CN109884070B - Aluminum-aluminum blister packaging tablet defect detection method based on machine vision - Google Patents
Aluminum-aluminum blister packaging tablet defect detection method based on machine vision Download PDFInfo
- Publication number
- CN109884070B CN109884070B CN201910174967.4A CN201910174967A CN109884070B CN 109884070 B CN109884070 B CN 109884070B CN 201910174967 A CN201910174967 A CN 201910174967A CN 109884070 B CN109884070 B CN 109884070B
- Authority
- CN
- China
- Prior art keywords
- image
- tablet
- tablets
- template
- aluminum
- 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.)
- Active
Links
Images
Abstract
The invention discloses a machine vision-based aluminum-aluminum blister packaging tablet defect detection method, which comprises the following steps: establishing a tablet template and storing template parameters; acquiring an image to be detected and carrying out image preprocessing operation; preliminarily positioning all normal tablets in an image to be detected; and finally, template matching and defect type detection are carried out. The invention overcomes the problems of false detection and missed detection caused by the light-reflecting characteristic of the aluminum-aluminum bubble cap, realizes the automation of tablet detection, can accurately detect the abnormal situations of defects, particle shortage, multiple particles and the like of tablets with different shapes, improves the detection precision and speed of assembly line packaging, and ensures the quality of the tablets and the production efficiency of enterprises.
Description
Technical Field
The invention belongs to the technical field of blister package defect detection, and particularly relates to an aluminum-aluminum blister package tablet defect detection method based on machine vision.
Background
Blister packaging is one of the most widely used and most rapidly developed flexible packaging materials in the pharmaceutical industry, and the common packaging methods include two types: aluminum-plastic blister packs and aluminum-aluminum blister packs. Both are encapsulated with an aluminum foil layer on a substrate, except that the substrate of the aluminum-plastic blister is plastic and the substrate of the aluminum-aluminum blister is aluminum. The aluminum foil has excellent light-shielding and moisture-proof properties, can effectively protect the packaged medicine, and is widely applied to medicine packaging. The quality of the tablets is an important component of the pharmaceutical industry, and is directly related to the life health of patients and the brand image of pharmaceutical enterprises, and in the processes of processing, manufacturing and blister packaging, the tablets can have the problems of defects, particle shortage, multiple particles, foreign matters and the like, so that the realization of automatic detection of the defects of the tablets and accurate elimination of unqualified products are very important.
The tablet defect detection work in actual production generally has the characteristics of repeatability, continuity, high precision requirement and the like. The currently commonly used detection methods are as follows: 1) the manual removal method needs workers to monitor the packaged medicine plates on the conveyor belt all the time, and unqualified products are removed when the unqualified products are found. Because the manual removal method has high labor intensity, low detection efficiency and higher requirement on workers, the condition of product omission or false detection is easy to occur when the workers are in a fatigue state; 2) the quality weighing method is to weigh the medicine with the help of a high-precision weight measuring instrument, and compare the weighed medicine with the quality of normal medicine to judge whether an abnormal condition occurs. The method is simple, but is easily influenced by industrial environment and measurement precision, and the abnormal condition that the total mass is unchanged due to the simultaneous existence of the lack particles and the multiple particles cannot be detected; 3) the image processing method based on machine vision collects images of the tablet plate through an industrial CCD camera and transmits the images to an industrial personal computer for image processing, and defects in the tablet images are identified and eliminated.
The image processing method based on machine vision has good adaptability, can replace manpower with a machine, effectively reduces the labor intensity of workers, realizes quick automatic detection of tablet defects and accurate judgment of tablet quality, and is widely adopted by enterprises in the market. For the defect detection of the aluminum-plastic blister package, because the substrate is made of transparent plastic, the image is acquired by adopting a mode of placing a light source at the bottom and taking a picture by a camera at the top, the obtained tablet image has a pure background, and the defect detection is relatively easy by using an image processing algorithm. And to aluminium blister packaging defect detection, because of the aluminium foil does not have the light transmissivity, can only adopt the mode collection image that the side was placed light source, top camera and was shot, the aluminium foil base plate has certain reflectivity moreover, and the image that obtains of shooing has local highlight, illumination inequality, impurity interference scheduling problem, leads to appearing the condition of false retrieval or hourglass inspection easily for medicine quality has the risk. Therefore, the accurate real-time detection method for the defects of the aluminum-aluminum blister packaging tablets is a key problem in the field of machine vision.
