CN114998205A - Method for detecting foreign matters in bottle in liquid filling process based on optical means - Google Patents
Method for detecting foreign matters in bottle in liquid filling process based on optical means Download PDFInfo
- Publication number
- CN114998205A CN114998205A CN202210453932.6A CN202210453932A CN114998205A CN 114998205 A CN114998205 A CN 114998205A CN 202210453932 A CN202210453932 A CN 202210453932A CN 114998205 A CN114998205 A CN 114998205A
- Authority
- CN
- China
- Prior art keywords
- image
- pixel point
- foreign matter
- value
- determining
- 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
- 239000007788 liquid Substances 0.000 title claims abstract description 154
- 238000000034 method Methods 0.000 title claims abstract description 36
- 230000003287 optical effect Effects 0.000 title claims abstract description 24
- 238000005429 filling process Methods 0.000 title claims abstract description 18
- 238000001914 filtration Methods 0.000 claims abstract description 129
- 230000011218 segmentation Effects 0.000 claims description 23
- 230000000877 morphologic effect Effects 0.000 claims description 22
- 238000004364 calculation method Methods 0.000 claims description 20
- 238000005315 distribution function Methods 0.000 claims description 18
- 238000012935 Averaging Methods 0.000 claims description 3
- 238000001514 detection method Methods 0.000 abstract description 20
- 238000004458 analytical method Methods 0.000 abstract description 6
- 238000012360 testing method Methods 0.000 abstract description 5
- 239000000463 material Substances 0.000 abstract description 4
- 238000004154 testing of material Methods 0.000 abstract description 2
- 230000008569 process Effects 0.000 description 12
- 210000000746 body region Anatomy 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 239000000243 solution Substances 0.000 description 4
- 238000009826 distribution Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 239000012263 liquid product Substances 0.000 description 3
- 230000003562 morphometric effect Effects 0.000 description 3
- 238000013425 morphometry Methods 0.000 description 3
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 239000011521 glass Substances 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 239000000047 product Substances 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 208000003464 asthenopia Diseases 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 239000003651 drinking water Substances 0.000 description 1
- 235000020188 drinking water Nutrition 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 239000003889 eye drop Substances 0.000 description 1
- 229940012356 eye drops Drugs 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 239000003978 infusion fluid Substances 0.000 description 1
- 238000002347 injection Methods 0.000 description 1
- 239000007924 injection Substances 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000004806 packaging method and process Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Quality & Reliability (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
The invention relates to the technical field of material testing and analysis, in particular to a method for detecting foreign matters in a liquid filling process bottle based on an optical means, which comprises the steps of collecting a visible light image of a filling liquid bottle to be detected by the optical means, carrying out material analysis and testing according to the visible light image of the filling liquid bottle to be detected, further determining a characteristic enhancement image corresponding to a multi-scale filtering image, and determining an optimal characteristic enhancement image from the characteristic enhancement image corresponding to the multi-scale filtering image; determining a binary image corresponding to the optimal feature enhancement image according to the optimal feature enhancement image; and determining each foreign matter area in each frame of filling liquid bottle image to be detected according to the filtering image and the binary image corresponding to the optimal characteristic enhancement image, and judging whether real foreign matters exist in the filling liquid bottle to be detected. The invention utilizes the visible light means to analyze and test materials, solves the problem of low detection accuracy of the foreign matters in the existing filling liquid bottle, and improves the accuracy of the detection result of the foreign matters in the filling liquid bottle.
Description
Technical Field
The invention relates to the technical field of material testing and analysis, in particular to a method for detecting foreign matters in a bottle in a liquid filling process based on an optical means.
Background
In order to meet the daily life requirements of consumers, most liquid products are filled in a form of glass bottles or plastic bottles, such as injection, infusion solutions, oral liquids, eye drops, drinking water and other various liquid products. The liquid products have visible foreign matters such as glass scraps, suspended matters and the like in the filling liquid bottle due to the reasons of poor filtering, poor container cleaning, collision during packaging and the like in the production and filling processes, and the quality of the products is greatly influenced by the foreign matters in the filling liquid bottle, and meanwhile, the safety problem is also brought. Therefore, manufacturers need to adopt a certain detection means to detect foreign matters in the sealed filling liquid bottles, and find out and remove unqualified products with foreign matters.
At present, the foreign matter detection method of the filling liquid bottle is mostly manual light inspection, and the manual light inspection means that the packaged filling liquid bottle is slightly shaken or inverted manually, and is aimed at a special light source to observe whether the liquid bottle contains suspended foreign matters or not. However, the detection method is time-consuming and labor-consuming, low in efficiency, visual fatigue is easy to generate when the detection method works for a long time, the accuracy of foreign matter detection can be reduced, and the manual lamp detection can cause harm to the bodies of workers to a certain extent. Along with the improvement of intelligent level, machine is examined to semi-automatic lamp and machine is examined to full-automatic lamp have appeared, and machine is examined to semi-automatic lamp has adopted simple machinery to rotate and optical system to realize the semi-automatization that artifical lamp was examined, and this foreign matter detection device has alleviateed artifical work to a certain extent, nevertheless still need examine the analysis through the lamp and judge the foreign matter, and the lamp is examined the foreign matter in the unable accurate differentiateing liquid bottle, leads to the foreign matter to detect the precision and the degree of accuracy and still not obtain the improvement.
Disclosure of Invention
In order to solve the problem of low accuracy of detecting foreign matters in the existing liquid filling bottle, the invention aims to provide a method for detecting foreign matters in a liquid filling process bottle based on an optical means.
The invention provides a method for detecting foreign matters in a bottle in a liquid filling process based on an optical means, which comprises the following steps:
collecting N frames of filling liquid bottle visible light images to be detected, and determining multi-scale filtering images of each frame of filling liquid bottle visible light images to be detected according to each frame of filling liquid bottle visible light images to be detected;
determining a characteristic enhancement image corresponding to each multi-scale filtering image according to the multi-scale filtering image of each frame of filling liquid bottle visible light image to be detected, and further determining an optimal characteristic enhancement image from the characteristic enhancement images corresponding to the multi-scale filtering images;
determining a binary image of the optimal characteristic enhancement image according to each pixel point and pixel value of the optimal characteristic enhancement image corresponding to each frame of filling liquid bottle visible image to be detected;
determining each foreign matter area in each frame of visible light image of the filling liquid bottle to be detected according to the filtering image and the binary image corresponding to the optimal characteristic enhancement image;
and judging whether real foreign matters exist in the filling liquid bottle to be detected or not according to the number and the gray value of pixel points in each foreign matter area in each frame of visible light image of the filling liquid bottle to be detected.
