clustering gray level image corresponding to the optimal preset clustering K value at each moment;
determining the oil spot boundary mixing fuzzy degree of the oil spot clustering gray level image corresponding to the optimal preset clustering K value at each moment according to the gradient value and the gradient direction of each pixel point in the oil spot clustering gray level image corresponding to the optimal preset clustering K value at each moment;
determining an optimal degree index corresponding to the oil spot clustering gray level image at each moment according to the oil spot boundary mixing fuzzy degree of the oil spot clustering gray level image corresponding to the time sequence number corresponding to each moment and the optimal preset clustering K value at each moment in a preset time period, and further determining the optimal oil spot clustering gray level image of the lubricating oil to be detected;
the average radius and the average gray value of the deposit ring in the optimal oil spot cluster gray image of the lubricating oil to be detected are obtained, the pollution degree evaluation value of the lubricating oil to be detected is determined according to the average radius and the average gray value of the deposit ring in the optimal oil spot cluster gray image of the lubricating oil to be detected, and then the pollution degree of the lubricating oil to be detected is determined.
2. The method for detecting the degree of contamination of lubricating oil based on image processing according to claim 1, wherein the step of determining the cluster effect evaluation value of the oil spot cluster gray level image corresponding to each preset cluster K value at each moment comprises:
determining the distance from each pixel point of each cluster to the image center and the distance from the cluster center to the image center according to the positions of each pixel point, the cluster center and the image center in each cluster in the oil spot clustering gray level image corresponding to each preset clustering K value at each moment;
determining a clustering distance corresponding to each pixel point in each clustering cluster in the oil spot clustering gray level image corresponding to each preset clustering K value at each moment according to the gray values of each pixel point and the clustering center in each clustering cluster in the oil spot clustering gray level image corresponding to each preset clustering K value at each moment, the distance from each pixel point to the image center and the distance from the clustering center to the image center;
determining the cluster clustering evaluation value of the oil spot clustering gray level image corresponding to each preset clustering K value at each moment according to the number of clustering clusters in the oil spot clustering gray level image corresponding to each preset clustering K value at each moment, the number of pixel points in each clustering cluster and the clustering distance corresponding to each pixel point in each clustering cluster;
calculating the distance between every two adjacent clustering centers in the oil stain clustering gray level image corresponding to every preset clustering K value at every moment, and determining the inter-class clustering evaluation value of the oil stain clustering gray level image corresponding to every preset clustering K value at every moment according to the number of the clustering clusters in the oil stain clustering gray level image corresponding to every preset clustering K value at every moment and the distance between every two adjacent clustering centers;
acquiring the area of each cluster in the oil stain clustering gray level image corresponding to each preset clustering K value at each moment, the maximum inscribed circle area and the minimum circumscribed circle area corresponding to the fitting outline of each cluster, wherein the area, the maximum inscribed circle area and the minimum circumscribed circle area of each cluster are the number of pixel points in the region, and determining the circular ring circularity of the oil stain clustering gray level image corresponding to each preset clustering K value at each moment according to the number of clusters in the oil stain clustering gray level image corresponding to each preset clustering K value at each moment, the maximum inscribed circle area and the minimum circumscribed circle area corresponding to the fitting outline of each cluster and the area of each cluster;
and determining the clustering effect evaluation value of the oil spot clustering gray level image corresponding to each preset clustering K value at each moment according to the cluster intra-cluster evaluation value, the cluster inter-cluster evaluation value and the annular circularity of the oil spot clustering gray level image corresponding to each preset clustering K value at each moment.
3. The method for detecting the degree of pollution of the lubricating oil based on the image processing as claimed in claim 2, wherein the formula for determining the circularity of the ring of the oil spot cluster gray level image corresponding to each preset cluster K value at each moment is as follows:
wherein the content of the first and second substances,Qthe circularity of the ring of the oil spot clustering gray level image corresponding to each preset clustering K value at each moment,Kthe number of the cluster clusters in the oil spot cluster gray level image corresponding to each preset cluster K value at each moment,clustering the oil spots in the gray level image corresponding to each preset cluster K value at each momentiThe area of each of the clusters to be clustered,clustering the oil spots in the gray level image corresponding to each preset cluster K value at each momentiThe minimum circumscribed circle area corresponding to the fitted contour of each cluster,clustering the oil spots in the gray level image corresponding to each preset cluster K value at each momenti-the maximum inscribed circle area for the fitted contours of 1 cluster.
