CN116297530A - Barrier film surface quality detection method based on optical technology - Google Patents
Barrier film surface quality detection method based on optical technology Download PDFInfo
- Publication number
- CN116297530A CN116297530A CN202310564883.8A CN202310564883A CN116297530A CN 116297530 A CN116297530 A CN 116297530A CN 202310564883 A CN202310564883 A CN 202310564883A CN 116297530 A CN116297530 A CN 116297530A
- Authority
- CN
- China
- Prior art keywords
- wave band
- band
- point
- adjustment
- characteristic
- 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
- 230000004888 barrier function Effects 0.000 title claims abstract description 89
- 238000001514 detection method Methods 0.000 title claims abstract description 23
- 238000005516 engineering process Methods 0.000 title claims abstract description 22
- 230000003287 optical effect Effects 0.000 title claims abstract description 20
- 230000003595 spectral effect Effects 0.000 claims abstract description 62
- 238000002310 reflectometry Methods 0.000 claims abstract description 40
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 13
- 238000000034 method Methods 0.000 claims description 38
- 238000012937 correction Methods 0.000 claims description 21
- 238000004364 calculation method Methods 0.000 claims description 15
- 238000009826 distribution Methods 0.000 claims description 12
- 238000012360 testing method Methods 0.000 abstract description 2
- 230000002159 abnormal effect Effects 0.000 description 5
- 238000001228 spectrum Methods 0.000 description 5
- 230000008859 change Effects 0.000 description 3
- 230000007547 defect Effects 0.000 description 3
- 238000010606 normalization Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000000701 chemical imaging Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 description 1
- 238000013329 compounding Methods 0.000 description 1
- 238000006073 displacement reaction Methods 0.000 description 1
- 229940079593 drug Drugs 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 235000013305 food Nutrition 0.000 description 1
- 229910052736 halogen Inorganic materials 0.000 description 1
- 150000002367 halogens Chemical class 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 229910052760 oxygen Inorganic materials 0.000 description 1
- 239000001301 oxygen Substances 0.000 description 1
- 239000005022 packaging material Substances 0.000 description 1
- 238000004806 packaging method and process Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Chemical compound O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
- 230000005428 wave function Effects 0.000 description 1
- 230000037303 wrinkles Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/95—Investigating the presence of flaws or contamination characterised by the material or shape of the object to be examined
-
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/42—Global feature extraction by analysis of the whole pattern, e.g. using frequency domain transformations or autocorrelation
- G06V10/422—Global feature extraction by analysis of the whole pattern, e.g. using frequency domain transformations or autocorrelation for representing the structure of the pattern or shape of an object therefor
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Analytical Chemistry (AREA)
- General Health & Medical Sciences (AREA)
- Biochemistry (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Chemical & Material Sciences (AREA)
- Quality & Reliability (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention relates to the technical field of optical means testing, and provides a barrier film surface quality detection method based on an optical technology, which comprises the following steps: collecting hyperspectral data of the surface of the barrier film; acquiring a plurality of characteristic wave band ranges of each position according to the spectral reflectivity of each wave band of each position on the surface of the barrier film, acquiring the classification degree of any two adjacent positions according to the characteristic wave band ranges of each position, and acquiring a plurality of areas on the surface of the barrier film and the data fluctuation degree of each area; acquiring initial adjustment of each wave band of each position according to the spectral reflectivity and the data fluctuation degree, correcting the initial adjustment according to the characteristic wave band range to obtain corrected adjustment of each wave band of each position, and acquiring adjusted hyperspectral data; and (3) denoising through EMD decomposition and wavelet threshold value to obtain denoised hyperspectral data, and finishing the surface quality detection of the barrier film. The invention aims to solve the problem that the hyperspectral data is affected by noise to cause inaccurate detection results.
Description
Technical Field
The invention relates to the technical field of optical means testing, in particular to a barrier film surface quality detection method based on an optical technology.
Background
The barrier film is a film material which is formed by compounding multiple layers of materials and can effectively isolate external substances such as oxygen, water vapor and the like from entering the package; the packaging material has good barrier property, is widely applied to packaging in various fields of foods, medicines and the like, is beneficial to protecting the quality of products and prolonging the shelf life of the products; the quality of the surface of the barrier film influences the barrier performance, so that the quality of the surface of the barrier film needs to be accurately detected; most of traditional barrier film surface quality detection methods are visual detection or microscopic observation, and detection efficiency is low.
In the existing method, a hyperspectral technology is adopted to acquire a plurality of wave band data on the surface of the barrier film, so that the data can be processed and analyzed under the condition of not contacting a detection object, further, the high-precision quality detection on the surface of the barrier film is realized, and the influence of human errors is avoided; however, hyperspectral data are easily affected by noise, so that the surface quality detection result of the barrier film is inaccurate; the traditional hyperspectral data denoising method comprises the steps of obtaining a plurality of IMF components by adopting an EMD (empirical mode decomposition) algorithm, denoising a wavelet threshold value of each IMF component, and reconstructing to obtain denoised hyperspectral data; however, since the plurality of IMF component numbers acquired by the EMD decomposition algorithm have modal aliasing, the threshold acquired by the wavelet threshold denoising algorithm cannot distinguish the detail information and the noise information, so that partial information is lost in the denoising result, and information adjustment is required to be carried out on hyperspectral data, so that the plurality of IMF components acquired by the EMD decomposition can accurately acquire the threshold to distinguish the detail information and the noise information, and further the wavelet threshold denoising is completed; and obtaining accurate detection results of the surface quality of the barrier film by obtaining the denoised hyperspectral data and comparing the hyperspectral data with ideal hyperspectral data.
Disclosure of Invention
The invention provides a barrier film surface quality detection method based on an optical technology, which aims to solve the problem that the detection result is inaccurate due to the influence of noise on the existing hyperspectral data, and adopts the following technical scheme:
one embodiment of the invention provides a barrier film surface quality detection method based on optical technology, which comprises the following steps:
collecting hyperspectral data of each position on the surface of the barrier film;
acquiring a plurality of wave band points and wave band curves of each position according to hyperspectral data of each position, acquiring a plurality of characteristic points of each position according to the wave band curves, acquiring a plurality of wave points and a plurality of characteristic wave band ranges of each position according to the distribution of the characteristic points in the wave band curves, acquiring classification degrees of any two adjacent positions according to the characteristic wave band ranges of each position, and acquiring a plurality of areas on the surface of the barrier film and the data fluctuation degree of each area according to the classification degrees;
obtaining reference weight of a local range of each position according to the classification degree and the data fluctuation degree, obtaining initial adjustment of each wave band of each position according to the reference weight and the hyperspectral data, obtaining correction adjustment of each wave band of each position according to the wave band curve and the fluctuation point, obtaining all adjustment wave bands of each position according to the correction adjustment of each wave band of each position, obtaining a plurality of categories according to correction adjustment clustering of all adjustment wave bands, and obtaining adjustment values of each adjustment wave band according to the adjustment wave bands, the hyperspectral data and the fluctuation points in each category, and obtaining adjusted hyperspectral data;
and carrying out EMD decomposition and wavelet threshold denoising according to the adjusted hyperspectral data to obtain denoised hyperspectral data, and finishing the surface quality detection of the barrier film.
