CN117576104A - Visual detection method for health state of ultrafiltration membrane in purification process - Google Patents
Visual detection method for health state of ultrafiltration membrane in purification process Download PDFInfo
- Publication number
- CN117576104A CN117576104A CN202410065311.XA CN202410065311A CN117576104A CN 117576104 A CN117576104 A CN 117576104A CN 202410065311 A CN202410065311 A CN 202410065311A CN 117576104 A CN117576104 A CN 117576104A
- Authority
- CN
- China
- Prior art keywords
- ultrafiltration membrane
- distribution
- area
- preset
- preset direction
- 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.)
- Pending
Links
- 239000012528 membrane Substances 0.000 title claims abstract description 135
- 238000000108 ultra-filtration Methods 0.000 title claims abstract description 131
- 238000001514 detection method Methods 0.000 title claims abstract description 51
- 238000000746 purification Methods 0.000 title claims abstract description 26
- 230000036541 health Effects 0.000 title claims abstract description 15
- 230000000007 visual effect Effects 0.000 title claims abstract description 11
- 238000009826 distribution Methods 0.000 claims abstract description 172
- 230000002159 abnormal effect Effects 0.000 claims abstract description 104
- 230000011218 segmentation Effects 0.000 claims abstract description 25
- 230000000875 corresponding effect Effects 0.000 claims description 63
- 238000000034 method Methods 0.000 claims description 53
- 230000003862 health status Effects 0.000 claims description 12
- 238000010606 normalization Methods 0.000 claims description 8
- 230000002596 correlated effect Effects 0.000 claims description 6
- 230000005856 abnormality Effects 0.000 claims description 5
- 238000012545 processing Methods 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 238000011179 visual inspection Methods 0.000 claims 9
- 238000007689 inspection Methods 0.000 claims 3
- 230000000694 effects Effects 0.000 abstract description 12
- 230000008569 process Effects 0.000 description 11
- 238000004364 calculation method Methods 0.000 description 8
- 238000004458 analytical method Methods 0.000 description 7
- 230000008859 change Effects 0.000 description 7
- 238000001914 filtration Methods 0.000 description 6
- 239000007788 liquid Substances 0.000 description 4
- 238000010586 diagram Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 239000000126 substance Substances 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000002547 anomalous effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 235000013361 beverage Nutrition 0.000 description 1
- 230000002146 bilateral effect Effects 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 238000010924 continuous production Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011010 flushing procedure Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/40—Analysis of texture
-
- 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
- G06T2207/10068—Endoscopic 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
- G06T2207/20028—Bilateral filtering
-
- 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
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
Abstract
The invention relates to the technical field of region segmentation, in particular to a visual detection method for the health state of an ultrafiltration membrane in a purification process. According to the invention, the texture rule index is obtained according to the distribution and arrangement conditions of the holes in the ultrafiltration membrane through the hole areas which are firstly divided; obtaining suspected abnormal areas according to the shape differences among the hole areas in different preset directions; obtaining the edge anomaly degree in different preset directions through the deviation degree of the distribution positions between the suspected anomaly region and other hole regions; and adjusting the gray value of the pixel points according to the position distribution and the edge anomaly degree of the pixel points in the suspected abnormal region and the texture rule index to obtain a new ultrafiltration membrane surface gray image, and carrying out state detection on a detection hole region obtained by dividing the new ultrafiltration membrane surface gray image. According to the invention, the accuracy of region segmentation is improved and the state detection effect is enhanced by improving the gray contrast of local details.
Description
Technical Field
The invention relates to the technical field of region segmentation, in particular to a visual detection method for the health state of an ultrafiltration membrane in a purification process.
Background
Ultrafiltration membranes are an important membrane technology commonly used to separate, filter and concentrate substances in solution. The film has a microporous structure, can selectively separate different molecules or particles, and is widely applied to the fields of water treatment, food and beverage, pharmacy, chemical industry, biotechnology and the like. Through periodic detection, the change of the performance of the membrane along with time can be known, so that measures can be taken in time to maintain or replace the membrane, the service life of the ultrafiltration membrane can be prolonged, and the maintenance and replacement cost is reduced, so that the health status monitoring of the filtration membrane is particularly important for industrial application of continuous production.
In the use process of the ultrafiltration membrane, the ultrafiltration membrane is deformed due to the impact of liquid, the adaptive segmentation based on gray level is realized by adopting an Ojin threshold segmentation method in the prior art, and the detection area is obtained for detection, but the ultrafiltration membrane has certain texture arrangement and is influenced by different liquid flushing in the purification process, so that the edge of the obtained image corresponding to deformation is fuzzy, the gray level contrast is not strong, the clearer and more accurate detection area can not be segmented, and the effect of detecting the state is poorer.
Disclosure of Invention
In order to solve the technical problems that a clearer and accurate detection area cannot be segmented in the prior art and the effect of detecting the state is poor, the invention aims to provide a visual detection method for the health state of an ultrafiltration membrane in the purification process, and the adopted technical scheme is as follows:
the invention provides a visual detection method for the health state of an ultrafiltration membrane in a purification process, which comprises the following steps:
acquiring a gray level image of the surface of the ultrafiltration membrane; dividing the gray level image on the surface of the ultrafiltration membrane to obtain a hole area; obtaining a texture rule index of the ultrafiltration membrane surface gray level image according to the distribution regularity of the hole area in the ultrafiltration membrane surface gray level image;
screening suspected abnormal areas from the hole areas according to the shape difference of each hole area and other hole areas in different preset directions; obtaining the edge anomaly degree of each suspected abnormal region in each preset direction according to the deviation degree of the distribution position between each suspected abnormal region and other hole regions in different preset directions;
updating the gray values of the pixels according to the position distribution of each pixel in the suspected abnormal region, the edge anomaly degree under different preset directions and the texture rule index of the gray image of the ultrafiltration membrane surface to obtain a new gray image of the ultrafiltration membrane surface; and dividing the new ultrafiltration membrane surface gray level image to obtain a detection hole area, and detecting the state according to the detection hole area.
Further, the method for obtaining the texture rule index comprises the following steps:
dividing the gray level image of the ultrafiltration membrane surface into a preset dividing number of dividing areas along a preset dividing direction; all the dividing areas have the same interval in the vertical direction of the preset dividing direction;
counting the number of the hole areas in each divided area as distribution number, and sequencing the distribution number according to the continuous adjacent distribution sequence of the divided areas to obtain a distribution sequence;
calculating the absolute value of kurtosis of the distribution quantity in the distribution sequence as a standard distribution index; calculating the absolute value of the deviation of the distribution quantity in the distribution sequence as a deviation distribution index; and obtaining the texture rule index of the ultrafiltration membrane surface gray level image according to the standard distribution index and the deviation distribution index, wherein the standard distribution index and the deviation distribution index are in negative correlation with the texture rule index.
