CN114529555A - Image recognition-based efficient cigarette box in-and-out detection method - Google Patents
Image recognition-based efficient cigarette box in-and-out detection method Download PDFInfo
- Publication number
- CN114529555A CN114529555A CN202111681201.9A CN202111681201A CN114529555A CN 114529555 A CN114529555 A CN 114529555A CN 202111681201 A CN202111681201 A CN 202111681201A CN 114529555 A CN114529555 A CN 114529555A
- Authority
- CN
- China
- Prior art keywords
- image
- smoke box
- gradient
- smoke
- gray
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 26
- 235000019504 cigarettes Nutrition 0.000 title claims abstract description 24
- 239000000779 smoke Substances 0.000 claims abstract description 88
- 230000007547 defect Effects 0.000 claims abstract description 22
- 238000005259 measurement Methods 0.000 claims description 11
- 238000000034 method Methods 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 7
- 238000001914 filtration Methods 0.000 claims description 6
- 230000000877 morphologic effect Effects 0.000 claims description 4
- 230000003628 erosive effect Effects 0.000 claims 1
- 230000005540 biological transmission Effects 0.000 abstract description 3
- 241000208125 Nicotiana Species 0.000 abstract description 2
- 235000002637 Nicotiana tabacum Nutrition 0.000 abstract description 2
- 230000000903 blocking effect Effects 0.000 abstract description 2
- 230000002596 correlated effect Effects 0.000 abstract 1
- 230000000875 corresponding effect Effects 0.000 abstract 1
- 230000007306 turnover Effects 0.000 description 3
- 230000009466 transformation Effects 0.000 description 2
- 238000005452 bending Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000007797 corrosion Effects 0.000 description 1
- 238000005260 corrosion Methods 0.000 description 1
- 230000001934 delay Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000003708 edge detection Methods 0.000 description 1
- 238000005530 etching Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000006467 substitution reaction Methods 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
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06K—GRAPHICAL DATA READING; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
- G06K17/00—Methods or arrangements for effecting co-operative working between equipments covered by two or more of main groups G06K1/00 - G06K15/00, e.g. automatic card files incorporating conveying and reading operations
- G06K17/0022—Methods or arrangements for effecting co-operative working between equipments covered by two or more of main groups G06K1/00 - G06K15/00, e.g. automatic card files incorporating conveying and reading operations arrangements or provisious for transferring data to distant stations, e.g. from a sensing device
- G06K17/0025—Methods or arrangements for effecting co-operative working between equipments covered by two or more of main groups G06K1/00 - G06K15/00, e.g. automatic card files incorporating conveying and reading operations arrangements or provisious for transferring data to distant stations, e.g. from a sensing device the arrangement consisting of a wireless interrogation device in combination with a device for optically marking the record carrier
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
- G06Q10/087—Inventory or stock management, e.g. order filling, procurement or balancing against orders
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/20—Image enhancement or restoration using local operators
-
- 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
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/155—Segmentation; Edge detection involving morphological operators
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02W—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO WASTEWATER TREATMENT OR WASTE MANAGEMENT
- Y02W90/00—Enabling technologies or technologies with a potential or indirect contribution to greenhouse gas [GHG] emissions mitigation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Business, Economics & Management (AREA)
- Economics (AREA)
- Quality & Reliability (AREA)
- Entrepreneurship & Innovation (AREA)
- Strategic Management (AREA)
- Finance (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Geometry (AREA)
- Computer Networks & Wireless Communication (AREA)
- General Engineering & Computer Science (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a high-efficiency cigarette box in-and-out detection method based on image recognition, which belongs to the technical field of tobacco logistics. The YCbCr format Y component gray level image is processed to generate an edge-detected binary image, appearance defects of the smoke box are judged, detection results of the appearance defects of the smoke box are correlated with corresponding bar code information, the bar code information of the smoke box and the appearance defect information of the smoke box are integrated, the problem that network transmission delay and cloud computing are prone to blocking is solved, the detection efficiency of the smoke box entering and exiting a warehouse is improved, and the detection accuracy of the smoke box entering and exiting the warehouse is improved.
Description
Technical Field
The invention belongs to the technical field of tobacco logistics, and particularly relates to a high-efficiency cigarette box in-and-out detection method based on image recognition.
