CN115423771B - Quasi-dynamic laser anti-counterfeit label identification method based on feature non-uniformity - Google Patents
Quasi-dynamic laser anti-counterfeit label identification method based on feature non-uniformity Download PDFInfo
- Publication number
- CN115423771B CN115423771B CN202211061606.7A CN202211061606A CN115423771B CN 115423771 B CN115423771 B CN 115423771B CN 202211061606 A CN202211061606 A CN 202211061606A CN 115423771 B CN115423771 B CN 115423771B
- Authority
- CN
- China
- Prior art keywords
- laser
- area
- frame
- image
- laser powder
- 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
- 238000000034 method Methods 0.000 title claims abstract description 48
- 239000000843 powder Substances 0.000 claims abstract description 180
- 230000008859 change Effects 0.000 claims abstract description 93
- 230000004044 response Effects 0.000 claims description 14
- 238000010586 diagram Methods 0.000 claims description 8
- 238000012545 processing Methods 0.000 claims description 7
- 239000003086 colorant Substances 0.000 claims description 6
- 238000004458 analytical method Methods 0.000 claims description 5
- 238000012216 screening Methods 0.000 claims description 4
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 claims description 3
- 238000004422 calculation algorithm Methods 0.000 claims description 3
- 238000003708 edge detection Methods 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 3
- 238000002347 injection Methods 0.000 claims description 2
- 239000007924 injection Substances 0.000 claims description 2
- 229910052705 radium Inorganic materials 0.000 claims description 2
- HCWPIIXVSYCSAN-UHFFFAOYSA-N radium atom Chemical compound [Ra] HCWPIIXVSYCSAN-UHFFFAOYSA-N 0.000 claims description 2
- 230000011218 segmentation Effects 0.000 claims description 2
- 239000000284 extract Substances 0.000 abstract description 4
- 238000004088 simulation Methods 0.000 description 13
- 238000000605 extraction Methods 0.000 description 6
- 230000007547 defect Effects 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 238000011156 evaluation Methods 0.000 description 3
- 230000002159 abnormal effect Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000013441 quality evaluation Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 1
- 229910052751 metal Inorganic materials 0.000 description 1
- 239000002184 metal Substances 0.000 description 1
- 238000012795 verification 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
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06K—GRAPHICAL DATA READING; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
- G06K7/00—Methods or arrangements for sensing record carriers, e.g. for reading patterns
- G06K7/10—Methods or arrangements for sensing record carriers, e.g. for reading patterns by electromagnetic radiation, e.g. optical sensing; by corpuscular radiation
- G06K7/14—Methods or arrangements for sensing record carriers, e.g. for reading patterns by electromagnetic radiation, e.g. optical sensing; by corpuscular radiation using light without selection of wavelength, e.g. sensing reflected white light
- G06K7/1404—Methods for optical code recognition
- G06K7/1408—Methods for optical code recognition the method being specifically adapted for the type of code
- G06K7/1417—2D bar codes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/28—Quantising the image, e.g. histogram thresholding for discrimination between background and foreground patterns
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
- G06V10/467—Encoded features or binary features, e.g. local binary patterns [LBP]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/50—Extraction of image or video features by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/56—Extraction of image or video features relating to colour
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/761—Proximity, similarity or dissimilarity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/46—Extracting features or characteristics from the video content, e.g. video fingerprints, representative shots or key frames
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/80—Recognising image objects characterised by unique random patterns
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30168—Image quality inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30204—Marker
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/09—Recognition of logos
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Electromagnetism (AREA)
- Toxicology (AREA)
- Quality & Reliability (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a quasi-dynamic laser anti-counterfeit label identification method based on feature non-uniformity, which comprises the following steps: collecting laser label information; extracting an anti-counterfeiting area in each frame of laser label image; calculating the ambiguity and average brightness of each frame of anti-counterfeiting area image; determining the center point position of each laser powder block according to each laser powder block area; extracting color distribution characteristics of each subarea of the anti-counterfeiting area image; extracting color change characteristics of the anti-counterfeiting area; and determining the authenticity of the laser anti-counterfeit label. The invention extracts the non-uniformity of the color distribution and the color change characteristics of the laser label, so as to describe the unique characteristics of the random distribution of the laser patterns of the real laser anti-counterfeit label, solve the problem of false identification of the false anti-counterfeit label caused by the adoption of the excessively single color change characteristics in the prior art, effectively improve the resistance of the false anti-counterfeit label and improve the accuracy of the identification of the laser anti-counterfeit label.