Optionally, the acquiring a plurality of band points and band curves of each position includes the following specific methods:
taking any one position on the surface of the barrier film as a target position, acquiring the spectral reflectivity of the target position in each wave band in hyperspectral data, taking the abscissa as the wave band, taking the ordinate as the spectral reflectivity, constructing a coordinate system, placing each wave band of the target position and the spectral reflectivity thereof in the coordinate system to obtain a plurality of wave band points, and connecting adjacent wave band points of the abscissa to obtain a wave band curve of the target position;
and acquiring a plurality of wave band points and wave band curves of each position.
Optionally, the method for obtaining a plurality of feature points at each position according to the band curve includes the following specific steps:
taking any one position of the surface of the barrier film as a target position, for two adjacent wave band points in a wave band curve of the target position, marking the ratio between the difference value obtained by subtracting the spectral reflectivity of the former wave band point from the spectral reflectivity of the latter wave band point and the absolute value of the difference value of the abscissa of the two wave band points as the wave band slope of the former wave band point, and obtaining the wave band slope of each wave band point of the target position;
taking any one of the wave band points as a target wave band point, if sign changes occur on the wave band slope of the target wave band point and the next adjacent wave band point, marking the target wave band point as a characteristic point, and acquiring all characteristic points in a wave band curve of a target position, wherein 2 times of the wave band distance between each characteristic point and the adjacent previous characteristic point is taken as the wave band width of each characteristic point, and the wave band distance represents the absolute value of the difference value of the abscissa between the two wave band points;
and acquiring a plurality of characteristic points of each position and the wave band width of each characteristic point.
Optionally, the method for acquiring the plurality of fluctuation points and the plurality of characteristic wave band ranges of each position includes the following specific steps:
taking any position on the surface of the barrier film as a target position, and the first position in a wave band curve of the target position
Degree of characteristic of individual characteristic points->
The calculation method of (1) is as follows:
wherein,,
indicate->
Band width of individual feature points, +.>
Band width maximum value in band curve representing target position,/->
Indicate->
Each band point in the band range of the feature points is associated with +.>
The absolute value of the difference between the spectral reflectances of the characteristic points is the smallest, the band range is expressed as +.>
The abscissa of each characteristic point is taken as the center, the wave band width is a window constructed by the size, the abscissa of the wave band point is in the window, and the wave band point is in the wave band range;
the characteristic degree of each characteristic point in the wave band curve of the target position is obtained, all the characteristic degrees are normalized, the characteristic points with the normalized value of the characteristic degree larger than a preset first threshold value are marked as fluctuation points, and the wave band range of the fluctuation points is marked as the characteristic wave band range; and acquiring a plurality of fluctuation points and a plurality of characteristic wave band ranges of each position.
Optionally, the method for obtaining the classification degree of any two adjacent positions according to the characteristic band range of each position includes the following specific steps:
taking any position on the surface of the barrier film as a target position, extracting wave band points included in all characteristic wave band ranges of the target position, arranging and connecting all wave band points according to the sequence from small wave band to large wave band, and recording the obtained result as a characteristic curve of the target position;
acquiring a characteristic curve of each position; marking any one position and any one position in eight adjacent domains as adjacent positions, and the first position
Individual position +.>
Classification degree of->
The calculation method of (1) is as follows:
wherein,,
indicate->
Characteristic curve of individual positions and adjacent positions +.>
Is set to be the DTW distance of the characteristic curve of (c),
indicate->
Individual position +.>
European distance,/, of->
An exponential function based on a natural constant;
and obtaining the classification degree of any two adjacent positions.
Optionally, the method for obtaining the plurality of areas on the surface of the barrier film and the data fluctuation degree of each area according to the classification degree includes the following specific steps:
if the classification degree of the two adjacent positions is greater than a preset second threshold value, the two positions belong to the same area; if the classification degree of the two adjacent positions is smaller than or equal to a preset second threshold value, the two positions belong to different areas; dividing the areas of all adjacent positions according to the classification degree and a preset second threshold value, and dividing the surface of the barrier film into a plurality of areas;
taking any one area as a target area, acquiring variances of all classification degrees in the target area, marking the variances as the classification variances of the target area, acquiring the classification variances of each area, normalizing all the classification variances, and marking the obtained result as the data fluctuation degree of each area.
Optionally, the method for obtaining the reference weight of the local range of each position according to the classification degree and the data fluctuation degree includes the following specific steps:
in the first place
The individual positions are the center and the local range is +.>
Rectangular window of size, wherein +.>
And->
Respectively represent +.>
Maximum value of range in row direction and maximum value of range in column direction in the region to which the position belongs, +.>
Indicating rounding-off the whole symbol, +.>
Is a super parameter;
acquisition of the first
The variance of all degree of classification in the local range of the individual positions is denoted as +.>
A first local variance of the individual locations; get->
Partial range of individual positions except for the relevant +.>
The variance of all other classification levels except the classification level of the individual position, the result obtained is recorded as +.>
A second local variance of the individual locations; first->
Reference weight of local range of individual positions +.>
The calculation method of (1) is as follows:
wherein,,
indicate->
First local variance of the individual positions, +.>
Indicate->
Second local variance of the individual positions, +.>
Indicate->
The degree of fluctuation of the data of the area to which the location belongs, +.>
Representing a maximum function>
Representing absolute values.
Optionally, the acquiring the initial adjustability of each band of each position includes the following specific methods:
wherein,,
indicate->
No. 5 of the individual positions>
Initial adjustability of the individual bands,/->
Indicate->
Position->
Spectral reflectance of individual bands, +.>
Indicate->
The region of the individual band is at +.>
Spectral reflectance mean value of individual bands, +.>
Indicate->
The region of the individual band is at +.>
Spectral reflectance maximum of the individual bands, +.>
Indicate->
The area of the individual wave band is in the first
Spectral reflectance minimum of the individual bands, < >>
Indicate->
Reference weights for local ranges of individual locations.