4. The method for detecting the degree of contamination of lubricating oil based on image processing according to claim 1, wherein the step of determining the oil stain clustering gray level image corresponding to the optimal preset clustering K value at each moment further comprises:
and if the cluster effect evaluation value of the oil spot cluster gray level image corresponding to a certain preset cluster K value at any moment is greater than a preset cluster effect evaluation threshold value, and the cluster effect evaluation value of the oil spot cluster gray level image corresponding to the preset cluster K value is the maximum value, judging that the oil spot cluster gray level image corresponding to the preset cluster K value at the moment is the oil spot cluster gray level image corresponding to the optimal preset cluster K value.
5. The method for detecting the degree of contamination of lubricating oil based on image processing according to claim 1, wherein the step of determining the degree of blur of the oil spot boundary of the oil spot cluster gray level image corresponding to the optimal preset cluster K value at each moment comprises:
when the optimal preset clustering K value at the nth moment is the first preset clustering K value, obtaining the oil spot boundary mixing fuzzy degree of the oil spot clustering gray level image corresponding to the first preset clustering K value at the nth moment;
when it comes toWhen the best preset clustering K value at each moment is the second preset clustering K value, the first preset clustering K value is adjustedCarrying out edge detection on the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment to obtain the first preset clustering K valueOil spot edge lines in the oil spot clustering gray level images corresponding to the second preset clustering K value at each moment;
according to the firstDetermining the gray value of each edge pixel point of the oil stain edge line in the oil stain cluster gray image corresponding to the second preset cluster K value at each moment and the number of the edge pixel pointsA first oil spot boundary mixing fuzzy factor of the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment;
get the firstGradient directions of all edge pixel points and eight neighborhood pixel points of an oil spot edge line in the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment are determined according to the gradient directions of the eight neighborhood pixel pointsDetermining the number of edge pixel points of an oil spot edge line in the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment, the gradient direction of each edge pixel point and the gradient direction of eight neighborhood pixel points of each edge pixel point, and determining the first preset clustering K valueA second oil spot boundary mixing fuzzy factor of the oil spot clustering gray level image corresponding to a second preset clustering K value at each moment;
according to the firstDetermining a first oil spot boundary mixed fuzzy factor and a second oil spot boundary mixed fuzzy factor of the oil spot clustering gray level image corresponding to a second preset clustering K value at each momentAnd mixing fuzzy degrees of oil spot boundaries of the oil spot clustering gray level images corresponding to the second preset clustering K value at each moment.
6. The method for detecting the degree of contamination of lubricating oil based on image processing according to claim 1, wherein the step of determining the degree of blur of the oil spot boundary of the oil spot cluster gray level image corresponding to the optimal preset cluster K value at each moment further comprises:
when it comes toOptimum prediction of individual timeWhen the cluster K value is set as a third preset cluster K value, the first preset cluster K value is set as a second preset cluster K valueCarrying out edge detection on the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment to obtain the first preset clustering K valueThe diffusion ring edge line in the oil spot cluster gray image corresponding to the third preset cluster K value at each moment is obtained, and then gradient values of all edge pixel points of the diffusion ring edge line and eight neighborhood pixel points of the edge line are obtained;
according to the firstGradient values of all edge pixel points of diffusion ring edge lines in the oil spot clustering gray level image corresponding to the third preset clustering K value at each moment and gradient values of eight neighborhood pixel points of the gradient pixel points, entropy values and variances of gradient co-occurrence matrixes corresponding to all edge pixel points of the diffusion ring edge lines are determined, and then the first preset clustering K value is determinedDiffusion ring boundary mixed fuzzy factors corresponding to the oil spot clustering gray level images corresponding to the third preset clustering K value at each moment;
according to the firstDetermining gradient values of all edge pixel points of diffusion ring edge lines in the oil spot clustering gray level image corresponding to the third preset clustering K value at each moment and gradient values of eight neighborhood pixel points of all edge pixel points in the oil spot clustering gray level image at each momentThe number of target neighborhood pixels corresponding to each edge pixel point of the diffusion ring edge line in the oil spot clustering gray level image corresponding to the third preset clustering K value at each moment;
according to the firstDetermining the number of target neighborhood pixels corresponding to each edge pixel point of the edge line of the diffusion ring in the oil spot clustering gray level image corresponding to the third preset clustering K value at each moment, the number of edge pixels of the edge line of the diffusion ring and a diffusion ring boundary mixed fuzzy factor, and determining the first preset clustering K value at each momentAnd mixing fuzzy degrees of oil spot boundaries of the oil spot clustering gray level images corresponding to the third preset clustering K value at each moment.