Description
Technical Field
The invention belongs to the technical field of image processing, and further relates to a quasi-dynamic laser anti-counterfeit label identification method based on feature non-uniformity in the technical field of image identification. The invention can be applied to the pseudo identification of the quasi-dynamic laser anti-counterfeit label of the commodity.
Background
The quasi-dynamic laser anti-fake label is one means of making anti-fake mark for commodity, and is especially one laser holographic technology to print complicated laser patterns on metal film and the patterns show different diffraction patterns. The identification of the quasi-dynamic laser anti-counterfeit label is to identify the quasi-dynamic laser anti-counterfeit label on the commodity, thereby determining the authenticity of the commodity. With the popularization of laser patterns, counterfeiters in the market forge the anti-counterfeiting label by a high-simulation method to realize the counterfeiting of products, so that the current demand for a robust counterfeit label counterfeiting method is continuously increased.
The Hangzhou Warewobu Internet of things technology Co., ltd discloses a pseudo-method based on identification of a quasi-dynamic laser tag in patent literature (patent application number: 201910663889.4, application publication number: CN 110428028A) based on a quasi-dynamic laser tag, a device, equipment and a medium. The method comprises the steps of firstly identifying and uploading bar code information, receiving seed information corresponding to the bar code information, comparing the characteristics of the laser label image with those of the seed information, and finally analyzing the color change of the multi-frame laser label image to judge whether the color change characteristics of the laser label are met. The method can achieve better fake verification results. However, the method still has the defects that the quality of the video frame image is not evaluated, namely whether the video frame image is clear or not and whether the brightness is uniform or not, so that the image with poor quality interferes with the color change feature extraction process, only the color change features of the laser tag are focused, the features are single, the method can not accurately identify the false anti-counterfeit tag with similar color change, and the identification precision of the false anti-counterfeit tag is influenced.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides a quasi-dynamic laser anti-counterfeit label identification method based on feature non-uniformity, which is used for solving the problem of false identification of the false anti-counterfeit label caused by neglecting quality evaluation of images in the identification process and single color change feature.
The method is characterized in that the method evaluates the image quality of the video frames of the laser label image by calculating the variance and average brightness of the laser label image, and screens out the video frames with uniform brightness and clear images from the evaluation result for subsequent identification. Because the image quality evaluation result is utilized to screen out the image with over-high brightness or uneven brightness, the interference of the abnormal brightness of the image to the color change feature extraction process is avoided, and the influence on the recognition precision is reduced. The invention extracts the non-uniformity of the color distribution and the color change characteristics of the laser label, namely extracts the color distribution of the laser pattern based on HSV color space analysis, extracts the color change of the laser pattern based on a frame difference method, and judges whether the color distribution or the variation non-uniformity exists according to the difference of the color distribution and the color change of each area of the laser label. The color distribution inconsistency indicates that the on-off state of each laser powder block of the laser label is random, and the color tone shows diversity, wherein the on-off state refers to whether the laser powder block shows color or not. Color variation non-uniformity indicates that the color tone variation of the laser patterns of the laser labels in the multi-frame images shows diversity. The unique characteristics of random distribution of laser patterns of the real laser anti-counterfeit label are further described by the two non-consistencies, so that the resistance to the false anti-counterfeit label is effectively enhanced.
The method comprises the following specific steps:
step 1, collecting laser anti-counterfeiting label information:
step 1.1, acquiring a two-dimensional code image in a laser anti-counterfeiting label, analyzing the two-dimensional code image by utilizing a two-dimensional code analysis algorithm, uploading obtained two-dimensional code information to a server, and calling seed information corresponding to the two-dimensional code information stored at the server;
step 1.2, collecting laser anti-counterfeit label video;
step 2, extracting anti-counterfeiting areas in each frame of laser label image:
step 2.1, carrying out gray processing on each frame of laser anti-counterfeit label RGB image in the laser anti-counterfeit label video to obtain a gray image of each frame of laser anti-counterfeit label, extracting edge information in the gray image of each frame of laser anti-counterfeit label by using an edge detection method, outputting the edge image of each frame of laser anti-counterfeit label, analyzing the topological structure of the edge image of each frame of laser anti-counterfeit label by using a raster scanning and marking method, searching for contours, and forming a candidate contour set;
step 2.2, screening the anti-counterfeiting areas conforming to the shape characteristics and the area characteristics in each frame of laser anti-counterfeiting label image from the candidate contour set by utilizing the shape characteristics and the area characteristics of the anti-counterfeiting areas recorded in the seed information, and correcting the anti-counterfeiting areas of each frame of laser anti-counterfeiting label image by utilizing affine transformation;
step 2.3, extracting and counting local binary pattern LBP characteristics of the corrected gray level diagram of each frame of anti-counterfeiting area to obtain an LBP characteristic histogram of the gray level diagram of each frame of anti-counterfeiting area, calculating similarity with the LBP characteristic histogram recorded in the seed information by using a correlation comparison method, reserving the frame anti-counterfeiting area image with the similarity larger than a similarity threshold, and discarding the frame anti-counterfeiting area image with the similarity smaller than or equal to the threshold, wherein the similarity threshold is a value selected in a range of [0.7,0.8 ].