Background
Cigarette case business turn over storehouse link in finished product cigarette commodity circulation garden is lower because of automatic level, often can produce the bar code and leak the inspection, the problem such as the broken damage of box and wrong tablet loads in mixture appears, cigarette case bar code identification and cigarette case appearance imperfections discernment are accomplished by different equipment respectively in the current cigarette case business turn over storehouse system, cigarette case image acquisition is through different cameras and sweep a yard equipment, the camera reaches and sweeps unable direct intercommunication between yard equipment, the identification result need be uploaded to high in the clouds server, match cigarette case bar code and cigarette case appearance imperfections through high in the clouds server, if network transmission delays appear in the actual production process, the special condition that high in the clouds data processing easily blocks up, can influence the normal business turn over storehouse of cigarette case.
Disclosure of Invention
The invention aims to identify the bar code of the smoke box and judge the appearance defect of the smoke box through the images of the surfaces of three sides of the smoke box, and the detection result of the appearance of the smoke box is optically connected with the corresponding bar code information and outputs the information of the smoke box entering and exiting the warehouse.
In order to achieve the above object, the present invention is achieved by: an efficient smoke box in-and-out detection method based on image recognition comprises the following steps: step 1, obtaining images of three side surfaces of a smoke box; step 2, processing the surface image of the smoke box and identifying a bar code of the smoke box; step 3, processing the surface image of the smoke box and judging the appearance defects of the smoke box; and 4, associating the detection result of the appearance defect of the smoke box with corresponding bar code information, and outputting the information of the smoke box entering and exiting the warehouse.
Further, the step 2 comprises: step 2-1, acquiring a surface image of the smoke box, performing gray level processing, and converting the surface image of the smoke box into a gray level image; step 2-2, obtaining gradient amplitude representation of the gray scale image in the horizontal and vertical directions by using a Scharr operator for the gray scale image; step 2-3, subtracting the y gradient from the x gradient to obtain images with high horizontal gradient and low vertical gradient; step 2-4, carrying out average blurring on the gradient map by using a 9 x 9 kernel, carrying out binarization processing on the image after the average blurring, setting the gray value of the pixel point with the gray value less than or equal to 255 as 0, and setting the gray value of the pixel point with the gray value greater than 255 as 255; step 2-5, performing morphological operation on the binary image to eliminate gaps; and 2-6, searching the maximum contour, fitting a contour rectangle, determining a bar code area and identifying, comparing the bar code data on three identified sides to judge whether the cigarette box license plate number is wrong, and 2-7, if the bar code area cannot be identified, restoring the motion blurred image and then identifying again.
Further, the specific steps of the steps 2 to 5 are to construct a rectangular kernel, the width of the kernel is larger than the length of the kernel, 4 times of expansion is carried out after 4 times of corrosion, and small spots in the image are removed.
Further, the concrete steps of steps 2 to 7 are to obtain the conveying speed of the conveying belt, determine the blurring length L and the blurring angle theta according to the conveying speed, obtain a point spread function, restore the motion blurred image, and identify the restored image.
Further, in step 2-2, the Scharr operator, the x-direction convolution kernel formula and the y-direction convolution kernel formula are as follows:
further, the step 3 comprises: step 3-1, reading the surface image of the smoke box, converting the RGB image into a YCbCr format, outputting a Y component of the YCbCr image, and converting the picture into a gray scale image; step 3-2, performing Gaussian filtering processing on the gray level image; 3-3, carrying out image gradient detection on the gray level image by using a Sobel operator; and 3-4, judging whether the detection results of the 3-direction images of the smoke box are all rectangular, simultaneously comparing the image characteristics with the image characteristics of the standard complete smoke box, if the detection results are all rectangular and the comparison results are qualified, indicating that the smoke box is complete, and otherwise, judging that the smoke box has appearance defects.
Further, in step 3-3, the Sobel operator, the x-direction convolution kernel formula and the y-direction convolution kernel formula are as follows:
further, the step 3-4 includes, 3-4-1, calculating a depth measurement error value Δi3-4-2, measuring error value Δ in depthiEstablishing a compensation value sequence tableConstructing a measured distance sequence table from the compensation value sequence table3-4-3, carrying out weighted average calculation on the compensation values to obtain pixel-based weighted compensation values delta d (u, v), and 3-4-4, carrying out depth measurement error compensation on the vertexes of the profile initial graph: zAfter compensation(u, v) + Δ d (u, v), step 3-4-5, generating a smoke box profile map from the compensated vertex coordinates.