Further, the step of determining the feature enhancement map corresponding to the multi-scale filtered image comprises:
determining a first characteristic value of each pixel point in a multi-scale filtering image according to the gray value of each pixel point in the multi-scale filtering image of each frame of filling liquid bottle visible light image to be detected;
determining HSV images corresponding to the multi-scale filtering images according to the multi-scale filtering images of the visible light images of the filling liquid bottles to be detected in each frame, and further determining second characteristic values of pixel points in the multi-scale filtering images;
determining the attention degree value of each pixel point in the multi-scale filtering image according to the first characteristic value and the second characteristic value of each pixel point in the multi-scale filtering image;
and determining a characteristic enhancement map corresponding to the multi-scale filtering image of the visible light image of the filling liquid bottle to be detected according to the attention degree value of each pixel point in the multi-scale filtering image.
Further, a calculation formula for determining the first characteristic value of each pixel point in the multi-scale filtering image is as follows:
wherein s is i 1 For the first feature value of the ith pixel point in the multi-scale filtered image,
the frequency of the gray value of the jth pixel except the gray value of the ith pixel point in the multi-scale filtering image appearing in the corresponding scale filtering image is I i Is the gray value, I, of the ith pixel point in the multi-scale filtering image j And exp () is an exponential function for the jth pixel gray value in the multi-scale filtering image except for the gray value of the ith pixel point.
The calculation formula for determining the second characteristic value of each pixel point in the multi-scale filtering image is as follows:
wherein s is i 2 Is the second characteristic value, H, of the ith pixel point in the multi-scale filtering image i Tone value S of ith pixel point in HSV image corresponding to multi-scale filtering image i Is the saturation value V of the ith pixel point in the HSV image corresponding to the multi-scale filtering image i The lightness value of the ith pixel point in the HSV image corresponding to the multi-scale filtering image,
the average value of the hue value of each pixel point in the HSV image corresponding to the multi-scale filtering image,
saturation of each pixel point in HSV image corresponding to multi-scale filtering imageThe average of the values,
and averaging the brightness values of all the pixel points in the HSV image corresponding to the multi-scale filtering image.
Further, the step of determining the binary image of the optimal feature enhancement map comprises:
constructing a sliding window with the size of C, sliding the sliding window on the optimal feature enhancement graph, obtaining the sliding window which corresponds to each pixel point in the optimal feature enhancement graph and takes the pixel point as the center, and further obtaining each pixel point in the optimal feature enhancement graph and each neighborhood pixel point;
determining a segmentation threshold corresponding to each pixel point in the optimal feature enhancement graph according to the position of each pixel point, the position of each pixel point corresponding to each neighborhood pixel point and the pixel value;
and determining a binary image of the optimal feature enhancement image according to the segmentation threshold corresponding to each pixel point in the optimal feature enhancement image.
Further, the calculation formula for determining the segmentation threshold corresponding to each pixel point in the optimal feature enhancement map is as follows:
wherein e (x, y) is a segmentation threshold corresponding to each pixel point in the optimal feature enhancement map, C is a region of a sliding window corresponding to each pixel point in the optimal feature enhancement map and centered on the pixel point, Bz (i, j) is a pixel value of each pixel point in the optimal feature enhancement map corresponding to each neighborhood pixel point, (x, y) is a position of each pixel point in the optimal feature enhancement map, (i, j) is a position of each neighborhood pixel point of each pixel point in the optimal feature enhancement map, and q is a bias term.
Further, the step of determining the optimal feature enhanced image from the feature enhanced image corresponding to the multi-scale filtered image comprises:
determining a judgment factor of the feature enhanced image corresponding to the multi-scale filter image according to the pixel value of each pixel point in the feature enhanced image corresponding to the multi-scale filter image;
and according to the judgment factor of the feature enhanced image corresponding to the multi-scale filter map, taking the feature enhanced image with the minimum judgment factor as the optimal feature enhanced image.
Further, the calculation formula for determining the decision factor of the feature enhanced image corresponding to the multi-scale filter map is as follows:
wherein rho is a judgment factor of the characteristic enhanced image corresponding to the multi-scale filter map, and g i The gray value of the ith pixel point in the multi-scale filter graph is shown, and K is the number of the pixel points corresponding to the multi-scale filter graph.
Further, the step of judging whether the filled liquid bottle to be detected has real foreign matters comprises:
determining the area corresponding to each foreign body area and the number of edge pixel points according to the number of the pixel points in each foreign body area in each frame of visible light image of the filling liquid bottle to be detected;
determining the morphological index value of each foreign matter region according to the area corresponding to each foreign matter region and the number of edge pixel points;
and judging whether the real foreign matter areas exist in the visible light image of the filling liquid bottle to be detected according to the morphological index values of the foreign matter areas.
Further, the step of judging whether the filled liquid bottle to be detected has real foreign matters comprises:
if the morphological index value of any foreign body area is smaller than the morphological index threshold value, the foreign body area is judged to be a real foreign body area, and if the morphological index value of any foreign body area is not smaller than the morphological index threshold value, the foreign body area is judged to be a suspected foreign body area;
if the suspected foreign matter area exists, determining a normal area in each frame of visible light image of the filling liquid bottle to be detected according to the filtering image and the binary image corresponding to the optimal feature enhancement image;
determining the credibility of each suspected foreign matter area as a real foreign matter area according to the gray value of each pixel point in each suspected foreign matter area in each frame of visible light image of the filling liquid bottle to be detected and the gray value of each pixel point in the normal area;
if the reliability of any suspected foreign matter area as a real foreign matter area is greater than the reliability threshold, the suspected foreign matter area is determined to be a real foreign matter area, and if the reliability of any suspected foreign matter area as a real foreign matter area is not greater than the reliability threshold, the suspected foreign matter area is determined not to be a real foreign matter area.