7. The method for detecting the degree of contamination of lubricating oil based on image processing as claimed in claim 5, wherein the determining step is a step ofThe calculation formula of the second oil spot boundary mixed fuzzy factor of the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment is as follows:
wherein the content of the first and second substances,Bis as followsA second oil spot boundary mixing fuzzy factor of the oil spot clustering gray level image corresponding to a second preset clustering K value at each moment,uis a firstThe number of edge pixel points of the oil spot edge line in the oil spot clustering gray level image corresponding to the second preset clustering K value at each moment,is as followsThe first of the oil stain edge lines in the oil stain clustering gray level image corresponding to the second preset clustering K value of each momentiThe gradient direction of the pixel points at each edge,is as followsThe first of the oil stain edge lines in the oil stain clustering gray level image corresponding to the second preset clustering K value of each momentiThe first of each edge pixelvGradient direction of each neighborhood pixel.
8. The method for detecting the degree of contamination of lubricating oil based on image processing as claimed in claim 6, wherein the first step is further determinedThe calculation formula of the diffusion ring boundary mixed fuzzy factor corresponding to the oil spot clustering gray level image corresponding to the third preset clustering K value at each moment is as follows:
wherein the content of the first and second substances,Cis as followsDiffusion ring boundary mixed fuzzy factors corresponding to the oil spot clustering gray level image corresponding to the third preset clustering K value at each moment,mis as followsExpansion in oil spot clustering gray level image corresponding to third preset clustering K value of each momentThe number of edge pixel points of the edge line of the scattering ring,is as followsThe third preset cluster K value at each moment corresponds to the first diffusion ring edge line in the oil spot cluster gray level imageiEntropy values of gradient co-occurrence matrixes corresponding to the edge pixel points,is a firstThe first diffusion ring edge line in the oil spot cluster gray level image corresponding to the third preset cluster K value of each momentiThe variance of the gradient co-occurrence matrix corresponding to each edge pixel point,based on a natural constant eIs used as the exponential function of (1).
9. The method for detecting the degree of contamination of lubricating oil based on image processing according to claim 1, wherein the calculation formula for determining the degree of preference index corresponding to the oil spot cluster gray level image at each moment is as follows:
wherein the content of the first and second substances,is composed oftOptimization degree corresponding to oil spot clustering gray level image at momentThe index is a function of the number of the target,is composed oftThe time sequence number corresponding to the time is,is composed oftThe oil spot boundary mixing fuzzy degree corresponding to the oil spot clustering gray level image at the moment,based on a natural constant eIs used as the exponential function of (1).