Further, the method for acquiring the suspected abnormal region comprises the following steps:
for any one preset direction, taking each hole area as a target area in sequence, and acquiring the edge of the target area; obtaining the region distance of the target region in the preset direction according to the distance between the lower edges of the target region in the preset direction;
Taking the variance of the area distance of all the hole areas in the preset direction as a distance fluctuation value; taking the average value of the area distances of all the hole areas in the preset direction as a standard distance;
taking the difference between the area distance of the target area in the preset direction and the standard distance as a target fluctuation value of the target area in the preset direction; taking the ratio of the target fluctuation value of the target area in the preset direction to the distance fluctuation value as the direction anomaly of the target area in the preset direction; normalizing the sum of the directional anomalies of the target area in all preset directions to obtain the overall anomaly of the target area;
when the overall anomaly degree is greater than a preset anomaly degree threshold value, the corresponding hole area is used as a suspected anomaly area.
Further, the method for acquiring the region distance comprises the following steps:
if two edge points exist on a straight line corresponding to the preset direction in the edge pixel point of the target area, the corresponding two edge points are used as an edge matching pair; all edge matching pairs corresponding to the target area are obtained on all straight lines corresponding to each preset direction;
calculating the distance between two edge points in each edge matching pair as the direction distance of each edge matching pair; and taking the average value of the direction distances of all the edge matching pairs in each preset direction as the area distance of the target area in the corresponding preset direction.
Further, the method for acquiring the edge anomaly degree comprises the following steps:
obtaining a distribution index of the hole areas in each preset direction according to the distance distribution condition between each hole area and the adjacent hole areas in each preset direction; sequentially taking the hole areas except the suspected abnormal areas as reference areas, and taking the average value of the distribution indexes of all the reference areas in each preset direction as the distribution influence index in each preset direction;
for any suspected abnormal region, acquiring a distribution index of the suspected abnormal region in each preset direction; and calculating the difference between the distribution index and the distribution influence index of the suspected abnormal region in each preset direction, and obtaining the edge anomaly degree of the suspected abnormal region in each preset direction.
Further, the method for obtaining the distribution index comprises the following steps:
calculating the distance between the center points corresponding to two adjacent hole areas in each preset direction to obtain the distribution distance of each hole area in the corresponding preset direction; and carrying out normalization processing on the ratio of the area distance to the distribution distance of each hole area in each preset direction to obtain the distribution index of the hole area in each preset direction.
Further, updating the gray value of the pixel point to obtain a new gray image of the ultrafiltration membrane surface according to the position distribution of each pixel point in the suspected abnormal region, the edge anomaly degree in different preset directions and the texture rule index of the gray image of the ultrafiltration membrane surface, including:
sequentially taking each suspected abnormal region as an adjustment region, and calculating the distance between each pixel point and two edges of the adjustment region in each preset direction for any pixel point in the adjustment region to obtain the pixel distribution distance of the pixel point in the corresponding preset direction; taking the minimum pixel distribution interval in each preset direction as the position weight of the pixel point in the corresponding preset direction;
obtaining local abnormal adjustment weights of the pixel points according to the position weights of the pixel points in each preset direction and the corresponding edge abnormal degrees in the adjustment areas; according to the local abnormal adjustment weight of the pixel point and the texture rule index of the ultrafiltration membrane surface gray level image, obtaining the gray level adjustment weight of the pixel point; the local abnormal adjustment weight and the gray adjustment weight are positively correlated, the texture rule index and the gray adjustment weight are negatively correlated, and the gray adjustment weight is a normalized value;
Taking the product of the gray scale adjustment weight of each pixel point and the corresponding gray scale value as the adjustment value of each pixel point; taking the sum value of the adjustment value and the gray value of each pixel point as a new gray value of each pixel point; and obtaining a new ultrafiltration membrane surface gray scale image based on the new gray scale value of each pixel point.
Further, the method for acquiring the local anomaly adjustment weight comprises the following steps:
taking the product of the position weight and the edge anomaly degree of each pixel point in each preset direction as an anomaly adjustment index of each pixel point in each preset direction;
and taking the accumulated value of the abnormal adjustment indexes of each pixel point in all preset directions as the local abnormal adjustment weight of each pixel point.
Further, the dividing the new ultrafiltration membrane surface gray level image to obtain a detection hole area, and performing state detection according to the detection hole area, including:
dividing a new ultrafiltration membrane surface gray level image by an Ojin threshold segmentation method to obtain a detection hole area;
matching each detection hole area with a preset standard hole area by adopting a shape matching algorithm to obtain the matching degree of each detection hole area; when the matching degree is smaller than a preset matching threshold value, the corresponding detection hole area is used as an abnormal area;
When the total number of the abnormal areas is larger than a preset abnormal threshold value, marking the health state of the corresponding ultrafiltration membrane as poor; and when the number of the abnormal areas is smaller than or equal to a preset abnormal threshold value, the health state of the corresponding ultrafiltration membrane is recorded as good.
Further, the dividing the gray scale image on the surface of the ultrafiltration membrane to obtain a hole area includes:
dividing the gray level image on the surface of the ultrafiltration membrane by using an Ojin threshold segmentation method, and taking the connected domain with the lowest gray level value as a hole connected domain;
and taking a region formed by the corresponding range of the hole communicating region in the ultrafiltration membrane surface gray level image as a hole region of the ultrafiltration membrane surface gray level image.