Step 3, calculating the ambiguity and average brightness of each frame of anti-counterfeiting area image:
step 3.1, performing binarization processing on the gray level map of each frame of anti-counterfeiting area by using a maximum inter-class variance method to obtain a black-and-white map of the anti-counterfeiting area, taking a white area in the black-and-white map of the anti-counterfeiting area as each laser powder block after segmentation, and calculating the area proportion of all the laser powder blocks;
step 3.2, calculating the blurring degree of the gray level map of each frame of anti-counterfeiting area;
step 3.3, calculating the average brightness of each frame of anti-counterfeiting area image;
step 4, determining the center point position of each laser powder block according to each laser powder block area;
step 5, extracting color distribution characteristics of each subarea of the anti-counterfeiting area image based on the color state of each laser powder block in each frame of the anti-counterfeiting area image:
step 5.1, dividing each frame of anti-counterfeiting area image into four subareas with the same size according to the horizontal direction and the vertical direction;
step 5.2, judging the on-off state of each laser powder block in each subarea according to the value of the tone S channel of the central point of each laser powder block and the range of the value of the brightness V channel;
step 5.3, generating a color histogram of each sub-area according to the colors of all the laser powder blocks in the bright state in each sub-area;
step 5.4, judging whether the brightness state of all laser powder blocks in each sub-area meets tau or not 1 ≤γ t ≤τ 2 If yes, judging the on-off states of all laser powder blocks in the subarea as non-uniformity, otherwise, judging the on-off states of all laser powder blocks in the subarea as uniformity, wherein gamma is calculated as the uniformity t Representing the proportion of the number of laser powder blocks in a bright state in the t th sub-region to the total number of laser powder blocks, and tau 1 For the luminance non-uniformity lower threshold, τ 2 The brightness non-uniformity upper threshold is the proportion gamma of the quantity of the laser powder blocks in the bright state of each subarea to the total quantity of the laser powder blocks under different angles by the statistically true laser label t And based on the plurality of gamma obtained t Taking the minimum value of the values as tau 1 Taking the maximum value of τ 2 ；
Step 5.5, judging whether the tone state of all laser powder blocks in each sub-area meets ρ t <ζ, if yes, determining that the tone states of all laser powder blocks in the subarea are non-uniform, otherwise, determining that the tone states of all laser powder blocks in the subarea are uniformThe inducibility, ρ t The ratio of the peak value of the color histogram of the t sub-area to the total number of laser powder blocks is represented, and xi is a tone non-consistency threshold value which is obtained by taking the maximum value from the ratio of the peak value of the color histogram of each sub-area to the total number of laser powder blocks under different angles of the statistically true laser label;
step 5.6, judging whether the on-off state or the tone state of all laser powder blocks in each sub-area is judged to be inconsistent, if so, judging that the color distribution characteristics of the sub-area are inconsistent, otherwise, judging that the color distribution characteristics of the sub-area are consistent;
step 6, extracting color change characteristics of the anti-counterfeiting area based on the color change of each laser powder block in the multi-frame anti-counterfeiting area image:
step 6.1, judging whether the saturation S channel value and the saturation V channel value of the central point of the laser powder block meet the channel threshold condition, if so, judging that the central point of the laser powder block has the on-off change, otherwise, judging that the central point of the laser powder block has no on-off change;
step 6.2, utilize
The formula calculates the tone variation of the central point of each laser powder block in the HSV image formed by the tone H channel, the saturation S channel and the brightness V channel in each frame of anti-counterfeiting area,
represents the tone variation of the center point of the p-th laser powder block in the HSV image in the t-th frame anti-counterfeiting area,/for the center point of the p-th laser powder block>
Respectively representing H channel values of HSV images of the p-th laser powder block center points in t-th frames and t-x-th frame anti-counterfeiting areas, wherein x is E (0, t);
step 6.3, judging the tone variation of the center point of each laser powder block
Whether the color change threshold value K is larger than the color change threshold value K is judged, if yes, the central point of the laser powder block is judged to have color change in the frame anti-counterfeiting area image, otherwise, the central point of the laser powder block is judged to have no color change in the frame anti-counterfeiting area, wherein the color change threshold value K is an average value of range values of H channels corresponding to seven basic colors in HSV color space;
step 6.3, counting the number of laser powder block center points with the brightness change and the tone change each time, and recording the frame number of the image frame of the anti-counterfeiting area with the change;
step 6.4, judging whether the number of the central points of the laser powder blocks with the brightness change and the tone change each time and the frame number of the image frame of the anti-counterfeiting area with the change meet the color change non-uniformity condition, if so, judging the color change characteristics of the laser label as non-uniformity, otherwise, judging the color change characteristics of the laser label as uniformity;
step 7, determining the authenticity of the laser anti-counterfeit label:
and judging the laser anti-counterfeit label meeting the duplicate non-uniformity condition as a true label, otherwise, judging the laser anti-counterfeit label as a false label.