10. The method for detecting the degree of contamination of lubricating oil based on image processing according to claim 1, wherein the step of further determining the degree of contamination of the lubricating oil to be detected comprises:
and if the evaluation value of the pollution degree of the lubricating oil to be detected is larger than the pollution degree threshold value, judging that the pollution degree of the lubricating oil to be detected is serious, otherwise, judging that the pollution degree of the lubricating oil to be detected is not serious.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211015805.4A CN115115621B (en) | 2022-08-24 | 2022-08-24 | Lubricating oil pollution degree detection method based on image processing |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211015805.4A CN115115621B (en) | 2022-08-24 | 2022-08-24 | Lubricating oil pollution degree detection method based on image processing |
Publications (2)
Publication Number | Publication Date |
---|---|
CN115115621A true CN115115621A (en) | 2022-09-27 |
CN115115621B CN115115621B (en) | 2022-11-11 |
Family
ID=83335730
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202211015805.4A Active CN115115621B (en) | 2022-08-24 | 2022-08-24 | Lubricating oil pollution degree detection method based on image processing |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115115621B (en) |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115311285A (en) * | 2022-10-12 | 2022-11-08 | 南通普盛动力有限公司 | Oil leakage identification method for diesel engine |
CN115690108A (en) * | 2023-01-04 | 2023-02-03 | 山东元旺电工科技有限公司 | Aluminum alloy rod production quality evaluation method based on image processing |
CN115841491A (en) * | 2023-02-24 | 2023-03-24 | 杭州电子科技大学 | Quality detection method of porous metal material |
CN115841488A (en) * | 2023-02-21 | 2023-03-24 | 聊城市飓风工业设计有限公司 | Hole checking method of PCB (printed Circuit Board) based on computer vision |
CN116152231A (en) * | 2023-04-17 | 2023-05-23 | 卡松科技股份有限公司 | Method for detecting impurities in lubricating oil based on image processing |
CN116229335A (en) * | 2023-05-10 | 2023-06-06 | 山东大佳机械有限公司 | Livestock and poultry farm environment recognition method based on image data |
CN116758071A (en) * | 2023-08-17 | 2023-09-15 | 青岛冠宝林活性炭有限公司 | Intelligent detection method for carbon electrode dirt under visual assistance |
CN116823834A (en) * | 2023-08-30 | 2023-09-29 | 济南玖通志恒信息技术有限公司 | Wastewater treatment grade detection method based on machine vision electrophoresis line discharge |
CN117115494B (en) * | 2023-10-23 | 2024-02-06 | 卡松科技股份有限公司 | Lubricating oil impurity pollution detection method and device based on artificial intelligence |
Citations (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2002296151A (en) * | 2001-03-29 | 2002-10-09 | Osaka Gas Co Ltd | Equipment diagnostic system, equipment diagnostic method and equipment diagnostic program using lubricating oil |
CN101196510A (en) * | 2007-12-25 | 2008-06-11 | 深圳市亚泰光电技术有限公司 | Method and device for detecting pollution degree of lubricating oil |
CN104865256A (en) * | 2015-04-01 | 2015-08-26 | 长安大学 | Multi-source information-fused quantitative determination method for detergency and dispersancy of automobile engine oil |
CN105758862A (en) * | 2016-01-26 | 2016-07-13 | 西安交通大学 | Online oxidative wear state monitoring method based on color extraction of wear particle image |
CN106908452A (en) * | 2017-04-24 | 2017-06-30 | 武汉理工大学 | Engine lubricating oil quality monitoring device based on machine vision |
JP2017215253A (en) * | 2016-06-01 | 2017-12-07 | 株式会社Ihi | Lubricant state identification device and lubricant state identification method |
CN108986096A (en) * | 2018-08-21 | 2018-12-11 | 燕山大学 | A kind of mixed oil liquid discrimination method based on three-dimensional fluorescence spectrum Yu Tchebichef square |
CN110415189A (en) * | 2019-07-22 | 2019-11-05 | 北京金交信息通信导航设计院 | A kind of marine oil overflow contamination monitoring system and method based on remote sensing |