The invention has the following beneficial effects:
according to the invention, through the hole areas which are firstly divided, according to the distribution arrangement condition of holes in the ultrafiltration membrane, the abnormal condition of the ultrafiltration membrane is analyzed from a macroscopic angle, so as to obtain a texture rule index, and the abnormal degree is analyzed integrally. Further, analysis is performed from local details, because the ultrafiltration membrane is generally provided with regularly arranged hole areas, and the hole areas are regular rectangles, the areas with anomalies can be primarily screened out according to the degree of shape difference, namely, the suspected anomaly areas are obtained according to the shape differences in different preset directions, the degree of deviation of distribution positions between the suspected anomaly areas and other hole areas is utilized to obtain the edge anomaly degree in different preset directions from the position deviation, and more adjustment needs to be performed on nearby pixel points at the places with the more anomalous edges so as to enhance gray scale contrast, thereby facilitating subsequent clear segmentation. The position distribution and the edge anomaly degree of the pixel points in the suspected anomaly region are utilized, the abnormal influence condition of each pixel point is obtained through local details, the gray value of the pixel point is comprehensively adjusted at a macroscopic angle by combining with a texture rule index, a new ultrafiltration membrane surface gray image is obtained, the gray values of the pixel points in the new ultrafiltration membrane surface gray image are optimized, the contrast of the local details is enhanced, the detection hole region obtained by dividing the new ultrafiltration membrane surface gray image is clearer and more accurate, and further the effect of state detection is better. According to the invention, the accuracy of region segmentation is improved and the state detection effect is enhanced by improving the gray contrast of local details.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions and advantages of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are only some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a visual detection method for the health status of an ultrafiltration membrane in a purification process according to an embodiment of the present invention;
fig. 2 is a schematic diagram of local texture distribution in a gray scale image of an ultrafiltration membrane surface according to an embodiment of the present invention.
Detailed Description
In order to further describe the technical means and effects adopted by the invention to achieve the preset aim, the following is a detailed description of specific embodiments, structures, features and effects of the visual detection method for the health state of the ultrafiltration membrane in the purification process according to the invention with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
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 following specifically describes a specific scheme of the visual detection method for the health state of the ultrafiltration membrane in the purification process provided by the invention with reference to the accompanying drawings.
Referring to fig. 1, a flow chart of a visual detection method for health status of an ultrafiltration membrane in a purification process according to an embodiment of the invention is shown, the method comprises the following steps:
s1: acquiring a gray level image of the surface of the ultrafiltration membrane; dividing the gray level image on the surface of the ultrafiltration membrane to obtain a hole area; and obtaining a texture rule index of the ultrafiltration membrane surface gray level image according to the distribution regularity of the hole area in the ultrafiltration membrane surface gray level image.
The method comprises the steps of collecting an ultrafiltration membrane surface image in a purification process through an endoscope, and carrying out image pretreatment to obtain an ultrafiltration membrane surface gray level image, wherein the image pretreatment process can specifically comprise image graying treatment, filtering denoising treatment, background removing treatment and the like, the image pretreatment process is a technical means well known to a person skilled in the art, graying can be specifically carried out by a gray weighting method, filtering denoising and background difference method are carried out by a bilateral filtering method, background removing and the like, and the description is omitted herein. In order to facilitate the accurate calculation of the ultrafiltration membrane surface, the Hough straight line is used for acquiring the straight line texture in the ultrafiltration membrane surface gray level image in the image preprocessing process, and the placement angle of the ultrafiltration membrane surface gray level image is adjusted, so that the straight line texture in the ultrafiltration membrane surface gray level image analyzed later is distributed in the horizontal and vertical directions, and the subsequent analysis is facilitated.
The arrangement of the apertures on the undeformed ultrafiltration membrane is regular, i.e. the arrangement of the pixels on the texture on the image is regular. Referring to fig. 2, a schematic diagram of local texture distribution in a gray scale image of an ultrafiltration membrane surface according to an embodiment of the present invention is shown, wherein the basic distribution arrangement of textures is in a grid shape, the white portion is a linear texture portion, and the linear textures in fig. 2 are distributed in horizontal and vertical directions.
When the ultrafiltration membrane is deformed or the like, the corresponding texture is distorted, and the shape of a filtered hole area is changed, so that the hole area in the ultrafiltration membrane surface gray image is further initially segmented, the ultrafiltration membrane surface gray image is segmented to obtain the hole area, preferably, the ultrafiltration membrane surface gray image is segmented by an Ojin threshold segmentation method, a connected area with the lowest gray value is used as the hole connected area, the Ojin threshold segmentation method is based on the gray level corresponding to the maximum inter-class variance as a segmentation threshold, the image can be divided into two parts, wherein the part with the low gray value corresponds to the hole area, the area formed by the corresponding range of the hole connected area in the ultrafiltration membrane surface gray image is used as the hole area of the ultrafiltration membrane surface gray image, the pixel points in the hole connected area are the pixel points of the hole part, the formed area range is the hole area, and the black regular rectangular part is the hole area in the view shown in fig. 2. It should be noted that, the oxford thresholding method is a technical means well known to those skilled in the art, and will not be described herein.
When the ultrafiltration membrane is impacted by liquid and foreign matters to generate abnormality, the hole area is deformed, and the corresponding edge texture is stretched, so that the edge characteristics of the edge part in the image are weakened, the gray scale contrast condition is not obvious, and a clear segmentation result cannot be obtained. Therefore, the abnormal situation is comprehensively analyzed through the macroscopic angle and the local detail angle, firstly, the macroscopic judgment is carried out, because the ultrafiltration membrane in the purification process is used, the image obtained through the endoscope is round, and the hole areas are regularly arranged in the round image, under normal conditions, the quantity distribution of the hole areas in the horizontal or vertical direction is normally distributed, namely, the quantity of the holes in the horizontal or vertical direction is in a change trend from small to large to small. Therefore, according to the distribution regularity of the hole area in the ultrafiltration membrane surface gray level image, the texture regularity index of the ultrafiltration membrane surface gray level image is obtained.
Preferably, the gray level image of the ultrafiltration membrane surface is divided into a preset number of divided areas along a preset dividing direction, and all the divided areas have the same interval in the vertical direction of the preset dividing direction. In the embodiment of the invention, the preset dividing direction is the vertical direction, the preset dividing number is 30, namely, the ultrafiltration membrane surface gray level image is divided into left and right parts by each dividing line along the vertical direction, the dividing line is continuously translated to lead the ultrafiltration membrane surface gray level image to be finally divided into 30 areas with complete parts in the vertical direction, each area is a dividing area, and the same interval of each divided area in the horizontal direction is ensured, so that the subsequent accurate analysis of the hole distribution arrangement condition is facilitated.
Counting the number of the hole areas in each divided area as distribution number, and sequencing the distribution number according to the continuous adjacent distribution sequence of the divided areas to obtain a distribution sequence. Because the ultrafiltration membrane is circularly distributed, when the ultrafiltration membrane is in a normal state, the number change of regularly distributed hole areas along one direction is increased and then reduced, namely, the ultrafiltration membrane is in standard normal distribution, so that the arrangement regularity is integrally judged from the macroscopic aspect through the change condition of the distribution number in a distribution sequence, and the number of holes in each divided area is arranged according to the distribution condition through the distribution sequence, thereby being convenient for analyzing the normal distribution condition.