Disclosure of Invention
The invention aims to provide a machine vision-based aluminum-aluminum blister packaging tablet defect detection method, aims to make up for the defects of the existing detection technology, solves the problems of easy false detection and missed detection caused by the easily-reflective characteristic of an aluminum substrate, and realizes the universality and real-time detection of tablets of different shapes and sizes.
In order to achieve the purpose, the technical scheme of the invention specifically comprises the following steps:
step 1, giving a completely normal template tablet image, wherein the shape of the tablet is not limited, preprocessing the template tablet image, calculating the center point coordinate of each tablet in the template and the connected domain area of each tablet after template image binarization, and storing all template parameters;
step 2, giving a tablet image to be detected of the same type, and carrying out image preprocessing operations such as cutting, binarization and the like on the image to obtain a binary image;
step 3, extracting connected domains from the binary image, calculating parameters such as the area and the central point coordinate of each connected domain, the frame coordinate of the minimum circumscribed rectangle, the compactness and the like, and realizing the primary positioning of the normal tablet;
and 4, matching the preliminarily positioned tablets with the tablets in the template, wherein the matched tablets are considered to be normal tablets, cutting the slices of each tablet according to the positions of the tablets which are not matched in the template, identifying the conditions of defects, lack of the tablets and multiple tablets by characteristic classification, and finally marking the defective tablets.
Further, the specific implementation manner of preprocessing the template tablet image in the step 1 is as follows,
step 1.1, size m0×n0Color template tablet image R0Conversion into a grayscale image G0In a gray scale image G0Taking the tablet plate size area as the interested area to obtain tablet plate image I with size of m × n0;
Step 1.2, obtaining tablet plate image I0Mean value μ, image I0Subtracting the mean value to obtain an image I1Obtaining image I by Otsu's method1For image I, η1Each pixel of (2) is furtherLine judgment shows that the value of 255 is larger than η and the value of 0 is smaller than η, and a binary image I with white tablet and black substrate is obtained2;
Step 1.3, for the binary image I2Performing morphological opening operation (corrosion first and expansion later) to obtain a result graph I after opening operation3。
Further, the specific implementation manner of obtaining all the template parameters in the template tablet image in step 1 is as follows,
step 1.4, for image I3Analyzing the connected domains, calculating the coordinates of the central points and the area information of all the connected domains in the image, and setting an area threshold η2Exclusion area less than threshold η2The rest connected domain is a normal tablet; finally, obtaining the total tablet number d of the template tablet image, and the central point coordinate of each tablet { (x)1,y1),(x2,y2),...,(xi,yi),...,(xd,yd) And connected domain area { S }1,S2,...,Si,...,SdAnd (5) calculating the maximum value S of the tablet area, wherein i is the serial number of a connected domain in the template tablet imagemaxAnd minimum value SminAnd saving all template parameters for later detection to call.
Further, in the step 3, connected domain analysis is performed on the binary image I in the step 2 to obtain an area of each connected domain, a center point coordinate and a frame coordinate of a minimum circumscribed rectangle, where the frame coordinate may be expressed as { x }L,yLW, h }, wherein (x)L,yL) Representing the coordinates of the top left corner of the minimum bounding rectangle, w and h represent the width and height of the rectangle, respectively.