CN112200854A (en) * | 2020-09-25 | 2021-01-08 | 华南农业大学 | Leaf vegetable three-dimensional phenotype measurement method based on video image |
CN113409271A (en) * | 2021-06-21 | 2021-09-17 | 广州文远知行科技有限公司 | Method, device and equipment for detecting oil stain on lens |
CN114219841A (en) * | 2022-02-23 | 2022-03-22 | 武汉欧耐德润滑油有限公司 | Automatic lubricating oil tank parameter identification method based on image processing |
CN114638827A (en) * | 2022-05-18 | 2022-06-17 | 卡松科技股份有限公司 | Visual detection method and device for impurities of lubricating oil machinery |
CN114782562A (en) * | 2022-06-18 | 2022-07-22 | 南通寝尚纺织品有限公司 | Garment fabric dip dyeing monitoring method based on data identification and artificial intelligence system |
CN114778553A (en) * | 2022-04-18 | 2022-07-22 | 中汽研汽车检验中心(天津)有限公司 | Oil spot detection system for rapidly identifying quality of lubricating oil |
CN114882265A (en) * | 2022-07-12 | 2022-08-09 | 济宁昆仑石油化工有限公司 | Method for detecting quality of finished oil |
-
2022
- 2022-08-24 CN CN202211015805.4A patent/CN115115621B/en active Active
Patent Citations (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2002296151A (en) * | 2001-03-29 | 2002-10-09 | Osaka Gas Co Ltd | Equipment diagnostic system, equipment diagnostic method and equipment diagnostic program using lubricating oil |
CN101196510A (en) * | 2007-12-25 | 2008-06-11 | 深圳市亚泰光电技术有限公司 | Method and device for detecting pollution degree of lubricating oil |
CN104865256A (en) * | 2015-04-01 | 2015-08-26 | 长安大学 | Multi-source information-fused quantitative determination method for detergency and dispersancy of automobile engine oil |
CN105758862A (en) * | 2016-01-26 | 2016-07-13 | 西安交通大学 | Online oxidative wear state monitoring method based on color extraction of wear particle image |
JP2017215253A (en) * | 2016-06-01 | 2017-12-07 | 株式会社Ihi | Lubricant state identification device and lubricant state identification method |
CN106908452A (en) * | 2017-04-24 | 2017-06-30 | 武汉理工大学 | Engine lubricating oil quality monitoring device based on machine vision |
CN108986096A (en) * | 2018-08-21 | 2018-12-11 | 燕山大学 | A kind of mixed oil liquid discrimination method based on three-dimensional fluorescence spectrum Yu Tchebichef square |
CN110415189A (en) * | 2019-07-22 | 2019-11-05 | 北京金交信息通信导航设计院 | A kind of marine oil overflow contamination monitoring system and method based on remote sensing |
CN112200854A (en) * | 2020-09-25 | 2021-01-08 | 华南农业大学 | Leaf vegetable three-dimensional phenotype measurement method based on video image |
CN113409271A (en) * | 2021-06-21 | 2021-09-17 | 广州文远知行科技有限公司 | Method, device and equipment for detecting oil stain on lens |
CN114219841A (en) * | 2022-02-23 | 2022-03-22 | 武汉欧耐德润滑油有限公司 | Automatic lubricating oil tank parameter identification method based on image processing |
CN114778553A (en) * | 2022-04-18 | 2022-07-22 | 中汽研汽车检验中心(天津)有限公司 | Oil spot detection system for rapidly identifying quality of lubricating oil |
CN114638827A (en) * | 2022-05-18 | 2022-06-17 | 卡松科技股份有限公司 | Visual detection method and device for impurities of lubricating oil machinery |
CN114782562A (en) * | 2022-06-18 | 2022-07-22 | 南通寝尚纺织品有限公司 | Garment fabric dip dyeing monitoring method based on data identification and artificial intelligence system |
CN114882265A (en) * | 2022-07-12 | 2022-08-09 | 济宁昆仑石油化工有限公司 | Method for detecting quality of finished oil |
Non-Patent Citations (4)
Title |
---|
J. POSTA ET AL: "APPROXIMATE TEST OF THE THERMAL DEGRADATION OF ENGINE OIL", 《6 TH INTERNATIONAL CONFERENCE ON TRENDS IN AGRICULTURAL ENGINEERING》 * |
M. S. LAGHARI ET AL: "Wear Particle Shape and Edge Detail Analysis", 《WEAR PARTICLE SHAPE AND EDGE DETAIL ANALYSIS》 * |
梁策 等: "主成分分析在柴油机润滑油磨粒分析中的应用", 《润滑与密封》 * |
肖梅 等: "汽车发动机润滑油清净分散性定量测定方法", 《长安大学学报(自然科学版)》 * |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115311285A (en) * | 2022-10-12 | 2022-11-08 | 南通普盛动力有限公司 | Oil leakage identification method for diesel engine |
CN115311285B (en) * | 2022-10-12 | 2023-04-14 | 南通普盛动力有限公司 | Oil leakage identification method for diesel engine |