Calculating the absolute value of kurtosis of the distribution quantity in the distribution sequence, and reflecting whether the distribution of the distribution quantity accords with normal distribution or not by using the standard distribution index as a standard distribution index, wherein in the embodiment of the invention, the expression of the standard distribution index is as follows:
in the method, in the process of the invention,standard distribution index expressed as a distribution sequence, +.>Expressed as total number of distribution numbers in the distribution sequence, < >>Expressed as +.>Data value of number of distributions +.>Mean value corresponding to data value expressed as number of distributions in the distribution sequence, +. >Variance corresponding to data values expressed as number of distributions in a distribution sequence, +.>Represented as an absolute value extraction function. It should be noted that, kurtosis calculation is a published technical formula familiar to those skilled in the art, wherein 3 represents kurtosis of standard normal distribution, and the kurtosis is also directly reflected by a difference from 3 in general kurtosis calculation to represent the situation of normal distribution, so specific meaning of the formula is not repeated.
Calculating the absolute value of the skewness of the distribution quantity in the distribution sequence, and taking the absolute value as a deviation distribution index, wherein whether the distribution of the distribution quantity is deviated or not is reflected by the deviation distribution index, and in the embodiment of the invention, the expression of the skewness distribution index is as follows:
in the method, in the process of the invention,expressed as a deviation distribution index in the distribution sequence, < >>Expressed as total number of distribution numbers in the distribution sequence, < >>Expressed as +.>Data value of number of distributions +.>Mean value corresponding to data value expressed as number of distributions in the distribution sequence, +.>Variance corresponding to data values expressed as number of distributions in a distribution sequence, +.>Represented as an absolute value extraction function. It should be noted that, the calculation of the skewness is a published technical formula familiar to those skilled in the art, so the specific meaning of the formula is not repeated.
When the standard distribution index is smaller, the deviation distribution index is smaller, which indicates that the distribution condition of the distribution quantity in the distribution sequence is closer to the standard normal distribution trend, so that the texture rule index of the gray level image of the ultrafiltration membrane surface is obtained according to the standard distribution index and the deviation distribution index, and the standard distribution index and the deviation distribution index are in negative correlation with the texture rule index. In the embodiment of the invention, the expression of the texture rule index is:
in the method, in the process of the invention,texture rule index expressed as ultrafiltration membrane surface gray level image, < >>Standard distribution index expressed as a distribution sequence, +.>Expressed as a deviation distribution index in the distribution sequence, < >>Represented as an exponential function with a base of natural constant.
In the embodiment of the invention, the standard distribution index and the deviation distribution index are reflected in a form of negative exponent power to be in negative correlation with the texture rule index, and in other embodiments of the invention, other basic mathematical operations can be adopted to reflect that the standard distribution index and the deviation distribution index are in negative correlation with the texture rule index, such as subtraction or division, etc., without limitation.
So far, the analysis of the macroscopic angle is completed, and when the index of the texture rule is larger, the distribution condition of the holes accords with normal distribution, and the possibility that the ultrafiltration membrane is in a normal state on the macroscopic angle is higher.
S2: screening suspected abnormal areas from the hole areas according to the shape difference of each hole area and other hole areas in different preset directions; and under different preset directions, obtaining the edge anomaly degree of each suspected abnormal region under each preset direction according to the deviation degree of the distribution position between each suspected abnormal region and other hole regions.
When the liquid impact is applied to the purification link, abnormal conditions such as deformation or damage and the like of the texture part can be only in local detail parts, and when the gray contrast in the local area is less obvious, the segmented area is also inaccurate, so that the gray contrast is increased for the part of the abnormal area possibly having deformation and the like, and better segmentation is achieved.
Since the hole area is a regular rectangular area, possible abnormal conditions of the hole area can be reflected according to the change of the shape. In the embodiment of the invention, the abnormal shape condition of the holes in different directions and other hole areas is reflected by the distances between the edges in the horizontal and vertical directions, namely, the overall abnormal degree of each hole area in the corresponding different preset directions is obtained according to the shape difference of each hole area and other hole areas in the different preset directions, the preset directions are set as the horizontal direction and the vertical direction, and an implementer can adjust according to specific implementation conditions.
Preferably, for any one preset direction, each hole area is sequentially taken as a target area, the edges of the target area are obtained, and the shape is reflected by the distance between the edges.
According to the distance between the lower edges of the target area in the preset direction, the area distance of the target area in the preset direction is obtained, preferably, if two edge points exist on one straight line corresponding to the preset direction in the edge pixel point of the target area, the corresponding two edge points are used as an edge matching pair, and when any one direction is the horizontal direction, namely on the corresponding straight line in the horizontal direction, when two edge pixel points exist on the straight line in the edge pixel point of the target area, the two edge pixel points are indicated to be on the same horizontal line, and the two pixel points are used as an edge matching pair. All edge matching pairs corresponding to the target area are obtained on all straight lines corresponding to each preset direction, namely the target area corresponds to the edge pixel points in a one-to-one mode in the horizontal direction, so that the distance between the subsequent calculation areas is facilitated. It should be noted that, the edge matching calculation method in the vertical direction is consistent, and redundant description is not made here.
Further, the distance between two edge points in each edge matching pair is calculated and used as the direction distance of each edge matching pair, the direction distance is the distance between target areas calculated by the edge points in one-to-one correspondence, the average value of the direction distances of all the edge matching pairs in each preset direction is used as the area distance of the target area in the corresponding preset direction, and the distance between the target area in the corresponding preset direction is reflected through the area distance. The calculation of the distance is a well-known technique for those skilled in the art, and may be, for example, a euclidean distance or the like, and is not limited thereto.
After the area distance is obtained, according to the difference condition among the analysis shapes of the area distance, further, the variance of the area distance of all the hole areas in the preset direction is taken as a distance fluctuation value, the discreteness of the distance in the whole area in the preset direction is reflected through the distance fluctuation value, when the possibility of abnormal conditions is smaller, the distance fluctuation value is smaller, and the texture can slightly change due to the occurrence of the purification and filtering process, so that the distance fluctuation value is necessarily different from zero. Taking the average value of the area distances of all the hole areas in the preset direction as a standard distance, and reflecting the distance degree of the whole hole area in the preset direction through the standard distance.