Further, the compactness of each connected domain is calculated in step 3, and the calculation formula is as follows:
wherein j represents the serial number of the connected domain in the image to be detected, SjDenotes the area of the jth connected component, wjAnd hjWidth and height, w, of the smallest bounding rectangle of the jth connected domainj×hjDenotes the area of the minimum bounding rectangle, cjThe value range of the compactness parameter is (0, 1)]。
Further, the specific implementation manner of achieving the preliminary positioning of the normal tablet in the step 3 is as follows,
setting a compactness threshold T, and combining an area threshold S of the templatemaxAnd SminAll connected domains are screened to satisfy the condition (S)min≤Sj≤SmaxAnd c isjNot less than T) connected domain is judged as a possible normal tablet, and the coordinates of the center point of the connected domain meeting the conditions are stored into a set;
and calculating the distance between the central points of every two connected domains in the set, judging the tablets to be multiple if the distance is smaller than a set threshold, deleting the two connected domains from the set, and obtaining the tablets which are completely normal in the image to be detected.
Further, the specific implementation manner of the step 4 is as follows,
step 4.1, matching the completely normal tablets obtained in the step 3 with the tablets in the template, calculating the distance between every two tablets, and setting a distance threshold η4If the distance between the tablet in the template and the center point of a certain tablet in the image to be detected is smaller than the threshold value η4If the matching is successful, otherwise, the matching is failed; number of tablets N for which matching failed is recordedcountAnd coordinates of a center point of the tablet in the template;
step 4.2, if NcountIf the image R to be detected is a normal image, the tablet plate has no abnormal condition; if N is presentcountIf not, indicating that the tablet at the corresponding position in the image R to be detected is possibly defective, lack of particles and multiple particles;
step 4.3, for NcountUnder the condition that the matching is not 0, taking the coordinate of the central point of the tablet which fails to be matched as the center, cutting a slice with the size of k multiplied by k on the image to be detected, carrying out binarization on the slice and solving the area S of the maximum connected domain in the binary imagepatchIf S ispatch<SminIt is judged as defective or defective, if Spatch>SmaxJudging the tablets to be a plurality of tablets, otherwise, judging the tablets to be normal tablets which are missed in the initial positioning;
and 4.4, marking the tablet positions with the defects, the lack particles, the multiple particles and other problems detected in the step 4.3 on the image to be detected, and sending a rejection command to a subsequent PLC device of the administration machine equipment.
Compared with the prior art for detecting the defects of tablets packaged by the aluminum-aluminum blister, the invention has the following advantages and beneficial effects:
(1) the invention designs an image processing method based on machine vision to replace a manual elimination method, which not only realizes the automation of tablet detection and ensures the medicine quality to be more guaranteed, but also greatly improves the detection precision and speed.
(2) The algorithm designed by the invention is suitable for tablets with different shapes and sizes, and is a universal aluminum-aluminum blister packaging tablet defect detection method.
(3) The method can effectively reduce the influence caused by the light-reflecting characteristic of the aluminum-aluminum bubble cap by extracting the characteristics of area, compactness and the like, accurately detect abnormal situations such as tablet defect, lack of granules, multiple granules and the like, and has the characteristics of small calculated amount, high processing speed, simplicity and effectiveness.
Drawings
FIG. 1 is a general flow diagram of the present invention.
FIG. 2 is a flow chart of locating a normal pill in an image to be detected.
Detailed Description
In order to more clearly illustrate the objects, technical solutions and advantages of the present invention, the following description is further provided with reference to the accompanying drawings and examples. It is to be understood that the invention is not to be limited by the disclosure of the embodiments, but is to be controlled by the scope of the appended claims.