Compared with the prior art, the invention has the following advantages:
firstly, as the invention calculates the ambiguity and average brightness of the anti-counterfeiting area of each frame of laser label image for the video frame, the evaluation result can screen out the video frame with uniform brightness and clear image, and overcomes the defect of reduced recognition precision caused by interference of the color change feature extraction process due to the video frame image with poor unfiltered quality in the prior art, so that the invention can effectively avoid the interference of the feature extraction process, thereby improving the recognition accuracy.
Secondly, due to the fact that the non-uniformity of the color distribution and the color change characteristics of the laser label is extracted, the unique characteristics of the random distribution of the laser patterns of the real laser anti-counterfeit label are further described, the defect of false identification of the false anti-counterfeit label caused by the fact that the color change characteristics are single in the prior art is overcome, and the countermeasures of the false anti-counterfeit label are effectively improved.
Drawings
FIG. 1 is a flow chart of the present invention;
fig. 2 is a simulation diagram of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and examples.
The specific steps of the present invention will be described in further detail with reference to fig. 1 and the embodiment.
And step 1, collecting laser anti-counterfeit label information.
Step 1.1, acquiring a two-dimensional code image in a laser anti-counterfeiting label, analyzing the two-dimensional code image by utilizing a two-dimensional code analysis algorithm, uploading the obtained two-dimensional code information to a server, and calling seed information corresponding to the two-dimensional code information stored at the server.
And 1.2, collecting laser anti-counterfeit label videos.
And 2, extracting an anti-counterfeiting area in each frame of laser label image.
And 2.1, carrying out gray processing on each frame of laser anti-counterfeit label RGB image in the laser anti-counterfeit label video to obtain a gray image of each frame of laser anti-counterfeit label, extracting edge information in the gray image of each frame of laser anti-counterfeit label by using an edge detection method, outputting the edge image of each frame of laser anti-counterfeit label, analyzing the topological structure of the edge image of each frame of laser anti-counterfeit label by using a raster scanning and marking method, and searching for contours to form a candidate contour set.
And 2.2, screening the anti-counterfeiting areas conforming to the shape characteristics and the area characteristics in each frame of laser anti-counterfeiting label image from the candidate contour set by utilizing the shape characteristics and the area characteristics of the anti-counterfeiting areas recorded in the seed information, and correcting the anti-counterfeiting areas of each frame of laser anti-counterfeiting label image by utilizing affine transformation so that the anti-counterfeiting areas are standard rectangles.
And 2.3, extracting and counting local binary pattern LBP characteristics of the corrected gray level map of each frame of anti-counterfeiting area to obtain an LBP characteristic histogram of the gray level map of each frame of anti-counterfeiting area, calculating similarity with the LBP characteristic histogram recorded in the seed information by using a correlation comparison method, and reserving the anti-counterfeiting area image if the similarity exceeds a set threshold value, wherein the texture characteristics of the anti-counterfeiting area image of the frame are matched with the seed information, otherwise, discarding the anti-counterfeiting area image of the frame.
And step 3, calculating the ambiguity and average brightness of the anti-counterfeiting area of each frame.
And 3.1, dividing the laser powder blocks in the anti-counterfeiting area from the anti-counterfeiting area of each frame, and calculating the area proportion of all the laser powder blocks.
And carrying out graying treatment on each frame of RGB image of the anti-counterfeiting area to obtain an anti-counterfeiting area gray scale image. And performing binarization processing on the gray level map by using a maximum inter-class variance method (Ostu) to obtain a black-and-white map of the anti-counterfeiting area of the frame. The white area in the black-and-white image is divided laser powder blocks.