And taking the difference between the area distance of the target area in the preset direction and the standard distance as a target fluctuation value of the target area in the preset direction, wherein the target fluctuation value reflects the difference degree between the distance of the target area and the standard distance. The ratio of the target fluctuation value and the distance fluctuation value of the target area in the preset direction is taken as the direction irregularity of the target area in the preset direction, the direction irregularity reflects the contribution degree of the difference of the target area in the corresponding preset direction to the total deviation, the sum of the direction irregularities of the target area in all preset directions is normalized, and the sum of the direction irregularities of the target area in all preset directions is taken as the total irregularity of the target area, and the total irregularity reflects the irregularity in all directions. In the embodiment of the invention, the expression of the overall anomaly is:
in the method, in the process of the invention,denoted as +.>Overall abnormality of the individual hole areas +.>Expressed as the total number of preset directions,denoted as +.>The hole area is at->Zone distance in a predetermined direction, +.>Indicated as being at->The number of hole areas in a predetermined direction, < >>Indicated as being at->The distance fluctuation value in the preset direction. />It should be noted that, normalization is a technical means well known to those skilled in the art, and the normalization function may be selected by linear normalization or standard normalization, and the specific normalization method is not limited herein.
Wherein,indicated as being at->A standard distance in a predetermined direction,denoted as +.>The hole area is at->The target fluctuation value in the preset direction,denoted as +.>The hole area is at->Directional anomalies in the preset directions. When the overall anomaly is larger, the deviation degree of the distance in the target area is larger, and the edge of the target area is dissimilar to the edges of other hole areas.
Therefore, when the overall anomaly degree is greater than a preset anomaly degree threshold value, the corresponding hole area is used as a suspected anomaly area. In the embodiment of the invention, the preset abnormality threshold is 0.6, and the implementer can adjust according to specific implementation conditions. The suspected abnormal region is a region with larger difference between the edge spacing condition in the region and other regions, and the shape of the suspected abnormal region is likely to be deformed.
Because the edges of the deformed areas may not be uniform, for example, because the length of the foreign matters in the horizontal direction is larger in the filtering, the hole areas are deformed only in the horizontal direction, and some foreign matters with larger whole area are extruded, so that the four edges of the hole areas are deformed, and the abnormal conditions in different directions in the suspected abnormal areas are analyzed.
When edge deformation occurs, the distribution between the suspected abnormal areas and other hole areas is changed, so that the edge abnormality degree of each suspected abnormal area in each preset direction is obtained according to the deviation degree of the distribution positions of each suspected abnormal area and other hole areas in different preset directions.
Preferably, in each preset direction, according to the distance distribution condition between each hole area and the adjacent hole area, the distribution index of the hole area in each preset direction is obtained, and the calculation of the distribution condition of the regular hole arrangement is performed through the characteristic of the regular hole arrangement on the ultrafiltration membrane surface.
In one embodiment of the invention, the distance between the center points corresponding to two adjacent hole areas is calculated in each preset direction to obtain the distribution distance of each hole area in the corresponding preset direction, the ratio of the area distance of each hole area in each preset direction to the distribution distance is normalized to obtain the distribution index of the hole area in each preset direction, the ratio of the hole area in the distribution distance is reflected by the distribution index, and the larger the distribution index is, the closer the hole area is to the adjacent hole area is.
Firstly, obtaining the distribution distance condition of normal holes in an ultrafiltration membrane according to distribution indexes, sequentially taking hole areas except suspected abnormal areas as reference areas, taking the average value of the distribution indexes of all the reference areas in each preset direction as the distribution influence index in each preset direction, wherein the distribution influence index is the distribution condition of the hole areas under the normal condition.
For any suspected abnormal region, acquiring a distribution index of the suspected abnormal region in each preset direction, calculating the difference between the distribution index and the distribution influence index of the suspected abnormal region in each preset direction, acquiring the edge anomaly degree of the suspected abnormal region in each preset direction, respectively analyzing according to the difference of the distribution conditions in each direction, reflecting the distribution deviation degree of the suspected abnormal region in the corresponding direction through the edge anomaly, and indicating that the anomaly of the suspected abnormal region in the lower edge of the corresponding direction is more obvious when the edge anomaly degree is larger.
S3: updating the gray values of the pixels according to the position distribution of each pixel in the suspected abnormal region, the edge anomaly degree under different preset directions and the texture rule index of the gray image of the ultrafiltration membrane surface to obtain a new gray image of the ultrafiltration membrane surface; and dividing the new ultrafiltration membrane surface gray level image to obtain a detection hole area, and detecting the state according to the detection hole area.
For the pixel points in the suspected abnormal region, when the pixel points are closer to the abnormal edge, the pixel points are blurred in gray level and are unfavorable for segmentation observation, so that the gray level contrast condition of the pixel points closer to the edge is improved, and when the whole abnormal condition in the ultrafiltration membrane is more serious, the contrast of the pixel points closer to the abnormal condition is required to be improved to a greater extent, so that the segmentation effect is enhanced, and the gray level value of the pixel points is updated according to the position distribution of each pixel point in the suspected abnormal region, the edge abnormal degree in different preset directions and the texture rule index of the gray level image of the surface of the ultrafiltration membrane to obtain a new gray level image of the surface of the ultrafiltration membrane.
Preferably, each suspected abnormal region is sequentially taken as an adjustment region, for any pixel point in the adjustment region, the distances between the pixel point and two edges of the adjustment region in each preset direction are calculated, and the pixel distribution distance of the pixel point in the corresponding preset direction is obtained, for example, for the horizontal direction, two distances exist between the pixel point in the horizontal direction and the edges of the adjustment region, and the two distances are the pixel distribution distance of the pixel point in the horizontal direction. The smallest pixel distribution distance in each preset direction is used as the position weight of the pixel point in the corresponding preset direction, and the closer the pixel point is to the edge position, the more the pixel point is affected by abnormal change, so that the nearest distance, namely the smallest pixel distribution distance, is used as the position weight of the pixel point to be analyzed.
Further, according to the position weight of the pixel point in each preset direction and the corresponding edge anomaly degree in the adjustment area, the local anomaly adjustment weight of the pixel point is obtained, and the local anomaly adjustment weight reflects the influence degree of the anomaly condition on each pixel point from the local detail angle. In one embodiment of the present invention, the product of the position weight and the edge anomaly degree of each pixel point in each preset direction is used as an anomaly adjustment index of each pixel point in each preset direction, and the accumulated value of the anomaly adjustment indexes of each pixel point in all preset directions is used as a local anomaly adjustment weight of each pixel point.