Further, the step of determining the reliability that each suspected foreign object region is a true foreign object region includes:
determining pixel gray mean values corresponding to each suspected foreign matter region and each normal region according to the gray value of each pixel point in each suspected foreign matter region and the gray value of each pixel point in each normal region in each frame of filling liquid bottle visible light image to be detected, and determining a standard pixel distribution function corresponding to each suspected foreign matter region and each normal region;
determining the credibility of each suspected foreign matter area as a real foreign matter area according to the corresponding pixel gray level mean value of each suspected foreign matter area and the normal area and the corresponding pixel distribution function of each suspected foreign matter area and the normal area, wherein the calculation formula is as follows:
wherein, P a The confidence that the a-th suspected foreign matter area is the real foreign matter,
is the gray level mean value of the pixel corresponding to the a-th suspected foreign matter area,
the mean value of the pixel gray levels corresponding to the normal region, f a (x) Is the pixel distribution function corresponding to the a-th suspected foreign matter area, f 0 (x) The standard pixel distribution function corresponding to the normal area.
The invention has the following beneficial effects:
the method comprises the steps of collecting N frames of visible light images of the filling liquid bottles to be detected by an optical means, carrying out material analysis and test according to each frame of visible light image of the filling liquid bottles to be detected, and further determining multi-scale filtering images of each frame of visible light image of the filling liquid bottles to be detected. According to the multi-scale filtering image of each frame of visible light image of the filling liquid bottle to be detected, the feature enhancement image corresponding to the multi-scale filtering image is determined, and then the optimal feature enhancement image is determined from the feature enhancement image corresponding to the multi-scale filtering image, so that the accuracy and the detection precision of the filling liquid detection bottle are improved in the process
According to each pixel point and pixel value of the optimal characteristic enhancement image corresponding to each frame of filling liquid bottle visible light image to be detected, the binary image of the optimal characteristic enhancement image is determined, and then each foreign matter area in each frame of filling liquid bottle visible light image to be detected is determined. And finally, judging whether the filled liquid bottle to be detected has real foreign matters or not according to the number and the gray value of the pixel points in each foreign matter area in each frame of visible light image of the filled liquid bottle to be detected.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions and advantages of the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
FIG. 1 is a flow chart of a method for detecting foreign matter in a bottle during a liquid filling process based on optical means according to the present invention;
fig. 2 is a flowchart of determining a feature enhancement map corresponding to a multi-scale filtered image according to an embodiment of the present invention.
Detailed Description
To further explain the technical means and effects of the present invention adopted to achieve the predetermined objects, the following detailed description of the embodiments, structures, features and effects of the technical solutions according to the present invention will be given with reference to the accompanying drawings and preferred embodiments. In the following description, different references to "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The embodiment provides a method for detecting foreign matters in a liquid filling process bottle based on an optical means, which comprises the following steps as shown in fig. 1:
(1) collecting N frames of filling liquid bottle visible light images to be detected, and determining multi-scale filtering images of each frame of filling liquid bottle visible light images to be detected according to each frame of filling liquid bottle visible light images to be detected, wherein the method comprises the following steps:
(1-1) the present embodiment detects the foreign matter in the filling liquid bottle mainly based on the image data information, and therefore, it is necessary to acquire the visible light image of the filling liquid bottle to be detected by using an optical means. When collecting images, the current filling liquid bottle to be detected is in a rotating state on the mechanical rotating device, in this embodiment, an optical CCD (charge coupled device) camera is used to continuously shoot N frames of visible light images of the filling liquid bottle to be detected, the visible light images of the filling liquid bottle to be detected are RGB images of the filling liquid bottle to be detected, the N frames of visible light images of the filling liquid bottle to be detected need to contain images of each direction of the filling liquid bottle to be detected, and the images of each direction should be images of the filling liquid bottle to be completely shot, so as to detect foreign matters in the filling liquid bottle to be detected.
The rotation speed of the mechanical rotating device and the time interval between two frames of images before and after the images are captured by the optical CCD camera can be set according to actual conditions. In addition, this embodiment is in one side of waiting to detect the filling liquid bottle with the optics CCD camera to the front view image of waiting to detect the filling liquid bottle is gathered, and the front view image data of waiting to detect the filling liquid bottle is the basic image data that the filling liquid bottle foreign matter detected.
(1-2) in order to eliminate the influence of gaussian noise on the detection of foreign matters in the visible light image of the filling liquid bottle to be detected, according to each frame of visible light image of the filling liquid bottle to be detected, filtering the visible light image of the filling liquid bottle to be detected by each frame through gaussian functions of different scales, so as to obtain a multi-scale filtering image of the visible light image of the filling liquid bottle to be detected by each frame, wherein the sizes of the gaussian templates selected in the embodiment are respectively: 5 × 5, 7 × 7, 9 × 9. It should be noted that the process of the multi-scale gaussian function filtering process is prior art and is not within the scope of the present invention, and will not be described in detail herein.
So far, each frame of filling liquid bottle visible light image that awaits measuring all corresponds the filtering image of 3 yards, is respectively: 5 × 5 filtered images, 7 × 7 filtered images, and 9 × 9 filtered images.
(2) Determining a characteristic enhancement image corresponding to each multi-scale filtering image according to the multi-scale filtering image of each frame of filling liquid bottle visible light image to be detected, and further determining an optimal characteristic enhancement image from the characteristic enhancement images corresponding to the multi-scale filtering images, wherein the method comprises the following steps:
(2-1) according to the multi-scale filtering image of each frame of visible light image of the filling liquid bottle to be detected, determining a characteristic enhancement diagram corresponding to the multi-scale filtering image, wherein a flow chart for determining the characteristic enhancement diagram corresponding to the multi-scale filtering image is shown in fig. 2, and the determining step comprises the following steps:
(2-1-1) determining a first characteristic value of each pixel point in the multi-scale filtering image according to the gray value of each pixel point in the multi-scale filtering image of each frame of filling liquid bottle visible light image to be detected.
In this embodiment, the frequency of each pixel gray value in the filtering image of 3 scales corresponding to the visible light image of each frame of filling liquid bottle to be detected appearing in the filtering image of the corresponding scale is counted, and according to each pixel gray value and the frequency of each pixel gray value appearing in the filtering image of the corresponding scale, the first characteristic value of each pixel point in the multi-scale filtering image is calculated, and the calculation formula is as follows:
wherein s is i 1 For the first feature value of the ith pixel point in the multi-scale filtered image,
the frequency of the gray value of the jth pixel except the gray value of the ith pixel point in the multi-scale filtering image appearing in the corresponding scale filtering image is I i Is the gray value, I, of the ith pixel point in the multi-scale filtering image j The exp () is an exponential function for the jth pixel gray value except the ith pixel gray value in the multi-scale filtering image.