The invention has the beneficial effects that: and respectively converting the surface images of three sides of the smoke box acquired at one time into a traditional grey-scale image and a YCbCr format Y component grey-scale image, processing the traditional grey-scale image, identifying the smoke box bar code, and judging whether the smoke box bar code is wrong. The YCbCr format Y component gray level image is processed to generate an edge-detected binary image, the appearance defect of the smoke box is judged, the detection result of the appearance defect of the smoke box is associated with corresponding bar code information, the bar code information of the smoke box and the appearance defect information of the smoke box can be integrated through one set of image acquisition module and image processing module, the problem that network transmission delay and cloud computing are prone to blocking is solved, the detection efficiency of the smoke box entering and exiting a warehouse is improved, and the detection accuracy of the smoke box entering and exiting the warehouse is improved.
Drawings
FIG. 1 is a flow chart of the steps of the present invention;
fig. 2 is a schematic diagram of depth error compensation.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings to facilitate understanding of the skilled person.
An image recognition-based efficient cigarette box in-and-out detection method comprises the following steps:
Step 2, processing the surface image of the smoke box and identifying a bar code of the smoke box;
further, the step 2 comprises: step 2-1, acquiring a surface image of the smoke box, performing gray level processing, and converting the surface image of the smoke box into a gray level image; specifically, the surface image of the smoke box is a color image, and the color image is converted into a gray formula according to RGB: gray 0.299+ G0.587 + B0.114, the surface image of the smoke box was converted to a grayscale.
Step 2-2, obtaining gradient amplitude representation of the gray scale image in the horizontal and vertical directions by using a Scharr operator for the gray scale image; specifically, the gray-scale image obtained in the step 2-1 and an x-direction convolution kernel G1xPerforming convolution to obtain a horizontal gradient map, formulaThe expression is as follows:
convolution kernel G of the gray level image obtained in the step 2-1 and the Y direction1yAnd (3) performing convolution to obtain a vertical direction gradient map, wherein the formula is expressed as follows:
and 2-3, subtracting the y gradient from the x gradient to obtain an image with a high horizontal gradient and a low vertical gradient. As specified by Scharr: the gradient map obtained by subtracting the vertical gradient map y from the horizontal gradient map x-gradient is an image with high horizontal gradient and low vertical gradient, so that the barcode region in the image is more obvious.
And 2-4, carrying out average blurring on the gradient map by using a 9-by-9 kernel, carrying out binarization processing on the image after the average blurring, setting the gray value of the pixel point with the gray value less than or equal to 255 as 0, and setting the gray value of the pixel point with the gray value greater than 255 as 255.
Step 2-5, performing morphological operation on the binary image to eliminate gaps; preferably, the specific steps 2 to 5 are to construct a rectangular kernel, the width of the kernel is larger than the length of the kernel, and after 4 times of etching, 4 times of expansion are carried out to remove small spots in the image. Specifically, a rectangular kernel is constructed by using the cv2.getstructuringelement, the width of the kernel is larger than the length of the kernel, so that gaps between vertical bars in the bar code can be eliminated, morphological operations are performed, and the kernel obtained in the previous step is applied to a binary image so as to eliminate the gaps between vertical bars.
And 2-6, searching the maximum outline, fitting an outline rectangle, determining a bar code area, identifying, comparing the bar code data on the three identified sides, and judging whether the cigarette box license plate number is wrong.
And 2-7, if the barcode region cannot be identified, restoring the motion blurred image and then identifying again. The cigarette box moving speed is high in the cigarette box conveying process, motion blurred images can be generated, and the intervals of bar codes in the motion blurred images are changed, so that the cigarette box cannot be identified. The conveying speed of the conveying belt is in different grades of high, medium and low, the fuzzy length L and the fuzzy angle theta of the different grades are fixed values, the light spot length and the angle of the high, medium and low grade image are obtained through experiments, the light spot length corresponds to the fuzzy length L, the light spot angle corresponds to the fuzzy angle theta, and a point spread function h (x, y) is obtained:
obtaining an original image through a point spread function:
g(x,y)=f(x,y)*h(x,y)
g (x, y) is the original barcode image, f (x, y) is the acquired image, h (x, y) is the point spread function, and x is the convolution.