As shown in FIG. 1, the invention discloses a machine vision-based aluminum-aluminum blister packaging tablet defect detection method, which comprises the following specific steps:
step 1, establishing a tablet template and storing template parameters:
step 1.1, size m0×n0Color template image R0Conversion into a grayscale image G0Because the bottom plate for placing the tablet plate can be shot when the camera collects the picture, the influence of the bottom plate on the detection is reduced, and the gray level image G is obtained0Taking the tablet plate size area as the interested area to obtain tablet plate image I with size of m × n0;
Step 1.2, obtaining tablet plate image I0Mean value μ, image I0Subtracting the mean value to obtain an image I1The averaging operation can overcome the problem of uneven illumination of the tablet plate image caused by the light-reflecting characteristic of the aluminum foil material. Obtaining image I by Otsu's method1For image I, η1Is 255 when the number of pixels is greater than η and 0 when the number of pixels is less than η, and a binary image I with a white tablet and a black substrate is obtained2Note that the original image R is now0The locally highlighted portion in (1) may also be white;
step 1.3, for the binary image I2Performing morphological opening operation (corrosion first and expansion later), eliminating interference of impurities on the tablet plate, disconnecting the tablet from the highlight part on the aluminum substrate, smoothing the contour of the tablet, and obtaining a result graph I after opening operation3;
Step 1.4, for image I3Analyzing the connected domains, calculating the coordinates of the central points and the area information of all the connected domains in the image, and setting an area threshold η2Exclusion area less than threshold η2The connected domain (i.e. the local highlight part around the tablet on the aluminum substrate is removed), and the remaining connected domain is the normal tablet. Finally, obtaining the total tablet number d of the template image, and the central point coordinate of each tablet { (x)1,y1),(x2,y2),...,(xi,yi),...,(xd,yd) And connected domain area { S }1,S2,...,Si,...,SdAnd f, wherein i is the serial number of the connected domain in the template image. Obtaining the maximum value S of the tablet areamaxAnd minimum value SminStoring all template parameters for calling in subsequent detection;
step 2, obtaining an image to be detected and carrying out image preprocessing operation:
giving an image R to be detected of the same type as the template tablet, and performing image preprocessing operation consistent with the step 1.1-step 1.3 on the image R, wherein the image preprocessing operation comprises the steps of converting a color image into a gray image, cutting, removing an average value, and then performing binary segmentation and morphological open operation to obtain a binary image I;
step 3, as shown in fig. 2, the flow of initially positioning the normal tablet in the image to be detected is as follows:
step 3.1, carrying out connected domain analysis on the binary image I to obtain the area and the central point coordinate of each connected domain and the frame coordinate of the minimum circumscribed rectangle, wherein the frame coordinate can be expressed as { x }L,yLW, h }, wherein (x)L,yL) Representing the coordinates of the top left corner of the minimum bounding rectangle, w and h represent the width and height of the rectangle, respectively.
Step 3.2, calculating the compactness of each connected domain, wherein the calculation formula is as follows:
wherein j represents the serial number of the connected domain in the image to be detected, SjDenotes the area of the jth connected component, wjAnd hjWidth and height, w, of the smallest bounding rectangle of the jth connected domainj×hjDenotes the area of the minimum bounding rectangle, cjThe value range of the compactness parameter is (0, 1)]. The amount of tightness can be used to distinguish between a normal tablet and a high-light portion around the tablet, since the high-light portion around the tablet is usually in an irregular arc or elongated shape and the tightness is usually small, whereas the tablet is usually in a circular or oval shape and the tightness is close to 1.