Optionally, the correcting adjustment of each band at each position is obtained by correcting the band curve and the fluctuation point, which comprises the following specific steps:
wherein,,
indicate->
Position->
Correction adjustability of individual bands,>
indicate->
Position->
Spectral reflectance of individual bands, +.>
Indicate->
Position->
Spectral reflectivity of the wavebands corresponding to the wavebands closest to the waveband point,
indicate->
Position->
The abscissa of the corresponding band point of the individual band, +.>
Indicate->
Position->
The abscissa of the nearest wave point to the corresponding band point of the individual band, +.>
Indicate->
No. 5 of the individual positions>
Initial adjustability of the individual bands,/->
Represents an exponential function based on natural constants, < ->
Representing absolute values.
Optionally, the method for obtaining the adjustment value of each adjustment band includes:
taking any one of the categories as a target, acquiring the average value of the absolute value of the difference value between each corresponding band point of the adjustment band in the target category and the adjacent next band point on the spectral reflectivity, and taking the obtained result as an adjustment reference value of the target category; acquiring an adjustment reference value of each category;
any one adjustment wave band in the target class is a target adjustment wave band, the absolute value of the difference value between the wave band point corresponding to the target adjustment wave band and the nearest fluctuation point in the spectral reflectivity is obtained, the absolute value of the difference value between the absolute value of the difference value and the adjustment reference value is obtained, and the obtained absolute value of the difference value is used as the adjustment size of the target adjustment wave band;
if the fluctuation point of the target adjustment wave band, which corresponds to the wave band point and is closest to the wave band point, is the wave crest point, the sum of the spectral reflectivity of the target adjustment wave band and the adjustment size is the adjustment value of the target adjustment wave band; if the fluctuation point of the target adjustment wave band, which corresponds to the wave band point and is closest to the wave band point, is the wave trough point, the difference of the spectral reflectivity of the target adjustment wave band minus the adjustment size is the adjustment value of the target adjustment wave band; and acquiring an adjustment value of each adjustment band.
The beneficial effects of the invention are as follows: according to the invention, the hyperspectral technology is adopted to detect the surface quality of the barrier film, and the area division of the surface of the barrier film is carried out on each position according to the area distribution relation of the hyperspectral data through the collected hyperspectral data of the surface of the barrier film; according to the distribution characteristics of the hyperspectral data in each region and the influence of the wave bands of the hyperspectral data on the characteristic structural characteristics of the characteristic wave band range of the hyperspectral data, the correction adjustability of each wave band of each position is obtained, and further adjustment is carried out on the premise of not changing the distribution characteristics of the hyperspectral data, so that a plurality of intrinsic mode functions obtained by EMD decomposition can accurately obtain a threshold value to distinguish detail information and noise information for accurate wavelet threshold denoising; the defect that part of information is lost in a denoising result caused by incapability of distinguishing detail information and noise information by a threshold value acquired in a wavelet threshold denoising algorithm due to modal aliasing of a plurality of eigen mode functions acquired by an EMD (empirical mode decomposition) algorithm is avoided, accurate denoising hyperspectral data is acquired, and the hyperspectral data is compared with ideal hyperspectral data to obtain an accurate detection result of the surface quality of the barrier film.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions of the prior art, the drawings which are used in the description of the embodiments or the prior art will be briefly described, it being obvious that the drawings in the description below are only some embodiments of the invention, and that other drawings can be obtained according to these drawings without inventive faculty for a person skilled in the art.
Fig. 1 is a schematic flow chart of a method for detecting surface quality of a barrier film based on an optical technology according to an embodiment of the invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Referring to fig. 1, a flowchart of a method for detecting surface quality of a barrier film based on optical technology according to an embodiment of the invention is shown, and the method includes the following steps:
and S001, collecting hyperspectral data of the surface of the barrier film.
The purpose of this embodiment is to perform quality detection on the surface of the barrier film by hyperspectral technology, so that hyperspectral data of the surface of the barrier film needs to be collected first; in order to prevent wrinkles, the barrier film is kept in a stretched state through a carrier roller on a production line, a black flannelette background is arranged below the barrier film, a hyperspectral imaging spectrometer is arranged above the barrier film, and hyperspectral data of the surface of the barrier film are collected; the equipment used for collecting the hyperspectral data of the surface of the barrier film is as follows: the specific model of the device, such as a hyperspectral imaging spectrometer (the wave band range is 400nm-100nm, the wavelength interval is about 2.8 nm), a halogen light source of 21V/200W, an electric control displacement cradle head, a computer and the like, is not set, and an implementer depends on actual conditions; it should be noted that the hyperspectral data is the spectral reflectivities of different positions of the surface of the barrier film in different wavebands, that is, each position of the surface of the barrier film has a corresponding spectral reflectivity in each waveband.
Thus, hyperspectral data of the surface of the barrier film are obtained.
Step S002, according to the spectral reflectivity of each wave band at each position of the surface of the barrier film, a plurality of characteristic wave band ranges of each position are obtained, the classification degree of any two adjacent positions is obtained according to the characteristic wave band ranges of each position, and a plurality of areas of the surface of the barrier film and the data fluctuation degree of each area are obtained according to the classification degree.
It should be noted that, the hyperspectral data may cause a larger difference in the spectral data of the surface of the barrier film due to the influence of the light source and the possible defect influence of the surface of the barrier film, so that the data fluctuation degrees of the different light source influence areas and the defect areas of the surface of the barrier film are different; the characteristic wave band range can be quantified according to the spectral reflectivity of each position in different wave bands, the surface of the barrier film is divided into a plurality of areas by combining the distribution relation of different positions of the surface of the barrier film, and the data fluctuation degree of each area is quantified; while the characteristic band range usually shows a peak or trough form, various pseudo peaks appear in hyperspectral data due to the influence of noise, so that calculation of the characteristic degree is required for all the peaks and troughs.
Specifically, taking any position on the surface of the barrier film as an example, acquiring the spectral reflectivity of the position in each wave band in hyperspectral data, taking an abscissa as the wave band, taking an ordinate as the spectral reflectivity to construct a coordinate system, placing each wave band of the position and the spectral reflectivity thereof in the coordinate system to obtain a plurality of wave band points, and connecting adjacent wave band points of the abscissa to obtain a wave band curve of the position; for two adjacent band points, the ratio between the difference value obtained by subtracting the spectral reflectance of the former band point from the spectral reflectance of the latter band point and the absolute value of the difference value of the abscissa of the two band points is recorded as the band slope of the former band point, and the band slope of each band point at the position is obtained according to the method; it should be noted that, the last band point, i.e., the point with the largest band value, does not need to calculate the band slope.