By referring to the determination process of the first characteristic value of the ith pixel point in the multi-scale filtering image, the first characteristic value of each pixel point in the multi-scale filtering image can be obtained, and the first characteristic value can represent the attention range of each pixel point. The first characteristic value refers to the difference between the gray value of each pixel point in the filtered image and the gray values of other pixels in the filtered image. For example, the larger the first characteristic value of a certain pixel point is, the larger the difference between the gray value of the pixel point and the gray values of other pixels in the filtered image is, and the higher the contrast is.
(2-1-2) determining HSV images corresponding to the multi-scale filtering images according to the multi-scale filtering images of the visible light images of the filling liquid bottles to be detected in each frame, and further determining a second characteristic value of each pixel point in the multi-scale filtering images.
In order to improve the characteristic enhancement effect of each pixel point in the filtering image subsequently, the embodiment performs color space conversion on the filtering image of 3 scales corresponding to the visible light image of each frame of the filling liquid bottle to be detected, converts the filtering image into an HSV image, and the converted HSV image has a good visual effect. The specific HSV color space conversion process is prior art and is not within the scope of the present invention, and will not be described in detail herein.
Calculating the mean values of three-channel components of all pixel points in the HSV image corresponding to the 3-scale filtering image according to the HSV image corresponding to the 3-scale filtering image corresponding to each frame of filling liquid bottle visible light image to be detected, wherein the mean values of the three-channel components are the mean value of hue values, the mean value of saturation values and the mean value of brightness values and are recorded as the mean value of hue values, the mean value of saturation values and the mean value of brightness values
Calculating a second characteristic value of each pixel point in the multi-scale filtering image according to the mean value of three-channel components of all pixel points in the HSV image corresponding to the filtering image with 3 scales and the hue value, the saturation value and the brightness value of each pixel point, wherein the calculation formula is as follows:
wherein s is i 2 Is the second characteristic value, H, of the ith pixel point in the multi-scale filtering image i Tone value S of ith pixel point in HSV image corresponding to multi-scale filtering image i Is the saturation value V of the ith pixel point in the HSV image corresponding to the multi-scale filtering image i The lightness value of the ith pixel point in the HSV image corresponding to the multi-scale filtering image,
the average value of the hue value of each pixel point in the HSV image corresponding to the multi-scale filtering image,
the average value of the saturation values of each pixel point in the HSV image corresponding to the multi-scale filtering image,
and averaging the brightness values of all pixel points in the HSV image corresponding to the multi-scale filtering image.
And then, referring to the determination process of the second characteristic value of the ith pixel point in the multi-scale filtering image, obtaining the second characteristic value of each pixel point in the multi-scale filtering image of each frame of filling liquid bottle visible light image to be detected. The second characteristic value is the difference between the three-channel component of each pixel point in the HSV image corresponding to the filtering image and the three-channel component mean value of all the pixel points in the HSV image corresponding to the filtering image, and the second characteristic value can be used for detecting and analyzing the attention degree of the pixel points subsequently.
(2-1-3) calculating the attention degree value of each pixel point in the multi-scale filtering image according to the first characteristic value and the second characteristic value of each pixel point in the multi-scale filtering image, wherein the calculation formula is as follows:
h i ＝s i 1 *s i 2
wherein h is i Is the attention range value, s, of the ith pixel point in the multi-scale filtered image i 1 Is the first characteristic value, s, of the ith pixel point in the multi-scale filtering image i 2 And the second characteristic value is the second characteristic value of the ith pixel point in the multi-scale filtering image.
It should be noted that the larger the first characteristic value and the second characteristic value of the ith pixel point in the multi-scale filtering image are, the higher the attention degree value of the pixel point is, that is, the more likely the pixel point is to be a pixel point of a foreign matter region in the visible light image of the filling liquid bottle to be detected, and the attention degree value refers to the possibility that each pixel point in the multi-scale filtering image is a foreign matter pixel point.
So far, the attention degree value of each pixel point in the multi-scale filtering image can be obtained by referring to the determination process of the attention degree value of the ith pixel point in the multi-scale filtering image. In addition, for convenience of calculation, according to the attention degree value of each pixel point in the multi-scale filtering image, the embodiment normalizes the attention degree value of each pixel point to make the attention degree value of each pixel point within the range of [0,1 ].
And (2-1-4) determining a characteristic enhancement map corresponding to the multi-scale filtering image of the visible light image of each frame of the filling liquid bottle to be detected according to the attention degree value of each pixel point in the multi-scale filtering image.
In order to improve the degree of significance of the foreign object region in the visible light image of the filling liquid bottle to be detected and highlight the feature information of the foreign object region, the attention degree value of each pixel point in the multi-scale filtering image is used as a feature enhancement value, and a feature enhancement image corresponding to the multi-scale filtering image of each frame of the visible light image of the filling liquid bottle to be detected is obtained according to the feature enhancement value of each pixel point and marked as T, so that the foreign object region in the visible light image of the filling liquid bottle to be detected can be detected later. It should be noted that the pixel value of the pixel point in each feature enhancement map refers to an attention degree value, that is, a feature enhancement value.
(2-2) according to the feature enhancement image corresponding to the multi-scale filtering image of each frame of visible light image of the filling liquid bottle to be detected, determining the optimal feature enhancement image from the feature enhancement image corresponding to the multi-scale filtering image, wherein the steps comprise:
and (2-2-1) determining a judgment factor of the characteristic enhanced image corresponding to the multi-scale filtering image according to the pixel value of each pixel point in the characteristic enhanced image corresponding to the multi-scale filtering image.
In order to improve the detection accuracy of the foreign matter detection of the filling liquid bottle, the feature enhancement image corresponding to the filter map with 3 scales is judged. Calculating a judgment factor of the feature enhanced image corresponding to the multi-scale filter image according to a pixel value of each pixel point in the feature enhanced image corresponding to the multi-scale filter image, namely the feature enhanced value of each pixel point, wherein the calculation formula is as follows:
wherein rho is a judgment factor of the characteristic enhanced image corresponding to the multi-scale filter graph, g i The pixel value of the ith pixel point in the multi-scale filter graph refers to a characteristic enhancement value, and K is the number of pixel points corresponding to the multi-scale filter graph.
It should be noted that the purpose of determining the determination factor of the feature-enhanced image is to determine whether the pixel value of each pixel in the feature-enhanced image can definitely approach to 0 or 1, so as to identify and segment the foreign object pixels in the foreign object region in the filtered image. For example, when the decision factor ρ of a certain feature-enhanced image is equal to 0, the pixel value of each pixel point in the feature-enhanced image may be close to 0 or close to 1.