And 3, processing the surface image of the smoke box and judging the appearance defects of the smoke box.
And 3-1, reading the surface image of the smoke box, converting the RGB image into a YCbCr format, outputting a Y component of the YCbCr image, and converting the picture into a gray-scale image. Specifically, YCbCr is a color space, which is composed of ternary Y, Cb and Cr, and the RGB to YCbCr image calculation formula is:
y represents brightness and concentration of the image, and Y component transformation of the YCbCr image can detect more tiny details than the conventional gray-scale image transformation, so that the accuracy of judging the appearance defects of the smoke box is improved.
And 3-2, performing Gaussian filtering processing on the gray level image. The small details of the image after the Y component of the YCbCr image is transformed are all displayed, which results in large noise in the image, and the image noise needs to be further eliminated by a filtering method. The principle of the Gaussian filtering algorithm is that firstly, a mask template is used for scanning all pixel points in an image, then weighted average operation is carried out on pixel values in a mask, and the gray value of the pixel points in the center of the mask is replaced by the operation result, the mask template is selected as a 3 x 3 template, and the mathematical calculation formula is as follows:
in the formula, g (x, y) represents the original pixel gray scale value of the (x, y) point, and f (x, y) represents the gray scale value of the (x, y) point after the gaussian filtering processing.
Step 3-3, carrying out image gradient detection on the gray image by using a Sobel operator, specifically, carrying out image gradient detection on the gray image obtained in the step 3-2 and the Sobel operator G2xConvolution is carried out to obtain a horizontal gradient map, and the formula is as follows:
3-2, obtaining the gray-scale image and Sobel operator G2yPerforming convolution to obtain a vertical gradient map, wherein the formula is as follows:
Comparing threshold values, namely, bending the size of the gradient value of the wind by a threshold setting method, wherein pixel points with the gradient value G larger than the threshold T are points with obvious gray change, the gradient is larger, the gray value of the points is set to be 0, the points are displayed as black, the points are marked, the gray value of the pixel points with the gradient value smaller than the threshold is set to be 255, the pixel points are displayed as white, and a binary image for edge detection is formed, wherein the formula is as follows:
the binary image is a three-side rectangular outline of the smoke box.
And 3-4, carrying out depth measurement error compensation on the vertex of the profile initial graph, and generating a smoke box profile graph according to the compensated vertex coordinates. Because the distance between the cameras arranged on the three sides of the smoke box and the smoke box is different, the depth data obtained under different measuring distances have obviously different errors, and the error of the depth measurement is increased along with the increase of the distance. Finally, the three-side contour dimension is obviously different and cannot be matched.
Said step 3-4 comprises, 3-4-1, calculating a depth measurement error value ΔiSetting the real depth value of a certain plane in space from the camera to be RiMeasuring the depth value of the plane as Z by using a cameraiThen the depth measurement error value ΔiCan be expressed as:
Δi=Ri-Zi(i=1,2,…,k)
3-4-2, measuring error value delta in depthiEstablishing a compensation value sequence tableIn the measuring distance range of the camera, K marking space planes are selected at equal intervals on average, and the depth measuring error value delta obtained by the formulaiOn the basis of which a sequence table of compensation values for depth error compensation is establishedConstructing a measured distance sequence table from the compensation value sequence table
And 3-4-3, carrying out weighted average calculation on the compensation values to obtain pixel-based weighted compensation values delta d (u, v). Fig. 2 is a schematic view of depth error compensation, and a weighted average calculation is performed on the compensation value to compensate the camera depth measurement error value in the ordered space. Let Z (u, v) be the depth value of a certain pixel in the depth image before compensation, which is in the sequence table of the measured distances Z (u, v) and Z (v)i+1(u, v) one neighborIn the field, the threshold is set as S (S is 1,2,3, …), and Z in the lookup table is calculated by the following formulaiAnd Zi+1The euclidean distance of all points in the sxs neighborhood of (c) to Z:
wherein m is more than or equal to u-s and less than or equal to u + s, n is more than or equal to v-s and less than or equal to v + s, Xi(m, n) and Yi(m, n) is ZiThe abscissa and ordinate, X, of the point at which (m, n) is locatedi+1(m, n) and Yi+1(m, n) is Zi+1The abscissa and ordinate of the point where (m, n) is located, and x (u, v) and y (u, v) are the abscissa and ordinate of the point where z (u, v) is located.