Further, for the wave band curve, the wave band slope of any one wave band point is a positive value, if the sign change of the wave band slope does not appear at the next wave band point, the wave band continues to extend until the sign change stops, and the stopped wave band point is marked as a wave crest point; if the slope of the next wave band point is a negative value, the wave peak characteristic is presented, and the wave band point is a wave peak point; conversely, if the wave band slope of any one wave band point is a negative value and the wave band slope of the next wave band point is a positive value, the wave trough characteristic is presented, and the wave band point is a wave trough point; according to the method, all wave peak points and wave trough points in the wave band curve are obtained, and for any wave peak point, 2 times of the wave band distance between the wave peak point and the previous adjacent wave trough point is used as the wave band width of the wave peak point, and the wave band distance is the absolute value of the difference value of the abscissa between two wave band points; for any trough point, taking 2 times of the wave band distance between the trough point and the previous adjacent wave peak point as the wave band width of the trough point; marking peak points and trough points as characteristic points, wherein the wave band width of a first characteristic point in a wave band curve is set as the wave band width of a second characteristic point; and acquiring the wave band width of each characteristic point in the wave band curve of the position according to the method.
Further, the first of the band curves at that position
For example, the characteristic points are characterized by the degree +.>
The calculation method of (1) is as follows:
wherein,,
indicate->
Band width of individual feature points, +.>
Maximum value of band width in band curve representing the position,/->
Indicate->
Each band point in the band range of the feature points is associated with +.>
The absolute value of the difference between the spectral reflectances of the characteristic points is the smallest, wherein the band range is expressed in +.>
The abscissa of each characteristic point is taken as the center, the wave band width is a window constructed by the size, the abscissa of the wave band point is in the window, and then the wave band point is in the wave band range; the larger the band width is, the smaller the probability that the band width is a false peak formed by noise influence is, the larger the probability that the band width is real data is, and the larger the characteristic degree is; the larger the difference of the spectral reflectivity, the more the feature points contain information, and the greater the feature degree; according to the method, the characteristic degree of each characteristic point in the wave band curve of the position is obtained, linear normalization is carried out on all the characteristic degrees, a preset first threshold value is given for judging the characteristic wave band range, the preset first threshold value of the embodiment is calculated by adopting 0.58, the characteristic points with the normalized value of the characteristic degree larger than the preset first threshold value are marked as fluctuation points, the wave band range of the fluctuation points is marked as the characteristic wave band range, and then a plurality of characteristic wave band ranges of the position are obtained; and acquiring a plurality of characteristic wave band ranges of each position on the surface of the barrier film according to the method.
Further, for any position, extracting the band points included in all the characteristic band ranges of the position, arranging all the band points in the order from small to large, andconnecting, and marking the obtained result as a characteristic curve of the position; acquiring a characteristic curve of each position in the surface of the barrier film according to the method; marking any one position and any one position in eight adjacent domains as adjacent positions, and marking the arbitrary two adjacent positions as the first position
Position and adjacent position->
For example, its degree of classification
The calculation method of (1) is as follows:
wherein,,
indicate->
Characteristic curve of individual positions and adjacent positions +.>
DTW distance of characteristic curve of +.>
Indicate->
Individual position +.>
European distance,/, of->
An exponential function based on a natural constant; the smaller the DTW distance is, the more similar the characteristic curves are, the greater the possibility that two positions belong to the same area is, meanwhile, the smaller the Euclidean distance is, the closer the two positions are, and the more should be dividedTo the same area; according to the method, the classification degree of any two adjacent positions in the surface of the barrier film is obtained, a preset second threshold value is given for area division, the preset second threshold value is calculated by 0.51 in the embodiment, and if the classification degree of the two adjacent positions is greater than the preset second threshold value, the two positions belong to the same area; if the classification degree of the two adjacent positions is smaller than or equal to a preset second threshold value, the two positions belong to different areas; and (3) dividing the areas of all adjacent positions according to the classification degree and a preset second threshold value, and finally dividing the surface of the barrier film into a plurality of areas.
Further, after obtaining a plurality of areas on the surface of the barrier film, taking any one area as an example, obtaining variances of all classification degrees in the area, marking the variances as classification variances of the area, obtaining classification variances of each area, performing linear normalization on all classification variances, and marking the obtained result as the data fluctuation degree of each area.
Thus, according to the spectral reflectivity of each wave band of each position, a plurality of characteristic wave band ranges of each position, a plurality of areas on the surface of the barrier film and the data fluctuation degree of each area are obtained.
Step S003, initial adjustment of each wave band of each position is obtained according to the spectral reflectivity and the data fluctuation degree, the initial adjustment is corrected according to the characteristic wave band range to obtain correction adjustment of each wave band of each position, and the adjusted hyperspectral data are obtained according to the correction adjustment.
The wavelet threshold denoising is to obtain wavelet coefficients by performing wavelet change on the IMF obtained by performing EMD decomposition on the hyperspectral data of each position, and performing threshold processing on the wavelet coefficients by setting a threshold; in order to accurately distinguish the detail information and noise of the hyperspectral data by the wavelet coefficient, accurate IMF needs to be obtained, namely, the hyperspectral data of each position needs to be adaptively adjusted correspondingly; the initial adjustment and correction adjustment of each wave band in the hyperspectral data of each position are obtained according to the distribution characteristics of the hyperspectral data of each position in the same region and the influence of the wave band of the hyperspectral data on the characteristic structure of the characteristic wave band range of the hyperspectral data.
It should be further noted that, the degree of data fluctuation obtained in step S002 is obtained by quantizing the whole region, and it is necessary to quantize the degree of data fluctuation in the local range of each position, thereby obtaining the reference weight; if the difference between the classification degree variances before and after removing a certain position in the local range is larger, the data difference between the position and other positions in the local range is larger, the influence on the local range is larger, and the influence on the data distribution characteristics of the area is larger.