And (2-2-2) according to the judgment factors of the feature enhanced images corresponding to the multi-scale filter maps, taking the feature enhanced image with the minimum judgment factor as the optimal feature enhanced image.
The reason why the feature enhanced image with the minimum determination factor is used as the optimal feature enhanced map in the present embodiment is that the feature enhanced map with the minimum determination factor has high purity, low image noise, and more prominent foreign object outline of the foreign object region, and the foreign object region is easy to extract and separate.
(3) And determining a binary image of the optimal characteristic enhancement image according to each pixel point and pixel value of the optimal characteristic enhancement image corresponding to each frame of filling liquid bottle visible image to be detected.
In this embodiment, after obtaining the optimal feature enhancement map corresponding to the visible light image of each frame of filling liquid bottle to be detected according to step (2-2), the foreign object region is extracted and segmented based on the optimal feature enhancement map, in order to avoid the influence of the fixed threshold on the extraction of the foreign object, the accuracy of image segmentation is improved, and the foreign object is extracted by setting the adaptive segmentation threshold, that is, determining the binary image of the optimal feature image, the steps include:
(3-1) constructing a sliding window with the size of C, wherein the size of C is C × C, the sliding window is set to 11 × 11, the sliding window slides on the optimal feature enhancement image, the sliding window which takes the pixel point as the center and corresponds to each pixel point in the optimal feature enhancement image is obtained, and then each neighborhood pixel point of each pixel point in the optimal feature enhancement image is obtained, and the neighborhood pixel points refer to other pixel points which do not contain the center pixel point in the sliding window area which takes each pixel point as the center. And calculating the segmentation threshold of the central pixel point of each sliding window in a self-adaptive manner according to each neighborhood pixel point of each pixel point in the optimal feature enhancement graph.
(3-2) calculating the segmentation threshold of the central pixel point of each sliding window in a self-adaptive manner, namely calculating the segmentation threshold corresponding to each pixel point in the optimal feature enhancement graph according to the position of each pixel point, the position of each pixel point corresponding to each neighborhood pixel point and the pixel value, wherein the pixel value refers to the feature enhancement value of each pixel point in the feature enhancement graph, and the calculation formula is as follows:
wherein e (x, y) is a segmentation threshold corresponding to each pixel point in the optimal feature enhancement map, C is a region of a sliding window corresponding to each pixel point in the optimal feature enhancement map and centered on the pixel point, Bz (i, j) is a pixel value of each pixel point in the optimal feature enhancement map corresponding to each neighborhood pixel point, (x, y) is a position of each pixel point in the optimal feature enhancement map, (i, j) is a position of each neighborhood pixel point of each pixel point in the optimal feature enhancement map, q is a bias term, and the bias term q can be set by an implementer and needs to be greater than zero, and the bias term q is set to 0.1 in this embodiment.
And (3-3) determining a binary image of the optimal feature enhancement map according to the segmentation threshold corresponding to each pixel point in the optimal feature enhancement map.
Establishing a segmentation model based on a segmentation threshold corresponding to each pixel point in the optimal feature enhancement image, and performing pixel-by-pixel judgment on each pixel point in the optimal feature image according to the segmentation model, namely performing segmentation processing on the optimal feature enhancement image so as to extract a foreign matter region in the optimal feature image, wherein the segmentation model is as follows:
wherein d (x, y) is the pixel value of the pixel point at the position (x, y) after the optimal feature enhancement map is subjected to the segmentation processing, Bz (x, y) is the pixel value of the pixel point at the position (x, y) in the optimal feature enhancement map, and e (x, y) is the segmentation threshold corresponding to the pixel point at the position (x, y) in the optimal feature enhancement map.
The optimal feature enhancement map after the segmentation processing is a binary image, so that a binary image of the optimal feature enhancement map corresponding to each frame of the visible light image of the filling liquid bottle to be detected is obtained. The binary image of the optimal feature enhancement map comprises a normal region and a foreign region, wherein pixel points in the normal region are marked as 0, and pixel points in the foreign region are marked as 1.
According to the method, the visible light image of the filling liquid bottle to be detected is subjected to feature enhancement processing based on the step (2), the significance of the outline of the foreign matter region in the visible light image of the filling liquid bottle to be detected is improved, the outline information of the foreign matter is enhanced, the foreign matter in the filling liquid bottle to be detected is detected and identified by utilizing the self-adaptive segmentation threshold value based on the step (3) through the feature enhancement image, the error introduced by the fixed threshold value is reduced, and the robustness is high for the visible light image of the filling liquid bottle to be detected with uneven illumination.
(4) And determining each foreign matter region in the visible light image of each frame of the filling liquid bottle to be detected according to the filtering image and the binary image corresponding to the optimal characteristic enhancement image.
In this embodiment, the filtered image of the optimal feature enhancement map corresponding to each frame of the visible light image of the filling liquid bottle to be detected is multiplied by the binary image, so that each foreign object region in each frame of the visible light image of the filling liquid bottle to be detected can be extracted. Because the filter image corresponding to the optimal characteristic enhancement image is an RGB image, the data of each foreign body region in the visible light image of the filling liquid bottle to be detected is also RGB image data, and the data of each foreign body region is used for further judging the foreign bodies in the filling liquid bottle to be detected, the detection precision of the real foreign bodies in the filling liquid bottle can be improved.
(5) And judging whether the real foreign matter area exists in the filling liquid bottle visible light image to be detected according to the number and the gray value of the pixel points in each foreign matter area in each frame of filling liquid bottle visible light image to be detected.
In this embodiment, for each foreign object region in the visible light image of each frame of filling liquid bottle to be detected obtained in step (4), it is considered that a small amount of small bubbles inevitably exist in the filling liquid bottle, and the bubbles may affect the detection of the foreign object in the filling liquid bottle, resulting in a high false detection rate. This embodiment will set up true foreign matter recognition model to reduce the influence of bubble etc. to the foreign matter testing result in the filling liquid bottle, the step of true foreign matter recognition model includes:
(5-1) determining the area and the number of edge pixel points corresponding to each foreign body region according to the number of the pixel points in each foreign body region in each frame of visible light image of the filling liquid bottle to be detected.