The calculation formula for deriving the pixel-weighted compensation value is:
and 3-4-4, performing depth measurement error compensation on the vertex of the profile initial graph: zAfter compensation(u,v)=Z(u,v)+Δd(u,v)
And 3-4-5, generating a smoke box contour map according to the compensated vertex coordinates. The compensated vertex coordinate is not influenced by the installation distance of the camera, so that the difference of the contour sizes of three sides is avoided, and the false alarm of the appearance defect detection of the smoke box is reduced.
And 3-5, judging whether the outline graphs of the three side smoke boxes of the smoke box are all rectangular, simultaneously comparing the image characteristics with the standard integral image characteristics of the smoke box, if the outline graphs are all rectangular and the comparison result is qualified, indicating that the smoke box is integral, and otherwise, judging that the smoke box has appearance defects.
And 4, associating the detection result of the appearance defect of the smoke box with corresponding bar code information, outputting the information of the smoke box entering and exiting the warehouse, recording and conveying qualified smoke boxes, and eliminating the appearance defect and the wrong-brand smoke box.
Specifically, the cigarette box appearance defect detection result obtained in the step 3 is associated with corresponding bar code information, the information of the cigarette box entering and exiting the warehouse is output to an entering and exiting warehouse management system and sorting and removing equipment, qualified cigarette boxes are recorded and conveyed, and appearance defects and misbranded cigarette boxes are removed.
The above description is only an embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that are not thought of through the inventive work should be included in the scope of the present invention.
Claims (8)
1. An image recognition-based smoke box efficient warehouse-in and warehouse-out detection method is characterized by comprising the following steps: step 1, obtaining images of three side surfaces of a smoke box; step 2, processing the surface image of the smoke box and identifying a bar code of the smoke box; step 3, processing the surface image of the smoke box and judging the appearance defects of the smoke box; and 4, associating the detection result of the appearance defect of the smoke box with corresponding bar code information, outputting the information of the smoke box entering and exiting the warehouse, recording and conveying qualified smoke boxes, and eliminating the appearance defect and the wrong-brand smoke box.
2. The method for detecting the efficient entering and exiting of the smoke box based on the image recognition as claimed in claim 1, wherein the step 2 comprises: step 2-1, acquiring a surface image of the smoke box, performing gray level processing, and converting the surface image of the smoke box into a gray level image; step 2-2, obtaining gradient amplitude representation of the gray scale image in the horizontal and vertical directions by using a Scharr operator for the gray scale image; step 2-3, subtracting the y gradient from the x gradient to obtain images with high horizontal gradient and low vertical gradient; step 2-4, carrying out average blurring on the gradient map by using a 9 x 9 kernel, carrying out binarization processing on the image after the average blurring, setting the gray value of the pixel point with the gray value less than or equal to 255 as 0, and setting the gray value of the pixel point with the gray value greater than 255 as 255; step 2-5, performing morphological operation on the binary image to eliminate gaps; step 2-6, searching a maximum outline, fitting an outline rectangle, determining a bar code area, identifying, comparing the bar code data on three identified sides, and judging whether the cigarette box license plate number is wrong; and 2-7, if the barcode region cannot be identified, restoring the motion blurred image and then identifying again.
3. The image recognition-based smoke box efficient warehouse entry and exit detection method according to claim 2, characterized in that the specific steps of steps 2-5 are to construct a rectangular kernel, the width of the kernel is larger than the length of the kernel, 4 times of erosion is performed, then 4 times of expansion is performed, and small spots in the image are removed.
4. The method as claimed in claim 2, wherein the steps 2-7 include obtaining a conveying speed of a conveyor belt, determining a blurring length L and a blurring angle θ according to the conveying speed, obtaining a point spread function, restoring a motion-blurred image, and recognizing the restored image.
6. the method for detecting the efficient entering and exiting of the smoke box based on the image recognition as claimed in claim 1, wherein the step 3 comprises: step 3-1, reading the surface image of the smoke box, converting the RGB image into a YCbCr format, outputting a Y component of the YCbCr image, and converting the picture into a gray scale image; step 3-2, performing Gaussian filtering processing on the gray level image; 3-3, performing image gradient detection on the gray level image by using a Sobel operator to generate a profile initial image; 3-4, performing depth measurement error compensation on the vertex of the profile initial graph, and generating a smoke box profile graph according to the compensated vertex coordinates; and 3-5, judging whether the outline graphs of the three side smoke boxes of the smoke box are all rectangular, simultaneously comparing the image characteristics with the standard integral image characteristics of the smoke box, if the outline graphs are all rectangular and the comparison result is qualified, indicating that the smoke box is integral, and otherwise, judging that the smoke box has appearance defects.