Specifically, by the first
The local area adopts +.>
Is described for a rectangular window range of size, wherein +.>
And->
Respectively represent +.>
Maximum value of range in row direction and maximum value of range in column direction in the region to which the position belongs, +.>
Indicating rounding-off the whole symbol, +.>
Is super-parameter, and the implementer can depend on the actual situation; obtaining +.>
The variance of all degree of classification in the local range of the individual positions is denoted as +.>
A first local variance of the individual locations; get->
Partial range of individual positions except for the relevant +.>
Variance of all other degrees of classification than the degree of classification of the individual position, i.e. remove +.>
All the classification degrees of the position participating in the calculation are calculated, the variance of other classification degrees is calculated, and the obtained result is marked as +.>
A second local variance of the individual locations; then->
Reference weight of local range of individual positions +.>
The calculation method of (1) is as follows:
wherein,,
indicate->
First local variance of the individual positions, +.>
Indicate->
Second local variance of the individual positions, +.>
Represent the first/>
The degree of fluctuation of the data of the area to which the location belongs, +.>
Representing a maximum function>
Representing absolute value; the larger the difference between the first local variance and the second local variance is, the more the position can reflect the data distribution characteristics of the area, and the larger the reference weight is.
Further, in the first step
No. 5 of the individual positions>
The individual bands are exemplified by their initial adjustability +.>
The calculation method of (1) is as follows:
wherein,,
indicate->
Position->
Spectral reflectance of individual bands, +.>
Indicate->
The region of the individual band is at +.>
Spectral reflectance mean value of individual bands, +.>
Indicate->
The region of the individual band is at +.>
Spectral reflectance maximum of the individual bands, +.>
Indicate->
The region of the individual band is at +.>
Spectral reflectance minimum of the individual bands, < >>
Indicate->
Reference weights for local ranges of individual locations; the larger the difference between the spectral reflectivity and the mean value is, the more the data of the wave band at the position should be adjusted so as to accord with the whole distribution of the region, and the larger the initial adjustment is; the larger the reference weight is, the more the data of the wave band at the position can reflect the overall data distribution characteristics of the region, the less the data distribution characteristics are required to be adjusted so as to be reserved, and the smaller the initial adjustment is; the initial adjustability of each band at each location is obtained as described above.
Further, since the hyperspectral data of each position has a characteristic band range, if the spectral characteristic information is lost after the band adjustment, the initial adjustment of each band needs to be corrected and optimized according to the characteristic band range to be the first
No. 5 of the individual positions>
For example, the initial adjustability of the individual bands is obtained with correction adjustability +.>
The calculation method of (1) is as follows:
wherein,,
indicate->
Position->
Spectral reflectance of individual bands, +.>
Indicate->
Position->
Spectral reflectivity of the nearest wavebands point to the corresponding band point of the individual band, +.>
Indicate->
Position->
The abscissa of the corresponding band point of the individual band, +.>
Indicate->
Position->
The abscissa of the nearest wave point to the corresponding band point of the individual band, +.>
Indicate->
No. 5 of the individual positions>
Initial adjustability of the individual bands,/->
Represents an exponential function based on natural constants, < ->
Representing absolute value; the larger the difference of the spectral reflectivity is, the larger the influence of the band on the characteristic band range is, the smaller the adjustment is, and the smaller the correction adjustment is; the larger the difference of the horizontal coordinates is, the larger the difference of the wave bands between the wave band point and the wave point is, the smaller the influence on the range of the characteristic wave band is, the larger the adjustment is, and the larger the correction adjustment is; the correction adjustability of each band at each position is obtained according to the above method.
Further, in the first step
The position is exemplified by>
Performing linear normalization on correction adjustability of all the wave bands at each position, marking the obtained result as an adjustment rate of each wave band, giving a preset third threshold value for judging the wave band needing to be adjusted, marking the adjustment rate larger than the preset third threshold value as an adjustment wave band, and marking the adjustment rate smaller than or equal to the preset third threshold value as a normal wave band; all adjustment bands of each position are obtained according to the method.
Further, correction adjustment properties of all adjustment bands are obtained, DBSCAN clustering is carried out on the adjustment bands according to the correction adjustment properties, the clustering distances adopt absolute values of differences between the correction adjustment properties, a plurality of categories are obtained, and each category comprises a plurality of adjustment bands; in order to obtain a better wavelet threshold denoising effect for each IMF component, the data of the same type of adjustment band should be adjusted, that is, an adjustment reference value of each type needs to be obtained; taking any one category as an example, acquiring the average value of the absolute value of the difference value between each corresponding band point of the adjusting band in the category and the adjacent next band point on the spectral reflectivity, and taking the obtained result as an adjusting reference value of the category; acquiring an adjustment reference value of each category according to the method; for any one adjustment wave band, obtaining the absolute value of the difference value between the wave band point corresponding to the adjustment wave band and the nearest fluctuation point on the spectral reflectivity, obtaining the absolute value of the difference value and the absolute value of the difference value of the adjustment reference value, and taking the obtained absolute value of the difference value as the adjustment size of the adjustment wave band; if the fluctuation point of the adjustment wave band, which corresponds to the wave band point and is closest to the wave band point, is the wave crest point, the sum of the spectral reflectivity of the adjustment wave band and the adjustment size is the adjustment value of the adjustment wave band; if the fluctuation point of the adjustment wave band, which corresponds to the wave band point and is closest to the wave band point, is the wave trough point, the difference of the spectral reflectivity of the adjustment wave band minus the adjustment size is the adjustment value of the adjustment wave band; according to the method, the adjustment value of each adjustment wave band is obtained, and the spectral reflectivity of the normal wave band is unchanged, so that the adjusted hyperspectral data are obtained.
Thus, the adjustment value of each adjustment band is obtained, and the adjustment of the hyperspectral data is completed.
And S004, performing EMD decomposition and wavelet threshold denoising according to the adjusted hyperspectral data to obtain denoised hyperspectral data, and finishing the surface quality detection of the barrier film.
EMD (empirical mode decomposition) is carried out on the adjusted hyperspectral data, all IMF components of each position are obtained, wavelet threshold denoising is carried out on each IMF component, and the denoised IMF components are reconstructed to obtain denoised hyperspectral data; the EMD decomposition, IMF component reconstruction and wavelet threshold denoising are all in the prior art, and the embodiment is not described in detail; among parameters of wavelet threshold denoising, the wavelet basis function is Mo Erxiao wave function, the wavelet decomposition layer number is 5, and the soft threshold form denoising is performed.