Firstly, count the number of pixel points in each foreign matter region in the filling liquid bottle visible light image that every frame is detected and is detected, according to the number of pixel points in each foreign matter region, calculate the area that each foreign matter region corresponds, be about to the area of pixel points number multiplication single pixel point in each foreign matter region, obtain the area that each foreign matter region corresponds, write as S a And a is 1,2, …, and a is the number of each foreign object region. In addition, the number of edge pixel points corresponding to each foreign object region is counted, that is, the perimeter of each foreign object region is counted and recorded as L a 。
(5-2) according to each pair of foreign matter regionsCorresponding area S a And the number of edge pixels L a And calculating the morphological index value of each foreign body area, wherein the calculation formula is as follows:
wherein, X a Is the morphological index value of the a-th foreign matter region, S a Is the area of the a-th foreign matter region, L a Is the perimeter of the a-th foreign matter region.
Referring to the determination process of the morphometric index value of the a-th foreign object region, the morphometric index value of each foreign object region can be obtained. It should be noted that the morphological index value can be used to preliminarily determine the bubble in the filling liquid bottle to be detected, and the larger the morphological index value is, the more the shape of the foreign object region corresponding to the morphological index value is close to a circle, the bubble region may be, otherwise, the more the foreign object region deviates from the circle. The process of calculating morphometric index values is prior art and is not within the scope of the present invention and will not be described in detail herein.
(5-3) judging whether a real foreign matter area exists in the visible light image of the filling liquid bottle to be detected according to the morphological index value of each foreign matter area, wherein the steps comprise:
(5-3-1) in this example, the form index threshold is denoted as X T ，X T When the morphological index value of any one of the foreign object regions is smaller than the morphological index threshold value, that is, X is 0.45 a <X T Judging the foreign body area as a real foreign body area, if the morphological index value of any one foreign body area is not less than the morphological index threshold value, namely X a ≥X T Then, the foreign object area is determined to be a suspected foreign object area. And according to the judgment result, each foreign matter area in the visible light image of the filling liquid bottle to be detected is subjected to primary identification so as to reduce the calculation amount of subsequent judgment and facilitate further analysis on the suspected foreign matter area.
(5-3-2) if the suspected foreign matter areas exist, determining the normal areas in the visible light images of the filling liquid bottles to be detected according to the filtering images and the binary images corresponding to the optimal feature enhancement images by referring to the step (4) in the process of determining the foreign matter areas in the visible light images of the filling liquid bottles to be detected in each frame.
(5-3-3) determining the credibility of each suspected foreign matter area as a real foreign matter area according to the gray value of each pixel point in each suspected foreign matter area in each frame of visible light image of the filling liquid bottle to be detected and the gray value of each pixel point in a normal area, wherein the credibility comprises the following steps:
(5-3-3-1) determining the pixel gray mean value corresponding to each suspected foreign matter region and each normal region according to the gray value of each pixel point in each suspected foreign matter region and the gray value of each pixel point in each normal region in the visible light image of each frame of the filling liquid bottle to be detected, and determining the pixel distribution function corresponding to each suspected foreign matter region and each normal region.
A. In this embodiment, the corresponding pixel gray level mean value of each suspected foreign matter region and each normal region can be obtained through the number of pixel points and the gray level value of each pixel point in each suspected foreign matter region, the number of pixel points in the normal region and the gray level value of each pixel point in each frame of visible light image of the filling liquid bottle to be detected.
B. In order to accurately identify each suspected foreign object region in the visible light image of each frame of filling liquid bottle to be detected, the present embodiment needs to acquire the distribution status of each suspected foreign object region. Forming a corresponding gray value sequence by gray values of the line pixel points in each suspected foreign matter area, fitting a pixel distribution function based on the gray value sequence corresponding to each suspected foreign matter area, wherein each suspected foreign matter area corresponds to one pixel distribution function and is marked as f (x), x is the gray value of the pixel points in each suspected foreign matter area, and fits a standard pixel distribution function by the gray values of the line pixel points in the normal area and is marked as f 0 (x) In that respect The process of fitting a distribution function according to a sequence is prior art and is not within the scope of the present invention and will not be described in detail here.
(5-3-3-2) determining the credibility of each suspected foreign matter area as a real foreign matter area according to the corresponding pixel gray level mean value of each suspected foreign matter area and the normal area and the corresponding pixel distribution function of each suspected foreign matter area and the normal area, wherein the calculation formula is as follows:
wherein, P a As the confidence level that the a-th suspected foreign object region is an actual foreign object,
is the gray level average value of the pixel corresponding to the a-th suspected foreign matter area,
the mean value of the pixel gray levels corresponding to the normal region, f a (x) Is the pixel distribution function corresponding to the a-th suspected foreign matter area, f 0 (x) The standard pixel distribution function corresponding to the normal area.
When the suspected foreign matter region is the bubble region, the pixel distribution in the bubble region is consistent with the pixel distribution in the normal region. In the calculation formula of the reliability that the suspected foreign matter region is the real foreign matter region,
the smaller the difference between the pixel gray level mean value corresponding to the a-th suspected foreign object area and the pixel gray level mean value corresponding to the normal area is,
the smaller the size is, the higher the similarity between the pixel distribution function corresponding to the a-th suspected foreign matter area and the standard pixel distribution function corresponding to the normal area is, that is, the lower the reliability that the a-th suspected foreign matter area is a real foreign matter is, the more likely the a-th suspected foreign matter area is to be a bubble area in a visible light image of the filling liquid bottle to be detected.
(5-3-4) this embodiment designates the reliability threshold as P T And P is T 0.65, if any one suspected foreign matter area is a real foreign matter areaIs greater than a confidence threshold, i.e. P a >P T If the reliability of any suspected foreign matter area as a real foreign matter area is not more than the reliability threshold, namely P a ≤P T Then the suspected foreign matter area is determined not to be the real foreign matter area. Based on the determination process, each suspected foreign matter area can be finally determined, and the accurate identification of the real foreign matters in the filling liquid bottle to be detected is realized.
Therefore, the real foreign matter region in the visible light image of each frame of filling liquid bottle to be detected is obtained, the real foreign matter region in the visible light image of each frame of filling liquid bottle to be detected is summarized and sorted, if the foreign matter region exists in the visible light image of any one or more frames of filling liquid bottle to be detected, the filling liquid bottle to be detected is judged to have the foreign matter, and reference opinions are provided for relevant detection personnel based on the final judgment result.