8. the method as claimed in claim 6, wherein the step 3-4 includes, 3-4-1, calculating a depth measurement error value Δi3-4-2, measuring error value Δ in depthiEstablishing a compensation value sequence tableConstructing a measured distance sequence table from the compensation value sequence table3-4-3, carrying out weighted average calculation on the compensation values to obtain pixel-based weighted compensation values delta d (u, v), and 3-4-4, carrying out depth measurement error compensation on the vertexes of the profile initial graph: zAfter compensation(u, v) + Δ d (u, v), step 3-4-5, generating a smoke box profile map from the compensated vertex coordinates.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111681201.9A CN114529555A (en) | 2021-12-31 | 2021-12-31 | Image recognition-based efficient cigarette box in-and-out detection method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111681201.9A CN114529555A (en) | 2021-12-31 | 2021-12-31 | Image recognition-based efficient cigarette box in-and-out detection method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN114529555A true CN114529555A (en) | 2022-05-24 |
Family
ID=81620993
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202111681201.9A Pending CN114529555A (en) | 2021-12-31 | 2021-12-31 | Image recognition-based efficient cigarette box in-and-out detection method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114529555A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114882370A (en) * | 2022-07-07 | 2022-08-09 | 西安超嗨网络科技有限公司 | Intelligent commodity identification method and device, terminal and storage medium |
Citations (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20090316009A1 (en) * | 2008-06-20 | 2009-12-24 | Atsushi Ito | Apparatus, Method, and Program for Processing Image |
CN105913395A (en) * | 2016-04-10 | 2016-08-31 | 中国人民解放军海军航空工程学院青岛校区 | Moving object observation and fuzzy restoration method |
CN106650546A (en) * | 2016-10-20 | 2017-05-10 | 国家电网公司 | Method for automatically identifying substation switchgear equipment object based on two-dimensional code |
CN107025639A (en) * | 2017-04-05 | 2017-08-08 | 中科微至智能制造科技江苏有限公司 | A kind of Bar code positioning method under complex environment |
CN108620343A (en) * | 2018-04-09 | 2018-10-09 | 河南中烟工业有限责任公司 | A kind of smoke box appearance quality detection method based on CCD |
CN208000278U (en) * | 2018-01-11 | 2018-10-23 | 河南中烟工业有限责任公司 | The smoke box appearance delection device recycled for smoke box |
KR20180128299A (en) * | 2017-05-23 | 2018-12-03 | (주) 루리텍 | Distance measuring error calibration apparatus for distance measuring camera |
CN109807071A (en) * | 2018-09-18 | 2019-05-28 | 红塔烟草(集团)有限责任公司 | A kind of smoke box appearance quality detection and offline checking system and its method |
CN110427792A (en) * | 2019-06-24 | 2019-11-08 | 广西中烟工业有限责任公司 | A kind of recognition matching process blocking two dimensional code and its application |
CN112435190A (en) * | 2020-11-24 | 2021-03-02 | 同济大学 | Motion fuzzy target detection method based on vehicle motion and data augmentation |
CN112785514A (en) * | 2020-12-31 | 2021-05-11 | 杭州荣旗科技有限公司 | Rapid bar code image motion reduction method |
CN113534733A (en) * | 2021-07-16 | 2021-10-22 | 广东中烟工业有限责任公司 | Cigarette box bar code identification system and method |
-
2021
- 2021-12-31 CN CN202111681201.9A patent/CN114529555A/en active Pending
Patent Citations (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20090316009A1 (en) * | 2008-06-20 | 2009-12-24 | Atsushi Ito | Apparatus, Method, and Program for Processing Image |
CN105913395A (en) * | 2016-04-10 | 2016-08-31 | 中国人民解放军海军航空工程学院青岛校区 | Moving object observation and fuzzy restoration method |
CN106650546A (en) * | 2016-10-20 | 2017-05-10 | 国家电网公司 | Method for automatically identifying substation switchgear equipment object based on two-dimensional code |
CN107025639A (en) * | 2017-04-05 | 2017-08-08 | 中科微至智能制造科技江苏有限公司 | A kind of Bar code positioning method under complex environment |