CN115690108A (en) * | 2023-01-04 | 2023-02-03 | 山东元旺电工科技有限公司 | Aluminum alloy rod production quality evaluation method based on image processing |
CN115841488A (en) * | 2023-02-21 | 2023-03-24 | 聊城市飓风工业设计有限公司 | Hole checking method of PCB (printed Circuit Board) based on computer vision |
CN115841491A (en) * | 2023-02-24 | 2023-03-24 | 杭州电子科技大学 | Quality detection method of porous metal material |
CN116152231A (en) * | 2023-04-17 | 2023-05-23 | 卡松科技股份有限公司 | Method for detecting impurities in lubricating oil based on image processing |
CN116229335A (en) * | 2023-05-10 | 2023-06-06 | 山东大佳机械有限公司 | Livestock and poultry farm environment recognition method based on image data |
CN116758071A (en) * | 2023-08-17 | 2023-09-15 | 青岛冠宝林活性炭有限公司 | Intelligent detection method for carbon electrode dirt under visual assistance |
CN116758071B (en) * | 2023-08-17 | 2023-11-03 | 青岛冠宝林活性炭有限公司 | Intelligent detection method for carbon electrode dirt under visual assistance |
CN116823834A (en) * | 2023-08-30 | 2023-09-29 | 济南玖通志恒信息技术有限公司 | Wastewater treatment grade detection method based on machine vision electrophoresis line discharge |
CN116823834B (en) * | 2023-08-30 | 2023-11-24 | 江苏锡澄环境科学研究院有限公司 | Wastewater treatment grade detection method based on machine vision electrophoresis line discharge |
CN117115494B (en) * | 2023-10-23 | 2024-02-06 | 卡松科技股份有限公司 | Lubricating oil impurity pollution detection method and device based on artificial intelligence |
Also Published As
Publication number | Publication date |
---|---|
CN115115621B (en) | 2022-11-11 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN115115621B (en) | Lubricating oil pollution degree detection method based on image processing | |
CN114723701B (en) | Gear defect detection method and system based on computer vision | |
Chen et al. | A self organizing map optimization based image recognition and processing model for bridge crack inspection | |
CN114529549B (en) | Cloth defect labeling method and system based on machine vision | |
Shanmugamani et al. | Detection and classification of surface defects of gun barrels using computer vision and machine learning | |
Yao et al. | Deep-learning-based bughole detection for concrete surface image | |
CN101620060B (en) | Automatic detection method of particle size distribution | |
CN110148162A (en) | A kind of heterologous image matching method based on composition operators | |
CN109444169A (en) | A kind of bearing defect detection method and system | |
CN115330783A (en) | Steel wire rope defect detection method | |
CN115222733A (en) | Metal component surface defect detection method based on image recognition | |
CN101153850A (en) | Method and system for detecting asphalt mixture | |
CN109035249B (en) | Pipeline fault parallel global threshold detection method based on image processing | |
CN116385450B (en) | PS sheet wear resistance detection method based on image processing | |
CN115170574B (en) | Sewage detection method and system based on threshold segmentation | |
CN114549981A (en) | Intelligent inspection pointer type instrument recognition and reading method based on deep learning | |
CN115115638B (en) | Oil leakage detection and judgment method for hydraulic system | |
CN112215790A (en) | KI67 index analysis method based on deep learning | |
Begelrnan et al. | Cell nuclei segmentation using fuzzy logic engine | |
CN114596551A (en) | Vehicle-mounted forward-looking image crack detection method | |
CN116823834B (en) | Wastewater treatment grade detection method based on machine vision electrophoresis line discharge | |
CN115496692A (en) | Lubricating oil abrasive particle image enhancement method | |
Wang et al. | Texture extraction of wear particles based on improved random Hough transform and visual saliency | |
CN115496794A (en) | Concrete crack width calculation method and system, computer device and storage medium | |
CN105405137A (en) | Shaft quality detection method based on machine vision |
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 |