According to the local abnormal adjustment weight of the pixel point and the texture rule index of the gray level image on the surface of the ultrafiltration membrane, the gray level adjustment weight of the pixel point is obtained, the local abnormal adjustment weight is positively correlated with the gray level adjustment weight, the texture rule index is negatively correlated with the gray level adjustment weight, and the gray level adjustment weight is normalized. The gray scale adjustment weight is the influence degree of each pixel point under the abnormal condition under the comprehensive local detail and the macroscopic angle. In the embodiment of the invention, the expression of the gray scale adjustment weight is:
In the method, in the process of the invention,denoted as +.>Gray scale adjustment weight of each pixel point, < ->Texture rule index expressed as ultrafiltration membrane surface gray level image, < >>Expressed as total number of preset directions, +.>Denoted as +.>The pixel point is at the +.>Position weights in a predetermined direction, +.>Denoted as +.>The suspected abnormal area where the pixel point is located is in the +.>Edge anomalies in a predetermined direction, < ->Expressed as a normalization function>Expressed as a preset adjustment parameter, in the embodiment of the present invention, set to 0.001, the purpose is to prevent the case where the denominator is zero to make the formula meaningless.
Wherein,denoted as +.>The pixel point is at the +.>Abnormal adjustment indexes in a preset direction,denoted as +.>Local anomalies of individual pixels adjust weights. When the gray scale adjustment weight is larger, the pixel point is required to be adjusted to a larger degree. Further, the product of the gray adjustment weight and the corresponding gray value of each pixel point is taken as an adjustment value of each pixel point, the adjustment value is the gray value which needs to be adjusted, the sum value of the adjustment value and the gray value of each pixel point is taken as a new gray value of each pixel point, updating of the gray value is completed based on anomaly analysis, and then a new ultrafiltration membrane surface gray image is obtained based on the new gray value of each pixel point.
In the new ultrafiltration membrane surface gray level image, the pixel points close to the edge of the abnormal condition are optimized, so that the local gray level contrast is improved, and the segmentation result can be better and more accurate during segmentation. Therefore, the new ultrafiltration membrane surface gray level image is segmented to obtain detection hole areas, and state detection is carried out according to the detection hole areas.
Preferably, the new ultrafiltration membrane surface gray level image is segmented by an Ojin threshold segmentation method, so that a detection hole area is obtained, and after the local gray level contrast is improved, the segmentation effect of an abnormal position can be more accurate by the Ojin threshold segmentation method, so that the hole area for further detection is obtained.
In the embodiment of the present invention, the matching degree of each detection hole area is obtained by matching each detection hole area with a preset standard hole area by using a shape matching algorithm, and in the embodiment of the present invention, the matching degree can be obtained by selecting the square difference matching of a template matching algorithm, which is a technical means well known to those skilled in the art, and not described herein, in other embodiments of the present invention, other shape matching algorithms, such as Hu moment, may be selected, without limitation. When the matching degree is larger, the detection hole area is more similar to the preset standard hole area, and the abnormal possibility is smaller.
When the matching degree is smaller than the preset matching threshold, the corresponding hole area is indicated to be abnormal, the corresponding detection hole area is taken as an abnormal area, and in the embodiment of the invention, the preset matching threshold is 0.6, and an implementer can adjust according to specific implementation conditions.
When the total number of abnormal areas is larger than a preset abnormal threshold, the abnormal areas are more, the ultrafiltration membrane needs to be processed, and the health state of the corresponding ultrafiltration membrane is marked as poor. When the number of the abnormal areas is smaller than or equal to a preset abnormal threshold value, the ultrafiltration membrane can meet the purification effect, and the health state of the corresponding ultrafiltration membrane is recorded as good. In the embodiment of the present invention, the preset abnormal threshold is set to 3, and the specific numerical value implementation can be adjusted by itself, which is not limited herein.
In summary, according to the invention, through the hole areas which are firstly divided, according to the distribution arrangement condition of holes in the ultrafiltration membrane, the abnormal condition is analyzed from a macroscopic angle, a texture rule index is obtained, and the abnormal degree is analyzed as a whole. Further, analysis is performed from local details, because the ultrafiltration membrane is generally provided with regularly arranged hole areas, and the hole areas are regular rectangles, the areas with anomalies can be primarily screened out according to the degree of shape difference, namely, the suspected anomaly areas are obtained according to the shape differences in different preset directions, the degree of deviation of distribution positions between the suspected anomaly areas and other hole areas is utilized to obtain the edge anomaly degree in different preset directions from the position deviation, and the positions of the edges, which are more abnormal, can be used for adjusting nearby pixel points so as to enhance gray scale contrast and facilitate subsequent segmentation. The position distribution and the edge anomaly degree of the pixel points in the suspected anomaly region are utilized, the abnormal influence condition of each pixel point is obtained through local details, the gray value of the pixel point is comprehensively adjusted at a macroscopic angle by combining with a texture rule index, a new ultrafiltration membrane surface gray image is obtained, the gray values of the pixel points in the new ultrafiltration membrane surface gray image are optimized, the contrast of the local details is enhanced, the detection hole region obtained by dividing the new ultrafiltration membrane surface gray image is clearer and more accurate, and further the effect of state detection is better. According to the invention, the accuracy of region segmentation is improved and the state detection effect is enhanced by improving the gray contrast of local details.
It should be noted that: the sequence of the embodiments of the present invention is only for description, and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments.
Claims (10)
1. A visual detection method for the health state of an ultrafiltration membrane in a purification process, which is characterized by comprising the following steps:
acquiring a gray level image of the surface of the ultrafiltration membrane; dividing the gray level image on the surface of the ultrafiltration membrane to obtain a hole area; obtaining a texture rule index of the ultrafiltration membrane surface gray level image according to the distribution regularity of the hole area in the ultrafiltration membrane surface gray level image;
screening suspected abnormal areas from the hole areas according to the shape difference of each hole area and other hole areas in different preset directions; obtaining the edge anomaly degree of each suspected abnormal region in each preset direction according to the deviation degree of the distribution position between each suspected abnormal region and other hole regions in different preset directions;
Updating the gray values of the pixels according to the position distribution of each pixel in the suspected abnormal region, the edge anomaly degree under different preset directions and the texture rule index of the gray image of the ultrafiltration membrane surface to obtain a new gray image of the ultrafiltration membrane surface; and dividing the new ultrafiltration membrane surface gray level image to obtain a detection hole area, and detecting the state according to the detection hole area.