KR20180128299A (en) * | 2017-05-23 | 2018-12-03 | (주) 루리텍 | Distance measuring error calibration apparatus for distance measuring camera |
CN208000278U (en) * | 2018-01-11 | 2018-10-23 | 河南中烟工业有限责任公司 | The smoke box appearance delection device recycled for smoke box |
CN108620343A (en) * | 2018-04-09 | 2018-10-09 | 河南中烟工业有限责任公司 | A kind of smoke box appearance quality detection method based on CCD |
CN109807071A (en) * | 2018-09-18 | 2019-05-28 | 红塔烟草(集团)有限责任公司 | A kind of smoke box appearance quality detection and offline checking system and its method |
CN110427792A (en) * | 2019-06-24 | 2019-11-08 | 广西中烟工业有限责任公司 | A kind of recognition matching process blocking two dimensional code and its application |
CN112435190A (en) * | 2020-11-24 | 2021-03-02 | 同济大学 | Motion fuzzy target detection method based on vehicle motion and data augmentation |
CN112785514A (en) * | 2020-12-31 | 2021-05-11 | 杭州荣旗科技有限公司 | Rapid bar code image motion reduction method |
CN113534733A (en) * | 2021-07-16 | 2021-10-22 | 广东中烟工业有限责任公司 | Cigarette box bar code identification system and method |
Non-Patent Citations (3)
Title |
---|
彭楷烽;朱厚耀;施冬冬;邓清文;马亮华;: "基于PCA的条形码定位识别方法", 机电工程技术, no. 06, 20 June 2020 (2020-06-20) * |
牛振岐 等: "基于像素级别的Kinect深度测量误差补偿方法", 《光电子• 激光》, vol. 27, no. 11, 15 November 2016 (2016-11-15), pages 1171 * |
袁鹏: "基于FPGA 的包装破损检测物流分拣***研究", 《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》, 15 November 2020 (2020-11-15), pages 11 - 12 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114882370A (en) * | 2022-07-07 | 2022-08-09 | 西安超嗨网络科技有限公司 | Intelligent commodity identification method and device, terminal and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110866924B (en) | Line structured light center line extraction method and storage medium | |
CN111179243A (en) | Small-size chip crack detection method and system based on computer vision | |
CN112651968B (en) | Wood board deformation and pit detection method based on depth information | |
CN113570605B (en) | Defect detection method and system based on liquid crystal display panel | |
CN110210477B (en) | Digital instrument reading identification method | |
CN111091544B (en) | Method for detecting breakage fault of side integrated framework of railway wagon bogie | |
CN111754466B (en) | Intelligent detection method for damage condition of conveyor belt | |
CN114972356B (en) | Plastic product surface defect detection and identification method and system | |
CN112446225B (en) | Determination of module size of optical code | |
CN115170669A (en) | Identification and positioning method and system based on edge feature point set registration and storage medium | |
CN110070523B (en) | Foreign matter detection method for bottle bottom | |
CN104331695A (en) | Robust round identifier shape quality detection method | |
CN112560538B (en) | Method for quickly positioning damaged QR (quick response) code according to image redundant information | |
JPH03257304A (en) | Position recognizing apparatus | |
CN115880699B (en) | Food packaging bag detection method and system | |
CN111861979A (en) | Positioning method, positioning equipment and computer readable storage medium | |
CN115471476A (en) | Method, device, equipment and medium for detecting component defects | |
CN115752969A (en) | Method, system and equipment for detecting sealing performance of aluminum foil seal | |
CN114529555A (en) | Image recognition-based efficient cigarette box in-and-out detection method | |
CN112561875A (en) | Photovoltaic cell panel coarse grid detection method based on artificial intelligence | |
CN111968082A (en) | Product packaging defect detection and identification method based on machine vision | |
CN116309532A (en) | Method, device, equipment and medium for detecting quality of target object | |
CN115239595A (en) | Method for detecting qualification of two-dimensional code of packaging printed matter | |
CN114419317A (en) | Light strip center extraction method for light with complex environment line structure | |
Yuhan et al. | Detection of road surface crack based on PYNQ |
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 |