Further, comparing the denoised hyperspectral data with ideal hyperspectral data of the surface of the barrier film obtained according to priori knowledge, wherein the spectrum comparison difference adopts Manhattan distance to carry out similarity measurement, the smaller the Manhattan distance is, the larger the similarity is, and the spectral similarity calculation method of the denoised hyperspectral data and the ideal hyperspectral data of each position is as follows:
，/>
the Manhattan distance of two spectrums corresponding to the same position is represented, and the purpose of adding 1 to the denominator is to avoid that the denominator is 0 to influence the calculation result; giving a preset fourth threshold value for judging the spectrum similarity of each position, calculating the preset fourth threshold value by adopting 0.6, obtaining the position of which the spectrum similarity of the surface of the barrier film is smaller than the preset fourth threshold value, marking the position as an abnormal position, obtaining the number of the abnormal positions in the surface of the barrier film, and marking the number as an abnormal number; the position with the spectrum similarity larger than or equal to a preset fourth threshold value is recorded as a normal position; giving a preset fifth threshold value for judging the surface quality of the barrier film, wherein the preset fifth threshold value is calculated by adopting 0.1, and if the abnormal number is more than or equal to the product of the preset fifth threshold value and the total amount of the surface position of the barrier film, the surface quality of the barrier film is poor; if the abnormal quantity is smaller than the product of the preset fifth threshold value and the total quantity of the surface positions of the barrier film, the surface quality of the barrier film is normal.
Thus, the quality detection of the surface of the barrier film is completed through hyperspectral technology.
The foregoing description of the preferred embodiments of the invention is not intended to be limiting, but rather is intended to cover all modifications, equivalents, alternatives, and improvements that fall within the spirit and scope of the invention.
Claims (10)
1. The barrier film surface quality detection method based on the optical technology is characterized by comprising the following steps of:
collecting hyperspectral data of each position on the surface of the barrier film;
acquiring a plurality of wave band points and wave band curves of each position according to hyperspectral data of each position, acquiring a plurality of characteristic points of each position according to the wave band curves, acquiring a plurality of wave points and a plurality of characteristic wave band ranges of each position according to the distribution of the characteristic points in the wave band curves, acquiring classification degrees of any two adjacent positions according to the characteristic wave band ranges of each position, and acquiring a plurality of areas on the surface of the barrier film and the data fluctuation degree of each area according to the classification degrees;
obtaining reference weight of a local range of each position according to the classification degree and the data fluctuation degree, obtaining initial adjustment of each wave band of each position according to the reference weight and the hyperspectral data, obtaining correction adjustment of each wave band of each position according to the wave band curve and the fluctuation point, obtaining all adjustment wave bands of each position according to the correction adjustment of each wave band of each position, obtaining a plurality of categories according to correction adjustment clustering of all adjustment wave bands, and obtaining adjustment values of each adjustment wave band according to the adjustment wave bands, the hyperspectral data and the fluctuation points in each category, and obtaining adjusted hyperspectral data;
and carrying out EMD decomposition and wavelet threshold denoising according to the adjusted hyperspectral data to obtain denoised hyperspectral data, and finishing the surface quality detection of the barrier film.
2. The method for detecting the surface quality of the barrier film based on the optical technology according to claim 1, wherein the steps of obtaining a plurality of band points and band curves of each position comprise the following specific steps:
taking any one position on the surface of the barrier film as a target position, acquiring the spectral reflectivity of the target position in each wave band in hyperspectral data, taking the abscissa as the wave band, taking the ordinate as the spectral reflectivity, constructing a coordinate system, placing each wave band of the target position and the spectral reflectivity thereof in the coordinate system to obtain a plurality of wave band points, and connecting adjacent wave band points of the abscissa to obtain a wave band curve of the target position;
and acquiring a plurality of wave band points and wave band curves of each position.
3. The method for detecting the surface quality of the barrier film based on the optical technology according to claim 1, wherein the method for obtaining the plurality of characteristic points of each position according to the band curve comprises the following specific steps:
taking any one position of the surface of the barrier film as a target position, for two adjacent wave band points in a wave band curve of the target position, marking the ratio between the difference value obtained by subtracting the spectral reflectivity of the former wave band point from the spectral reflectivity of the latter wave band point and the absolute value of the difference value of the abscissa of the two wave band points as the wave band slope of the former wave band point, and obtaining the wave band slope of each wave band point of the target position;
taking any one of the wave band points as a target wave band point, if sign changes occur on the wave band slope of the target wave band point and the next adjacent wave band point, marking the target wave band point as a characteristic point, and acquiring all characteristic points in a wave band curve of a target position, wherein 2 times of the wave band distance between each characteristic point and the adjacent previous characteristic point is taken as the wave band width of each characteristic point, and the wave band distance represents the absolute value of the difference value of the abscissa between the two wave band points;
and acquiring a plurality of characteristic points of each position and the wave band width of each characteristic point.
4. The method for detecting the surface quality of the barrier film based on the optical technology according to claim 3, wherein the steps of obtaining a plurality of fluctuation points and a plurality of characteristic wave band ranges of each position comprise the following specific steps:
taking any position on the surface of the barrier film as a target position, and the first position in a wave band curve of the target position
Degree of characteristic of individual characteristic points->
The calculation method of (1) is as follows:
wherein,,
indicate->
Band width of individual feature points, +.>
Band width maximum value in band curve representing target position,/->
Indicate->
Each band point in the band range of the feature points is associated with +.>
The absolute value of the difference between the spectral reflectances of the characteristic points is the smallest, the band range is expressed as +.>
The abscissa of each characteristic point is taken as the center, the wave band width is a window constructed by the size, the abscissa of the wave band point is in the window, and the wave band point is in the wave band range;
the characteristic degree of each characteristic point in the wave band curve of the target position is obtained, all the characteristic degrees are normalized, the characteristic points with the normalized value of the characteristic degree larger than a preset first threshold value are marked as fluctuation points, and the wave band range of the fluctuation points is marked as the characteristic wave band range; and acquiring a plurality of fluctuation points and a plurality of characteristic wave band ranges of each position.
5. The method for detecting the surface quality of the barrier film based on the optical technology according to claim 1, wherein the method for obtaining the classification degree of any two adjacent positions according to the characteristic wave band range of each position comprises the following specific steps:
taking any position on the surface of the barrier film as a target position, extracting wave band points included in all characteristic wave band ranges of the target position, arranging and connecting all wave band points according to the sequence from small wave band to large wave band, and recording the obtained result as a characteristic curve of the target position;
acquiring a characteristic curve of each position; marking any one position and any one position in eight adjacent domains as adjacent positions, and the first position
Individual position +.>
Classification degree of->
The calculation method of (1) is as follows:
wherein,,
indicate->
Characteristic curve of individual positions and adjacent positions +.>
DTW distance of characteristic curve of +.>
Indicate->
Individual position +.>
European distance,/, of->
An exponential function based on a natural constant;
and obtaining the classification degree of any two adjacent positions.