2. The visual inspection method for health status of ultrafiltration membrane in purification process according to claim 1, wherein said method for obtaining texture rule index comprises:
dividing the gray level image of the ultrafiltration membrane surface into a preset dividing number of dividing areas along a preset dividing direction; all the dividing areas have the same interval in the vertical direction of the preset dividing direction;
counting the number of the hole areas in each divided area as distribution number, and sequencing the distribution number according to the continuous adjacent distribution sequence of the divided areas to obtain a distribution sequence;
calculating the absolute value of kurtosis of the distribution quantity in the distribution sequence as a standard distribution index; calculating the absolute value of the deviation of the distribution quantity in the distribution sequence as a deviation distribution index; and obtaining the texture rule index of the ultrafiltration membrane surface gray level image according to the standard distribution index and the deviation distribution index, wherein the standard distribution index and the deviation distribution index are in negative correlation with the texture rule index.
3. The visual inspection method for health status of ultrafiltration membrane in purification process according to claim 1, wherein said method for obtaining suspected abnormal region comprises:
for any one preset direction, taking each hole area as a target area in sequence, and acquiring the edge of the target area; obtaining the region distance of the target region in the preset direction according to the distance between the lower edges of the target region in the preset direction;
taking the variance of the area distance of all the hole areas in the preset direction as a distance fluctuation value; taking the average value of the area distances of all the hole areas in the preset direction as a standard distance;
taking the difference between the area distance of the target area in the preset direction and the standard distance as a target fluctuation value of the target area in the preset direction; taking the ratio of the target fluctuation value of the target area in the preset direction to the distance fluctuation value as the direction anomaly of the target area in the preset direction; normalizing the sum of the directional anomalies of the target area in all preset directions to obtain the overall anomaly of the target area;
when the overall anomaly degree is greater than a preset anomaly degree threshold value, the corresponding hole area is used as a suspected anomaly area.
4. A visual inspection method for health status of ultrafiltration membrane in purification process according to claim 3, wherein said obtaining method for regional distance comprises:
if two edge points exist on a straight line corresponding to the preset direction in the edge pixel point of the target area, the corresponding two edge points are used as an edge matching pair; all edge matching pairs corresponding to the target area are obtained on all straight lines corresponding to each preset direction;
calculating the distance between two edge points in each edge matching pair as the direction distance of each edge matching pair; and taking the average value of the direction distances of all the edge matching pairs in each preset direction as the area distance of the target area in the corresponding preset direction.
5. A visual inspection method for health status of ultrafiltration membrane in purification process according to claim 3, wherein said method for obtaining edge anomaly comprises:
obtaining a distribution index of the hole areas in each preset direction according to the distance distribution condition between each hole area and the adjacent hole areas in each preset direction; sequentially taking the hole areas except the suspected abnormal areas as reference areas, and taking the average value of the distribution indexes of all the reference areas in each preset direction as the distribution influence index in each preset direction;
For any suspected abnormal region, acquiring a distribution index of the suspected abnormal region in each preset direction; and calculating the difference between the distribution index and the distribution influence index of the suspected abnormal region in each preset direction, and obtaining the edge anomaly degree of the suspected abnormal region in each preset direction.
6. The visual inspection method for health status of ultrafiltration membrane in purification process according to claim 5, wherein said obtaining method for distribution index comprises:
calculating the distance between the center points corresponding to two adjacent hole areas in each preset direction to obtain the distribution distance of each hole area in the corresponding preset direction; and carrying out normalization processing on the ratio of the area distance to the distribution distance of each hole area in each preset direction to obtain the distribution index of the hole area in each preset direction.
7. The visual inspection method of the health status of an ultrafiltration membrane in the purification process according to claim 1, wherein updating the gray value of each pixel to obtain a new ultrafiltration membrane surface gray image according to the position distribution of each pixel in the suspected abnormal region, the edge anomaly degree in different preset directions, and the texture rule index of the ultrafiltration membrane surface gray image comprises the following steps:
Sequentially taking each suspected abnormal region as an adjustment region, and calculating the distance between each pixel point and two edges of the adjustment region in each preset direction for any pixel point in the adjustment region to obtain the pixel distribution distance of the pixel point in the corresponding preset direction; taking the minimum pixel distribution interval in each preset direction as the position weight of the pixel point in the corresponding preset direction;
obtaining local abnormal adjustment weights of the pixel points according to the position weights of the pixel points in each preset direction and the corresponding edge abnormal degrees in the adjustment areas; according to the local abnormal adjustment weight of the pixel point and the texture rule index of the ultrafiltration membrane surface gray level image, obtaining the gray level adjustment weight of the pixel point; the local abnormal adjustment weight and the gray adjustment weight are positively correlated, the texture rule index and the gray adjustment weight are negatively correlated, and the gray adjustment weight is a normalized value;
taking the product of the gray scale adjustment weight of each pixel point and the corresponding gray scale value as the adjustment value of each pixel point; taking the sum value of the adjustment value and the gray value of each pixel point as a new gray value of each pixel point; and obtaining a new ultrafiltration membrane surface gray scale image based on the new gray scale value of each pixel point.
8. The visual inspection method for health status of ultrafiltration membrane in purification process according to claim 7, wherein said obtaining method for local abnormality adjustment weight comprises:
taking the product of the position weight and the edge anomaly degree of each pixel point in each preset direction as an anomaly adjustment index of each pixel point in each preset direction;
and taking the accumulated value of the abnormal adjustment indexes of each pixel point in all preset directions as the local abnormal adjustment weight of each pixel point.
9. The visual inspection method for health status of ultrafiltration membrane in purification process according to claim 1, wherein said segmenting new ultrafiltration membrane surface gray scale image to obtain inspection hole area, and performing status inspection according to inspection hole area comprises:
dividing a new ultrafiltration membrane surface gray level image by an Ojin threshold segmentation method to obtain a detection hole area;
matching each detection hole area with a preset standard hole area by adopting a shape matching algorithm to obtain the matching degree of each detection hole area; when the matching degree is smaller than a preset matching threshold value, the corresponding detection hole area is used as an abnormal area;
When the total number of the abnormal areas is larger than a preset abnormal threshold value, marking the health state of the corresponding ultrafiltration membrane as poor; and when the number of the abnormal areas is smaller than or equal to a preset abnormal threshold value, the health state of the corresponding ultrafiltration membrane is recorded as good.