Step 3.3, set the compactness threshold T (in this example, T is 0.6), combined with the area threshold S of the templatemaxAnd SminFor all connected domainsLine screening, satisfying the condition (S)min≤Sj≤SmaxAnd c isjNot less than T) connected domain is judged as a possible normal tablet, and the coordinates of the center point of the connected domain meeting the conditions are stored into a set;
step 3.4, considering the multiple particles that may occur during the manufacturing process (i.e. two or more tablets are placed in a single tablet groove), the eligible tablets in step 3.3 may contain multiple particles. And (3) calculating the distance between the central points of every two connected domains in the set, judging the connected domains to be multiple if the distance is smaller than a set threshold, deleting the two connected domains from the set in the step (3.3), and obtaining the rest of the tablets which are completely normal in the image to be detected. Let b denote the number of completely normal tablets in the image to be detected, and the coordinates of their center points can be expressed asWherein j is the serial number of the Chinese medicine tablet in the image to be detected;
and 4, template matching and defect type detection:
and 4.1, because the positions of the tablets of the same type in each tablet on the production line are relatively fixed, the positions of the normal tablets in the image to be detected are basically consistent with the positions of the corresponding tablets in the template. The completely normal tablet obtained in step 3.4Tablet with template { (x)1,y1),(x2,y2),...,(xi,yi),...,(xd,yd) Matching, and calculating the distance between every two tablets to obtain a distance matrix P with the size of d multiplied by b, wherein the expression is as follows:
wherein i is more than or equal to 1 and less than or equal to d, j is more than or equal to 1 and less than or equal to b
Wherein d represents the number of tablets in the template, b represents the number of normal tablets in the image to be detected (b is less than or equal to d), and pijIndicating the ith tablet and waiting tablet in the templateDetecting the distance between the jth tablet center points in the image, and calculating the formula as follows:
set distance threshold η4(η is provided in the present embodiment)4Equal to the width of the minimum bounding rectangle of the first tablet of the template image), the minimum value of each row of the distance matrix P is calculated, and the minimum value is compared with the distance threshold value. Let the minimum value in row i be piaA is 1. ltoreq. b, if pia<η4If so, the matching between the ith tablet in the template and the a-th tablet in the image to be detected is successful, otherwise, the matching is failed. The matching failure indicates that the tablet at the ith position in the template corresponding to the image R to be detected has defects, and the serial number i of the tablet which fails to be matched in the template and the total number N of the tablets which fail to be matched are recordedcount;
Step 4.2, if NcountIf the image R to be detected is a normal image, the tablet plate has no abnormal condition; if N is presentcountIf not, indicating that the tablet at the corresponding position in the image R to be detected is possibly defective, lack of particles and multiple particles;
step 4.3, for NcountCase other than 0, to match the tablet that failed to match with the coordinate (x) of the center point of the templatei,yi) As the center, cutting a slice with the size of k multiplied by k on an image to be detected, binarizing the slice and solving the area S of the maximum connected domain in the binary imagepatchIf S ispatch<SminIt is judged as defective or defective, if Spatch>SmaxIf the tablets are multiple, the tablets are judged to be normal tablets which are missed in the initial positioning.
And 4.4, drawing the tablets with defects, particles lack, multiple particles and other problems detected in the step 4.3 on the image to be detected by using a red frame, and sending an eliminating instruction to a subsequent PLC device of the administration machine equipment.
The method designed by the invention overcomes the problem that the aluminum substrate is easy to miss detection and miss detection due to the characteristic of easy light reflection, can detect tablets with different shapes and sizes, has certain universality, and improves the precision and the real-time performance of the method for detecting the defects of the tablets packaged by the aluminum-aluminum blister.
The specific embodiments described above are merely illustrative of the spirit of the invention. Various modifications or additions may be made to the described embodiments or alternatives may be employed by those skilled in the art without departing from the spirit or ambit of the invention as defined in the appended claims.