6. The method for detecting the surface quality of the barrier film based on the optical technology according to claim 1, wherein the method for obtaining the plurality of areas on the surface of the barrier film and the data fluctuation degree of each area according to the classification degree comprises the following specific steps:
if the classification degree of the two adjacent positions is greater than a preset second threshold value, the two positions belong to the same area; if the classification degree of the two adjacent positions is smaller than or equal to a preset second threshold value, the two positions belong to different areas; dividing the areas of all adjacent positions according to the classification degree and a preset second threshold value, and dividing the surface of the barrier film into a plurality of areas;
taking any one area as a target area, acquiring variances of all classification degrees in the target area, marking the variances as the classification variances of the target area, acquiring the classification variances of each area, normalizing all the classification variances, and marking the obtained result as the data fluctuation degree of each area.
7. The method for detecting the surface quality of the barrier film based on the optical technology according to claim 1, wherein the method for obtaining the reference weight of the local range of each position according to the classification degree and the data fluctuation degree comprises the following specific steps:
in the first place
The individual positions are the center and the local range is +.>
Rectangular window of size, wherein +.>
And->
Respectively represent +.>
Maximum value of range in row direction and maximum value of range in column direction in the region to which the position belongs, +.>
Indicating rounding-off the whole symbol, +.>
Is a super parameter;
acquisition of the first
The variance of all degree of classification in the local range of the individual positions is denoted as +.>
A first local variance of the individual locations; get->
Partial range of individual positions except for the relevant +.>
The variance of all other classification levels except the classification level of the individual position, the result obtained is recorded as +.>
A second local variance of the individual locations; first->
Reference weight of local range of individual positions +.>
The calculation method of (1) is as follows:
wherein,,
indicate->
First local variance of the individual positions, +.>
Indicate->
Second local variance of the individual positions, +.>
Indicate->
The degree of fluctuation of the data of the area to which the location belongs, +.>
Representing a maximum function>
Representing absolute values.
8. The method for detecting the surface quality of the barrier film based on the optical technology according to claim 1, wherein the method for obtaining the initial adjustability of each band of each position comprises the following specific steps:
wherein,,
indicate->
No. 5 of the individual positions>
Initial adjustability of the individual bands,/->
Indicate->
Position->
Spectral reflectance of individual bands, +.>
Indicate->
The region of the individual band is at +.>
Spectral reflectance mean value of individual bands, +.>
Indicate->
The region of the individual band is at +.>
Spectral reflectance maximum of the individual bands, +.>
Indicate->
The region of the individual band is at +.>
Spectral reflectance minimum of the individual bands, < >>
Indicate->
Reference weights for local ranges of individual locations.
9. The method for detecting the surface quality of the barrier film based on the optical technology according to claim 1, wherein the correction adjustment of each band at each position is obtained by correcting the band curve and the fluctuation point, comprising the following specific steps:
wherein,,
indicate->
Position->
Correction adjustability of individual bands,>
indicate->
Position->
Spectral reflectance of individual bands, +.>
Indicate->
Position->
Spectral reflectivity of the nearest wavebands point to the corresponding band point of the individual band, +.>
Indicate->
Position->
The abscissa of the corresponding band point of the individual band, +.>
Indicate->
Position->
The abscissa of the nearest wave point to the corresponding band point of the individual band, +.>
Indicate->
No. 5 of the individual positions>
Initial adjustability of the individual bands,/->
Represents an exponential function based on natural constants, < ->
Representing absolute values.
10. The method for detecting the surface quality of the barrier film based on the optical technology according to claim 1, wherein the step of obtaining the adjustment value of each adjustment band comprises the following specific steps:
taking any one of the categories as a target, acquiring the average value of the absolute value of the difference value between each corresponding band point of the adjustment band in the target category and the adjacent next band point on the spectral reflectivity, and taking the obtained result as an adjustment reference value of the target category; acquiring an adjustment reference value of each category;
any one adjustment wave band in the target class is a target adjustment wave band, the absolute value of the difference value between the wave band point corresponding to the target adjustment wave band and the nearest fluctuation point in the spectral reflectivity is obtained, the absolute value of the difference value between the absolute value of the difference value and the adjustment reference value is obtained, and the obtained absolute value of the difference value is used as the adjustment size of the target adjustment wave band;
if the fluctuation point of the target adjustment wave band, which corresponds to the wave band point and is closest to the wave band point, is the wave crest point, the sum of the spectral reflectivity of the target adjustment wave band and the adjustment size is the adjustment value of the target adjustment wave band; if the fluctuation point of the target adjustment wave band, which corresponds to the wave band point and is closest to the wave band point, is the wave trough point, the difference of the spectral reflectivity of the target adjustment wave band minus the adjustment size is the adjustment value of the target adjustment wave band; and acquiring an adjustment value of each adjustment band.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310564883.8A CN116297530B (en) | 2023-05-19 | 2023-05-19 | Barrier film surface quality detection method based on optical technology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310564883.8A CN116297530B (en) | 2023-05-19 | 2023-05-19 | Barrier film surface quality detection method based on optical technology |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116297530A true CN116297530A (en) | 2023-06-23 |
| CN116297530B CN116297530B (en) | 2023-07-25 |
Family
ID=86801701
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310564883.8A Active CN116297530B (en) | 2023-05-19 | 2023-05-19 | Barrier film surface quality detection method based on optical technology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116297530B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116754511A (en) * | 2023-08-18 | 2023-09-15 | 天津博霆光电技术有限公司 | Indocyanine green detection method and system based on spectrum technology |