The method can identify the real foreign matter area based on the image data of the filling liquid bottle, determine the characteristic enhanced image of the filling liquid bottle through optical means and material test, extract the foreign matter area formed by pixel points of each foreign matter in the image of the filling liquid bottle, and after each foreign matter area is obtained, to avoid the influence of the bubble area and irrelevant factors in the image of the liquid bottle on the extracted foreign matter area, the method finely judges each foreign matter area, and realizes the identification of the real foreign matter area. The invention has the characteristics of small calculated amount, high detection speed and high foreign body detection precision.
The above-mentioned embodiments are only used for illustrating the technical solutions of the present application, and not for limiting the same; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; such modifications and substitutions do not substantially depart from the spirit and scope of the embodiments of the present application and are intended to be included within the scope of the present application.
Claims (10)
1. A method for detecting foreign matters in a bottle in a liquid filling process based on an optical means is characterized by comprising the following steps:
collecting N frames of filling liquid bottle visible light images to be detected, and determining multi-scale filtering images of each frame of filling liquid bottle visible light images to be detected according to each frame of filling liquid bottle visible light images to be detected;
determining a characteristic enhancement image corresponding to each multi-scale filtering image according to the multi-scale filtering image of each frame of filling liquid bottle visible light image to be detected, and further determining an optimal characteristic enhancement image from the characteristic enhancement images corresponding to the multi-scale filtering images;
determining a binary image of the optimal characteristic enhancement image according to each pixel point and pixel value of the optimal characteristic enhancement image corresponding to each frame of filling liquid bottle visible image to be detected;
determining each foreign matter area in each frame of visible light image of the filling liquid bottle to be detected according to the filtering image and the binary image corresponding to the optimal characteristic enhancement image;
and judging whether real foreign matters exist in the filling liquid bottle to be detected or not according to the number and the gray value of pixel points in each foreign matter area in each frame of visible light image of the filling liquid bottle to be detected.
2. The method for detecting the foreign matter in the liquid filling process bottle based on the optical means as claimed in claim 1, wherein the step of determining the feature enhancement map corresponding to the multi-scale filtering image comprises:
determining a first characteristic value of each pixel point in a multi-scale filtering image according to the gray value of each pixel point in the multi-scale filtering image of each frame of filling liquid bottle visible light image to be detected;
determining HSV images corresponding to the multi-scale filtering images according to the multi-scale filtering images of the visible light images of the filling liquid bottles to be detected in each frame, and further determining second characteristic values of pixel points in the multi-scale filtering images;
determining the attention degree value of each pixel point in the multi-scale filtering image according to the first characteristic value and the second characteristic value of each pixel point in the multi-scale filtering image;
and determining a characteristic enhancement map corresponding to the multi-scale filtering image of the visible light image of the filling liquid bottle to be detected according to the attention degree value of each pixel point in the multi-scale filtering image.
3. The method for detecting the foreign matter in the liquid filling process bottle based on the optical means as claimed in claim 2, wherein the calculation formula for determining the first characteristic value of each pixel point in the multi-scale filtering image is as follows:
wherein s is i 1 For the first feature value of the ith pixel point in the multi-scale filtered image,
the frequency of the gray value of the jth pixel except the gray value of the ith pixel point in the multi-scale filtering image appearing in the corresponding scale filtering image is I i Is the gray value, I, of the ith pixel point in the multi-scale filtering image j The exp () is an exponential function for the jth pixel gray value except the ith pixel gray value in the multi-scale filtering image.
The calculation formula for determining the second characteristic value of each pixel point in the multi-scale filtering image is as follows:
wherein s is i 2 Is the second characteristic value, H, of the ith pixel point in the multi-scale filtering image i Tone value S of ith pixel point in HSV image corresponding to multi-scale filtering image i Filtering image pairs for multiple scalesSaturation value V of ith pixel point in HSV image i The lightness value of the ith pixel point in the HSV image corresponding to the multi-scale filtering image,
the average value of the hue value of each pixel point in the HSV image corresponding to the multi-scale filtering image,
the average value of the saturation values of each pixel point in the HSV image corresponding to the multi-scale filtering image,
and averaging the brightness values of all pixel points in the HSV image corresponding to the multi-scale filtering image.
4. The method for detecting the foreign matter in the liquid filling process bottle based on the optical means as claimed in claim 1, wherein the step of determining the binary image of the optimal feature enhancement map comprises:
constructing a sliding window with the size of C, sliding the sliding window on the optimal feature enhancement graph, obtaining the sliding window which corresponds to each pixel point in the optimal feature enhancement graph and takes the pixel point as the center, and further obtaining each pixel point in the optimal feature enhancement graph and each neighborhood pixel point;
determining a segmentation threshold corresponding to each pixel point in the optimal feature enhancement graph according to the position of each pixel point, the position of each pixel point corresponding to each neighborhood pixel point and the pixel value;
and determining a binary image of the optimal feature enhancement map according to the segmentation threshold corresponding to each pixel point in the optimal feature enhancement map.
5. The method for detecting the foreign matter in the liquid filling process bottle based on the optical means as claimed in claim 4, wherein the calculation formula for determining the segmentation threshold corresponding to each pixel point in the optimal feature enhancement map is as follows:
wherein e (x, y) is a segmentation threshold corresponding to each pixel point in the optimal feature enhancement map, C is a region of a sliding window corresponding to each pixel point in the optimal feature enhancement map and centered on the pixel point, Bz (i, j) is a pixel value of each pixel point in the optimal feature enhancement map corresponding to each neighborhood pixel point, (x, y) is a position of each pixel point in the optimal feature enhancement map, (i, j) is a position of each neighborhood pixel point of each pixel point in the optimal feature enhancement map, and q is a bias term.
6. The method for detecting the foreign matter in the liquid filling process bottle based on the optical means as claimed in claim 1, wherein the step of determining the optimal feature enhancement image from the feature enhancement map corresponding to the multi-scale filtering image further comprises:
determining a judgment factor of the feature enhanced image corresponding to the multi-scale filter image according to the pixel value of each pixel point in the feature enhanced image corresponding to the multi-scale filter image;
and according to the judgment factor of the feature enhanced image corresponding to the multi-scale filter map, taking the feature enhanced image with the minimum judgment factor as the optimal feature enhanced image.
7. The method for detecting the foreign matter in the liquid filling process bottle based on the optical means as claimed in claim 6, wherein the calculation formula for determining the judgment factor of the feature enhanced image corresponding to the multi-scale filter map is as follows:
wherein rho is a judgment factor of the characteristic enhanced image corresponding to the multi-scale filter map, and g i For multi-scale filter mapsAnd K is the number of pixel points corresponding to the multi-scale filter graph.