Claims (5)
1. An aluminum-aluminum blister package tablet defect detection method based on machine vision is characterized by comprising the following steps:
step 1, establishing a template tablet image, and preprocessing the template tablet image to obtain all template parameters in the template tablet image;
step 2, giving an image to be detected which is the same as the type of the template tablet, and preprocessing the image to be detected to obtain a binary image;
step 3, extracting connected domains of the binary image, calculating the area and the center point coordinate of each connected domain, the frame coordinate of the minimum circumscribed rectangle and the compactness parameter, and realizing the primary positioning of the normal tablet; the concrete implementation mode is as follows,
calculating the compactness of each connected domain in the step 3, wherein the calculation formula is as follows:
wherein j represents the serial number of the connected domain in the image to be detected, SjDenotes the area of the jth connected component, wjAnd hjWidth and height, w, of the smallest bounding rectangle of the jth connected domainj×hjDenotes the area of the minimum bounding rectangle, cjThe compactness parameter is obtained in the value range of (0, 1)];
Setting a compactness threshold T, and combining an area threshold S of the templatemaxAnd SminScreening all connected domains SmaxAnd SminRespectively the maximum value and the minimum value of the tablet area of the template, and satisfies the condition Smin≤Sj≤SmaxAnd c isjJudging the connected domain of more than or equal to T as a possible normal tablet, wherein j represents the serial number of the connected domain in the image to be detected, cjRepresenting the compactness parameter of the jth connected domain, and storing the center point coordinates of the connected domains meeting the conditions into a set;
calculating the distance between the central points of every two connected domains in the set, judging the connected domains to be multiple if the distance is smaller than a set threshold, deleting the two connected domains from the set, and obtaining the rest of the tablets which are completely normal in the image to be detected;
and 4, matching the preliminarily positioned tablets with the tablets in the template tablet image, wherein the matched tablets are considered to be normal tablets, cutting the slices of each tablet according to the positions of the tablets which are not matched in the template tablet image, identifying the conditions of defects, lack of the tablets and multiple tablets by characteristic classification, and finally marking the defective tablets.
2. The method for detecting the defects of the aluminum-aluminum blister package tablets based on the machine vision as claimed in claim 1, wherein the method comprises the following steps: the specific implementation manner of preprocessing the template tablet image in the step 1 is as follows,
step 1.1, size m0×n0Color template tablet image R0Conversion into a grayscale image G0In a gray scale image G0Taking the tablet plate size area as the interested area to obtain tablet plate image I with size of m × n0;
Step 1.2, obtaining tablet plate image I0Mean value μ, image I0Subtracting the mean value to obtain an image I1Obtaining image I by Otsu's method1For image I, η1Is 255 when the number of pixels is greater than η and 0 when the number of pixels is less than η, and a binary image I with a white tablet and a black substrate is obtained2;
Step 1.3, for the binary image I2Form of proceedingThe arithmetic operation of learning to open, namely firstly corroding and then expanding, and a result graph I after the operation of opening is obtained3。
3. The method for detecting the defects of the aluminum-aluminum blister package tablets based on the machine vision as claimed in claim 2, wherein: the specific implementation of obtaining all the template parameters in the template tablet image in step 1 is as follows,
step 1.4, for image I3Analyzing the connected domains, calculating the coordinates of the central points and the area information of all the connected domains in the image, and setting an area threshold η2Exclusion area less than threshold η2The rest connected domain is a normal tablet; finally, obtaining the total tablet number d of the template tablet image, and the central point coordinate of each tablet { (x)1,y1),(x2,y2),...,(xi,yi),...,(xd,yd) And connected domain area { S }1,S2,...,Si,...,SdAnd (5) calculating the maximum value S of the tablet area, wherein i is the serial number of a connected domain in the template tablet imagemaxAnd minimum value SminAnd saving all template parameters for later detection to call.
4. The method for detecting the defects of the aluminum-aluminum blister package tablets based on the machine vision as claimed in claim 1, wherein the method comprises the following steps: in the step 3, connected domain analysis is performed on the binary image I in the step 2 to obtain the area of each connected domain, the coordinates of the central point and the coordinates of the frame of the minimum circumscribed rectangle, wherein the coordinates of the frame can be expressed as { x }L,yLW, h }, wherein (x)L,yL) Representing the coordinates of the top left corner of the minimum bounding rectangle, w and h represent the width and height of the rectangle, respectively.