| CN117112979A (en) * | 2023-10-20 | 2023-11-24 | 昆山尚瑞智能科技有限公司 | Error compensation optimization method in spectrum measurement process |
| CN117330521A (en) * | 2023-12-01 | 2024-01-02 | 黑龙江中医药大学 | Clinical laboratory uses blood smear system |
| CN118032695A (en) * | 2024-04-02 | 2024-05-14 | 山东创宇能源科技股份有限公司 | Nitrogen-sulfur in-situ measurement method and system based on differential ultraviolet spectrum technology |
| CN118051862A (en) * | 2024-04-16 | 2024-05-17 | 洛阳禾安工程技术服务有限公司 | Detection method and system for building paint |
| CN118329737A (en) * | 2024-06-12 | 2024-07-12 | 广东正一包装股份有限公司 | Intelligent detection method and system for leakage rate of barrier film |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5121337A (en) * | 1990-10-15 | 1992-06-09 | Exxon Research And Engineering Company | Method for correcting spectral data for data due to the spectral measurement process itself and estimating unknown property and/or composition data of a sample using such method |
| WO2016000088A1 (en) * | 2014-07-02 | 2016-01-07 | 中国林业科学研究院林业研究所 | Hyperspectral waveband extraction method based on optimal index factor-correlation coefficient method |
| CN105989592A (en) * | 2015-02-11 | 2016-10-05 | 中国科学院西安光学精密机械研究所 | Hyperspectral image waveband selection method based on double clustering and neighborhood analysis |
| CN106096571A (en) * | 2016-06-22 | 2016-11-09 | 北京化工大学 | A kind of based on EMD feature extraction with the cell sorting method of rarefaction representation |
| CN107132205A (en) * | 2017-06-12 | 2017-09-05 | 华中科技大学 | One kind is based on the enhanced Spectral matching method and system of all band feature |
| CN109583469A (en) * | 2018-10-17 | 2019-04-05 | 浙江工业大学 | K mean value high spectrum image wave band clustering method based on mutual information |
| CN109871768A (en) * | 2019-01-18 | 2019-06-11 | 西北工业大学 | The optimal band selection method of EO-1 hyperion based on shared arest neighbors |
-
2023
- 2023-05-19 CN CN202310564883.8A patent/CN116297530B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5121337A (en) * | 1990-10-15 | 1992-06-09 | Exxon Research And Engineering Company | Method for correcting spectral data for data due to the spectral measurement process itself and estimating unknown property and/or composition data of a sample using such method |
| WO2016000088A1 (en) * | 2014-07-02 | 2016-01-07 | 中国林业科学研究院林业研究所 | Hyperspectral waveband extraction method based on optimal index factor-correlation coefficient method |
| CN105989592A (en) * | 2015-02-11 | 2016-10-05 | 中国科学院西安光学精密机械研究所 | Hyperspectral image waveband selection method based on double clustering and neighborhood analysis |
| CN106096571A (en) * | 2016-06-22 | 2016-11-09 | 北京化工大学 | A kind of based on EMD feature extraction with the cell sorting method of rarefaction representation |
| CN107132205A (en) * | 2017-06-12 | 2017-09-05 | 华中科技大学 | One kind is based on the enhanced Spectral matching method and system of all band feature |
| CN109583469A (en) * | 2018-10-17 | 2019-04-05 | 浙江工业大学 | K mean value high spectrum image wave band clustering method based on mutual information |
| CN109871768A (en) * | 2019-01-18 | 2019-06-11 | 西北工业大学 | The optimal band selection method of EO-1 hyperion based on shared arest neighbors |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116754511A (en) * | 2023-08-18 | 2023-09-15 | 天津博霆光电技术有限公司 | Indocyanine green detection method and system based on spectrum technology |
| CN116754511B (en) * | 2023-08-18 | 2023-10-27 | 天津博霆光电技术有限公司 | Indocyanine green detection method and system based on spectrum technology |
| CN117112979A (en) * | 2023-10-20 | 2023-11-24 | 昆山尚瑞智能科技有限公司 | Error compensation optimization method in spectrum measurement process |
| CN117112979B (en) * | 2023-10-20 | 2024-02-06 | 昆山尚瑞智能科技有限公司 | Error compensation optimization method in spectrum measurement process |
| CN117330521A (en) * | 2023-12-01 | 2024-01-02 | 黑龙江中医药大学 | Clinical laboratory uses blood smear system |
| CN117330521B (en) * | 2023-12-01 | 2024-02-20 | 黑龙江中医药大学 | Clinical laboratory uses blood smear system |
| CN118032695A (en) * | 2024-04-02 | 2024-05-14 | 山东创宇能源科技股份有限公司 | Nitrogen-sulfur in-situ measurement method and system based on differential ultraviolet spectrum technology |
| CN118051862A (en) * | 2024-04-16 | 2024-05-17 | 洛阳禾安工程技术服务有限公司 | Detection method and system for building paint |
| CN118329737A (en) * | 2024-06-12 | 2024-07-12 | 广东正一包装股份有限公司 | Intelligent detection method and system for leakage rate of barrier film |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116297530B (en) | 2023-07-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116297530B (en) | Barrier film surface quality detection method based on optical technology | |
| CN115018844B (en) | Plastic film quality evaluation method based on artificial intelligence | |
| CN105891230B (en) | Fruit appearance detection method based on spectral image analysis | |
| CN116246174B (en) | Sweet potato variety identification method based on image processing | |
| CN117011292B (en) | Method for rapidly detecting surface quality of composite board | |
| CN113406015B (en) | Transparency calculation method and system for water bodies of near-shore and inland waters | |
| CN116935384B (en) | Intelligent detection method for cell abnormality sample | |
| CN114445386B (en) | PVC pipe quality detection and evaluation method and system based on artificial intelligence | |
| CN115131683B (en) | Forestry information identification method based on high-resolution remote sensing image | |
| CN117591905B (en) | Pork safety detection method based on hyperspectral features | |
| CN116245880A (en) | Electric vehicle charging pile fire risk detection method based on infrared identification | |
| CN113989639A (en) | Hyperspectral image analysis processing method-based automatic litchi disease identification method and device | |
| CN118225711B (en) | Intelligent soil water potential detection method based on spectral analysis | |
| CN117333489B (en) | Film damage detection device and detection system | |
| CN117974671B (en) | Watch dial defect intelligent detection method based on artificial intelligence | |
| CN115170958A (en) | Method for calculating wilting degree of soybean canopy by adopting multi-spectral image Fourier transform | |
| CN117191716A (en) | Magnetic printed matter quality real-time detection method based on artificial intelligence | |
| CN117330521B (en) | Clinical laboratory uses blood smear system | |
| CN114965316A (en) | Hyperspectral multi-parameter based water quality monitoring method and computer program product | |
| CN116883674A (en) | Multispectral image denoising device and food quality detection system using same | |
| CN109946246B (en) | Detection method and device for soluble solid of apple | |
| Liao et al. | Rapeseed storage quality detection using hyperspectral image technology-an application for future smart cities | |
| CN117783051A (en) | Methane gas leakage detection method based on multi-sensor data fusion | |
| CN117522744A (en) | Spectrum quantization correction method based on supervised manifold learning | |
| CN116994675B (en) | Brocade based on near infrared data Lantern calyx epidermis detection method |
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 |