8. The method for detecting the foreign matter in the liquid filling process bottle based on the optical means as claimed in claim 1, wherein the step of judging whether the filled liquid bottle to be detected has the real foreign matter comprises the following steps:
determining the area corresponding to each foreign matter region and the number of edge pixel points according to the number of the pixel points in each foreign matter region in each frame of visible light image of the filling liquid bottle to be detected;
determining the morphological index value of each foreign matter region according to the area corresponding to each foreign matter region and the number of edge pixel points;
and judging whether the visible light image of the filling liquid bottle to be detected has a real foreign matter area or not according to the morphological index value of each foreign matter area.
9. The method for detecting the foreign matters in the liquid filling process bottle based on the optical means as claimed in claim 8, wherein the step of judging whether the filled liquid bottle to be detected has the real foreign matters comprises the following steps:
if the morphological index value of any foreign body area is smaller than the morphological index threshold value, the foreign body area is judged to be a real foreign body area, and if the morphological index value of any foreign body area is not smaller than the morphological index threshold value, the foreign body area is judged to be a suspected foreign body area;
if the suspected foreign matter area exists, determining a normal area in each frame of visible light image of the filling liquid bottle to be detected according to the filtering image and the binary image corresponding to the optimal feature enhancement image;
determining the credibility of each suspected foreign matter area as a real foreign matter area according to the gray value of each pixel point in each suspected foreign matter area in each frame of visible light image of the filling liquid bottle to be detected and the gray value of each pixel point in the normal area;
if the reliability of any suspected foreign matter area as a real foreign matter area is greater than the reliability threshold, the suspected foreign matter area is determined to be a real foreign matter area, and if the reliability of any suspected foreign matter area as a real foreign matter area is not greater than the reliability threshold, the suspected foreign matter area is determined not to be a real foreign matter area.
10. The method of claim 9, wherein the step of determining the confidence level that each suspected foreign object region is a true foreign object region comprises:
determining a pixel gray mean value corresponding to each suspected foreign matter area and each normal area according to a gray value of each pixel point in each suspected foreign matter area and a gray value of each pixel point in each normal area in each frame of visible light image of the filling liquid bottle to be detected, and determining a standard pixel distribution function corresponding to each suspected foreign matter area and each normal area;
determining the credibility of each suspected foreign matter area as a real foreign matter area according to the corresponding pixel gray level mean value of each suspected foreign matter area and the normal area and the corresponding pixel distribution function of each suspected foreign matter area and the normal area, wherein the calculation formula is as follows:
wherein, P a As the confidence level that the a-th suspected foreign object region is an actual foreign object,
is the gray level average value of the pixel corresponding to the a-th suspected foreign matter area,
the mean value of the pixel gray levels corresponding to the normal region, f a (x) Is the pixel distribution function corresponding to the a-th suspected foreign matter area, f 0 (x) The standard pixel distribution function corresponding to the normal area.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210453932.6A CN114998205A (en) | 2022-04-24 | 2022-04-24 | Method for detecting foreign matters in bottle in liquid filling process based on optical means |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210453932.6A CN114998205A (en) | 2022-04-24 | 2022-04-24 | Method for detecting foreign matters in bottle in liquid filling process based on optical means |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114998205A true CN114998205A (en) | 2022-09-02 |
Family
ID=83025309
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210453932.6A Pending CN114998205A (en) | 2022-04-24 | 2022-04-24 | Method for detecting foreign matters in bottle in liquid filling process based on optical means |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114998205A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116071364A (en) * | 2023-03-29 | 2023-05-05 | 苏州飞搜科技有限公司 | Method for detecting foreign matters contained in liquid in filling syringe |
| CN116823835A (en) * | 2023-08-30 | 2023-09-29 | 山东省永星食品饮料有限公司 | Bottled water impurity detection method based on machine vision |
-
2022
- 2022-04-24 CN CN202210453932.6A patent/CN114998205A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116071364A (en) * | 2023-03-29 | 2023-05-05 | 苏州飞搜科技有限公司 | Method for detecting foreign matters contained in liquid in filling syringe |
| CN116823835A (en) * | 2023-08-30 | 2023-09-29 | 山东省永星食品饮料有限公司 | Bottled water impurity detection method based on machine vision |
| CN116823835B (en) * | 2023-08-30 | 2023-11-10 | 山东省永星食品饮料有限公司 | Bottled water impurity detection method based on machine vision |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116721106B (en) | Profile flaw visual detection method based on image processing | |
| Zhou et al. | Automated visual inspection of glass bottle bottom with saliency detection and template matching | |
| CN115082419B (en) | Blow-molded luggage production defect detection method | |
| CN108896574B (en) | Bottled liquor impurity detection method and system based on machine vision | |
| CN114998205A (en) | Method for detecting foreign matters in bottle in liquid filling process based on optical means | |
| CN109816644A (en) | A kind of bearing defect automatic checkout system based on multi-angle light source image | |
| CN109550712A (en) | A kind of chemical fiber wire tailfiber open defect detection system and method | |
| CN115082683A (en) | Injection molding defect detection method based on image processing | |
| CN108537099A (en) | A kind of licence plate recognition method of complex background | |
| CN105403147A (en) | Embedded bottle pre-form detection system and detection method | |
| CN113658131B (en) | Machine vision-based tour ring spinning broken yarn detection method | |
| CN109741314A (en) | A kind of visible detection method and system of part | |
| CN110070523B (en) | Foreign matter detection method for bottle bottom | |
| CN114972356A (en) | Plastic product surface defect detection and identification method and system | |
| CN114445707A (en) | Intelligent visual fine detection method for defects of bottled water labels | |
| CN114495098B (en) | Diaxing algae cell statistical method and system based on microscope image | |
| CN112581452B (en) | Industrial accessory surface defect detection method, system, intelligent equipment and storage medium | |
| CN106780437B (en) | A kind of quick QFN chip plastic packaging image obtains and amplification method | |
| CN114119603A (en) | Image processing-based snack box short shot defect detection method | |
| CN115082776A (en) | Electric energy meter automatic detection system and method based on image recognition | |
| CN109387524A (en) | Thread defect detection method and device based on linearly polarized photon | |
| CN117330582A (en) | Polymer PE film surface crystal point detecting system | |
| CN117237747A (en) | Hardware defect classification and identification method based on artificial intelligence | |
| CN108171691B (en) | Detection method of plastic container | |
| CN117036259A (en) | Metal plate surface defect detection method based on deep learning |
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 |