5. The method for detecting the defects of the aluminum-aluminum blister package tablets based on the machine vision as claimed in claim 1, wherein the method comprises the following steps: the specific implementation of the step 4 is as follows,
step 4.1, the completely normal tablets obtained in step 3 are placed in a templateMatching the tablets, calculating the distance between every two tablets, and setting a distance threshold value η4If the distance between the tablet in the template and the center point of a certain tablet in the image to be detected is smaller than the threshold value η4If the matching is successful, otherwise, the matching is failed; number of tablets N for which matching failed is recordedcountAnd coordinates of a center point of the tablet in the template;
step 4.2, if NcountIf the image R to be detected is a normal image, the tablet plate has no abnormal condition; if N is presentcountIf not, indicating that the tablet at the corresponding position in the image R to be detected is possibly defective, lack of particles and multiple particles;
step 4.3, for NcountUnder the condition that the matching is not 0, taking the coordinate of the central point of the tablet which fails to be matched as the center, cutting a slice with the size of k multiplied by k on the image to be detected, carrying out binarization on the slice and solving the area S of the maximum connected domain in the binary imagepatchIf S ispatch<SminIt is judged as defective or defective, if Spatch>SmaxJudging the tablets to be a plurality of tablets, otherwise, judging the tablets to be normal tablets which are missed in the initial positioning;
and 4.4, marking the tablet positions with the defects, the lack particles, the multiple particles and other problems detected in the step 4.3 on the image to be detected, and sending a rejection command to a subsequent PLC device of the administration machine equipment.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910174967.4A CN109884070B (en) | 2019-03-08 | 2019-03-08 | Aluminum-aluminum blister packaging tablet defect detection method based on machine vision |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910174967.4A CN109884070B (en) | 2019-03-08 | 2019-03-08 | Aluminum-aluminum blister packaging tablet defect detection method based on machine vision |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109884070A CN109884070A (en) | 2019-06-14 |
CN109884070B true CN109884070B (en) | 2020-06-09 |
Family
ID=66931316
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910174967.4A Active CN109884070B (en) | 2019-03-08 | 2019-03-08 | Aluminum-aluminum blister packaging tablet defect detection method based on machine vision |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109884070B (en) |
Families Citing this family (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110473197A (en) * | 2019-08-13 | 2019-11-19 | 树根互联技术有限公司 | Material surface defect detection method, device, equipment and storage medium |
CN110441319B (en) * | 2019-09-09 | 2022-05-03 | 凌云光技术股份有限公司 | Method and device for detecting appearance defects |
CN111008960B (en) * | 2019-11-22 | 2023-03-28 | 广州大学 | Aluminum electrolytic capacitor bottom appearance detection method and device based on machine vision |
CN111060519A (en) * | 2019-12-30 | 2020-04-24 | 研祥智能科技股份有限公司 | LED support defect judgment method and system |
CN111398308B (en) * | 2020-03-27 | 2023-01-17 | 上海健康医学院 | Automatic detection method and system for packaging quality of aluminum-plastic bubble caps of tablets and capsules |
CN111504198B (en) * | 2020-06-08 | 2022-04-22 | 无锡和博永新科技有限公司 | Method for detecting resistance cutting edge |
CN111896697A (en) * | 2020-07-31 | 2020-11-06 | 山东理工职业学院 | Integrated carbon nanotube electric sensor based on medicine monitoring and system thereof |
CN112345534B (en) * | 2020-10-30 | 2023-08-04 | 上海电机学院 | Defect detection method and system for particles in bubble plate based on vision |
CN112362673A (en) * | 2020-11-17 | 2021-02-12 | 清华大学天津高端装备研究院洛阳先进制造产业研发基地 | Visual detection method and system for dumplings |
CN112712496A (en) * | 2020-12-16 | 2021-04-27 | 广州大学 | Aluminum blister package pill defect detection method, system and storage medium |
CN115578380B (en) * | 2022-11-18 | 2023-03-17 | 菲特(天津)检测技术有限公司 | Universal bubble cap defect detection method based on machine vision |