10. The visual inspection method for health status of ultrafiltration membrane in purification process according to claim 1, wherein said dividing the gray scale image of the ultrafiltration membrane surface to obtain the hole area comprises:
dividing the gray level image on the surface of the ultrafiltration membrane by using an Ojin threshold segmentation method, and taking the connected domain with the lowest gray level value as a hole connected domain;
and taking a region formed by the corresponding range of the hole communicating region in the ultrafiltration membrane surface gray level image as a hole region of the ultrafiltration membrane surface gray level image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410065311.XA CN117576104A (en) | 2024-01-17 | 2024-01-17 | Visual detection method for health state of ultrafiltration membrane in purification process |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410065311.XA CN117576104A (en) | 2024-01-17 | 2024-01-17 | Visual detection method for health state of ultrafiltration membrane in purification process |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117576104A true CN117576104A (en) | 2024-02-20 |
Family
ID=89886791
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410065311.XA Pending CN117576104A (en) | 2024-01-17 | 2024-01-17 | Visual detection method for health state of ultrafiltration membrane in purification process |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117576104A (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20020081025A1 (en) * | 1998-12-18 | 2002-06-27 | Wagman Adam H. | System and method for in-line inspection of stencil aperture blockage |
| WO2020038124A1 (en) * | 2018-08-21 | 2020-02-27 | 深圳创维－Rgb电子有限公司 | Image contrast enhancement method and apparatus, and device and storage medium |
| WO2021000524A1 (en) * | 2019-07-03 | 2021-01-07 | 研祥智能科技股份有限公司 | Hole protection cap detection method and apparatus, computer device and storage medium |
| CN116071387A (en) * | 2023-03-06 | 2023-05-05 | 曲阜市巨力铁路轨道工程股份有限公司 | Sleeper rail production quality detection method based on machine vision |
| CN116758064A (en) * | 2023-08-14 | 2023-09-15 | 深圳天眼新能源科技有限公司 | Lithium battery diaphragm quality detection method based on electron scanning microscope |
| CN116823822A (en) * | 2023-08-29 | 2023-09-29 | 山东航宇游艇发展有限公司 | Ship sheet metal part welding defect detection method based on visual characteristics |
| CN116843688A (en) * | 2023-09-01 | 2023-10-03 | 山东虹纬纺织有限公司 | Visual detection method for quality of textile |
| CN116934749A (en) * | 2023-09-15 | 2023-10-24 | 山东虹纬纺织有限公司 | Textile flaw rapid detection method based on image characteristics |
-
2024
- 2024-01-17 CN CN202410065311.XA patent/CN117576104A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20020081025A1 (en) * | 1998-12-18 | 2002-06-27 | Wagman Adam H. | System and method for in-line inspection of stencil aperture blockage |
| WO2020038124A1 (en) * | 2018-08-21 | 2020-02-27 | 深圳创维－Rgb电子有限公司 | Image contrast enhancement method and apparatus, and device and storage medium |
| WO2021000524A1 (en) * | 2019-07-03 | 2021-01-07 | 研祥智能科技股份有限公司 | Hole protection cap detection method and apparatus, computer device and storage medium |
| CN116071387A (en) * | 2023-03-06 | 2023-05-05 | 曲阜市巨力铁路轨道工程股份有限公司 | Sleeper rail production quality detection method based on machine vision |
| CN116758064A (en) * | 2023-08-14 | 2023-09-15 | 深圳天眼新能源科技有限公司 | Lithium battery diaphragm quality detection method based on electron scanning microscope |
| CN116823822A (en) * | 2023-08-29 | 2023-09-29 | 山东航宇游艇发展有限公司 | Ship sheet metal part welding defect detection method based on visual characteristics |
| CN116843688A (en) * | 2023-09-01 | 2023-10-03 | 山东虹纬纺织有限公司 | Visual detection method for quality of textile |
| CN116934749A (en) * | 2023-09-15 | 2023-10-24 | 山东虹纬纺织有限公司 | Textile flaw rapid detection method based on image characteristics |
Non-Patent Citations (3)
| Title |
|---|
| 兰红;方治屿;: "3维灰度矩阵的钢板缺陷图像识别", 中国图象图形学报, no. 06, 16 June 2019 (2019-06-16) * |
| 曾毅;郭龙源;罗百通;: "基于BP神经网络和均值差分的TFT-LCD MURA缺陷检测方法", 湖南理工学院学报(自然科学版), no. 01, 15 March 2017 (2017-03-15) * |
| 郭忠峰;杨子豪;王赫莹;: "结合最小均方差滤波和目标灰度的边缘检测", 机械工程与自动化, no. 04, 15 August 2020 (2020-08-15) * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115829883B (en) | Surface image denoising method for special-shaped metal structural member | |
| CN107255641B (en) | A method of Machine Vision Detection is carried out for self-focusing lens surface defect | |
| CN105067638B (en) | Tire fetal membrane face character defect inspection method based on machine vision | |
| WO2022110804A1 (en) | Image noise measurement method and device based on local statistical information | |
| CN116030061B (en) | Silica gel molding effect detection method based on vision | |
| CN116611748B (en) | Titanium alloy furniture production quality monitoring system | |
| CN115018844B (en) | Plastic film quality evaluation method based on artificial intelligence | |
| CN115131348B (en) | Method and system for detecting textile surface defects | |
| CN112857471B (en) | Industrial Internet of things chemical wastewater treatment emission on-line monitoring and early warning management cloud platform | |
| CN114820625B (en) | Automobile top block defect detection method | |
| CN116416252B (en) | Method for detecting sedimentation image of wastewater in boehmite production process | |
| CN115115642A (en) | Strip steel scab defect detection method based on image processing | |
| CN115861307B (en) | Fascia gun power supply driving plate welding fault detection method based on artificial intelligence | |
| CN116630813B (en) | Highway road surface construction quality intelligent detection system | |
| CN114881960A (en) | Feature enhancement-based cloth linear defect detection method and system | |
| CN115049671A (en) | Cloth surface defect detection method and system based on computer vision | |
| CN116309649A (en) | Throat lesion abnormal region segmentation system based on CT image | |
| CN116137036A (en) | Gene detection data intelligent processing system based on machine learning | |
| CN114817228A (en) | Abnormal value removing method for long-time sequence monitoring data of lake water level | |
| CN113643276A (en) | Textile texture defect automatic detection method based on statistical analysis | |
| CN116309599A (en) | Water quality visual monitoring method based on sewage pretreatment | |
| CN116524205A (en) | Sewage aeration automatic detection and identification method | |
| CN114663393A (en) | Convolutional neural network-based detection method for defect of poor spacing between photovoltaic modules | |
| CN117291919B (en) | Visual detection method for lubricating grease on surface of steel rail bolt | |
| CN116823125B (en) | White spirit storage abnormality early warning system based on multisource data |
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 |