CN116151691A (en) * | 2023-04-13 | 2023-05-23 | 山东一方制药有限公司 | Traditional Chinese medicine formula granule preparation quality supervision system based on artificial intelligence |
CN116698860B (en) * | 2023-08-08 | 2023-10-27 | 山东鲁地源天然药物有限公司 | Method for realizing mass solid root type traditional Chinese medicine slice quality analysis based on image processing |
CN117274251B (en) * | 2023-11-20 | 2024-02-06 | 山东鲁抗医药集团赛特有限责任公司 | Tablet quality detection method in medicine production process based on image data |
Family Cites Families (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
SG10201501672PA (en) * | 2015-03-05 | 2016-10-28 | Emage Vision Pte Ltd | Inspection of sealing quality in blister packages |
CN105021628B (en) * | 2015-07-19 | 2019-12-17 | 中北大学 | Method for detecting surface defects of optical fiber image inverter |
CN105894036B (en) * | 2016-04-19 | 2019-04-09 | 武汉大学 | A kind of characteristics of image template matching method applied to mobile phone screen defects detection |
CN106290382A (en) * | 2016-07-01 | 2017-01-04 | 广东技术师范学院 | Bubble-cap tablet package defective vision detection device and method |
CN107525808A (en) * | 2017-07-27 | 2017-12-29 | 佛山市南海区广工大数控装备协同创新研究院 | Blister medication classification and the online visible detection method of defect on a kind of production line |
JP6450815B1 (en) * | 2017-08-24 | 2019-01-09 | Ckd株式会社 | Appearance inspection device and blister packaging machine |
CN108805854B (en) * | 2018-01-09 | 2022-02-08 | 湖南科技大学 | Method for rapidly counting tablets and detecting completeness of tablets in complex environment |
CN108344743A (en) * | 2018-02-02 | 2018-07-31 | 佛山职业技术学院 | One kind being based on machine vision drug blister package defect inspection method and system |
CN108414533A (en) * | 2018-04-26 | 2018-08-17 | 德清炬诚电子科技有限公司 | Bubble-cap class packs drug visual detection equipment and method |
CN109297984B (en) * | 2018-11-13 | 2021-02-19 | 正大天晴药业集团股份有限公司 | Bubble cap packaging defect detection method, device and equipment |
-
2019
- 2019-03-08 CN CN201910174967.4A patent/CN109884070B/en active Active
Also Published As
Publication number | Publication date |
---|---|
CN109884070A (en) | 2019-06-14 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109884070B (en) | Aluminum-aluminum blister packaging tablet defect detection method based on machine vision | |
CN107402216B (en) | Film-coated product detection system and method | |
Sun et al. | Electric contacts inspection using machine vision | |
US5515159A (en) | Package seal inspection system | |
US20070052795A1 (en) | System and method for evaluating a machined surface of a cast metal component | |
CN106651857B (en) | A kind of printed circuit board patch defect inspection method | |
CN110441316B (en) | Battery surface defect detection method and detection system | |
CN111429408B (en) | Gold wire detection method for packaged chip | |
CN114372980A (en) | Industrial defect detection method and system | |
CN111766245A (en) | Button battery negative electrode shell defect detection method based on machine vision | |
JP5352615B2 (en) | Tablet inspection device and PTP packaging machine | |
CN111353986A (en) | Die-cutting product quality detection method | |
TWI558999B (en) | Defect inspection method and apparatus thereof | |
JP2004020254A (en) | Inspecting apparatus for transparent electrode film substrate | |
US11176419B1 (en) | Method for labeling image | |
CN111833350A (en) | Machine vision detection method and system | |
CN116205860A (en) | Wine bottle packaging defect detection system based on machine vision | |
CN114608458B (en) | Device and method for detecting thickness of die attach adhesive | |
CN116008294A (en) | Key cap surface particle defect detection method based on machine vision | |
JP2016102717A (en) | Method for inspecting screw | |
JP5692790B2 (en) | Tablet appearance inspection device and PTP packaging machine | |
CN102914546A (en) | System and method for detecting mist spray tank covers on the basis of machine vision | |
CN113793321A (en) | Casting surface defect dynamic detection method and device based on machine vision | |
KR102196396B1 (en) | A method for detecting surface defects of variable display panel glass | |
JPH07239940A (en) | Method for processing closed loop image and defect inspection method using the 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 |