And calculating the proportion of the area of all the laser powder blocks divided in each frame of black-and-white image to the total area of the black-and-white image to obtain the proportion of the number of white pixels to the number of total pixels in the black-and-white image. Comparing the proportion with a true proportion value recorded in seed information stored when the corresponding laser anti-counterfeiting label leaves the factory, wherein the true proportion value refers to the proportion of the area of all laser powder blocks actually contained in the anti-counterfeiting area relative to the total area of the anti-counterfeiting area. The difference between the area proportion of the laser powder block calculated from the black-and-white image of each frame of anti-counterfeiting area and the recorded real proportion value is denoted as epsilon, if the difference between the two proportion values is larger than a set threshold value mu, namely epsilon is larger than mu, the area proportion of the laser powder block is calculated based on the divided laser powder block areas, and if the difference between the proportion values is large, the fact that all laser powder block areas cannot be correctly divided from the frame of anti-counterfeiting area image is indicated, and the follow-up identification process needs to be based on the accurately divided laser powder block areas, so that the frame of anti-counterfeiting area RGB image is abandoned. If the ratio difference is less than or equal to the set threshold μ, then the calculation of the subsequent step is performed. In the embodiment of the present invention, μ=0.3.
And 3.2, calculating the blurring degree of the gray level map of the anti-counterfeiting area.
And convolving each pixel in the gray level map of the anti-counterfeiting area with a Laplacian operator to obtain a second derivative response map. A variance of the second derivative response map is calculated from each pixel value of the second derivative response map. Because the second derivative response map reflects the degree of response to the image edge, the edge of the image is clear, and a high response value exists in the second derivative response map, so that the variance of the response map is large, and the edge of the image is blurred, and the response values in the second derivative response map are low, so that the variance of the response map is small, and therefore the magnitude of the variance represents the degree of sharpness of the image edge. If the variance is larger than the set threshold value theta, the edge of the anti-counterfeiting area image is clear, and the calculation of the subsequent steps can be performed; if the variance is smaller than the set threshold value theta, the image edge of the anti-counterfeiting area is fuzzy and needs to be discarded. In the embodiment of the present invention, θ=50.
And 3.3, calculating the average brightness of each frame of anti-counterfeiting area image.
And converting the RGB image of the anti-counterfeiting area into an HSV color space to obtain the HSV image of the anti-counterfeiting area. And (5) calculating the average value of the V channel values of all the pixel points of the HSV image. Since the V-channel represents brightness, if the average value of the V-channel is greater than or equal to the set threshold value delta 1 Indicating that the image brightness is too bright; if the average value is too low, less than or equal to the set threshold value delta 2 Indicating that the image brightness is too low. Because the brightness of the image is too high or too low, when the color change characteristics are extracted from the images of the multi-frame anti-counterfeiting area, the color change caused by abnormal brightness is generated, but the color change is not the effect of the laser powder block under different display angles of the laser label due to light diffraction, so that the recognition result is interfered, and the images of the frame anti-counterfeiting area are abandoned so as to ensure the accuracy of the recognition result. Delta 1 And delta 2 Is an empirical value for detecting brightness anomalies in engineering practice. In embodiments of the invention, delta 1 ＝220，δ 2 ＝20。
And 4, determining the center point position of each laser powder block according to each laser powder block area.
And 4.1, performing an AND operation on the anti-counterfeiting area black-and-white image and the binary image in the corresponding seed information pixel by pixel, and eliminating an incorrect white area in each laser powder block to obtain a public area black-and-white image. Because the binary image in the seed information is a real black-and-white image of the anti-counterfeiting area, the black-and-white image of the anti-counterfeiting area is provided with a reference for the correctness of each laser powder block which is divided. In the common area black-and-white map, each white area represents one laser patch. Analyzing the topological structure of the black-and-white map of the public area by using a raster scanning and marking method, namely, the connectivity of each pixel and the neighborhood pixels thereof, extracting the outline of each white area, and determining the center point position of each outline, namely, obtaining the center point position of each laser powder block.
And 5, extracting color distribution characteristics of each subarea of the anti-counterfeiting area image.
And 5.1, dividing each frame of anti-counterfeiting area image into four subareas with the same size according to the horizontal direction and the vertical direction.
And 5.2, calculating the on-off state of each laser powder block in each sub-area.
And converting the RGB image of each frame of anti-counterfeiting area into an HSV color space, wherein H in the HSV color space represents hue, S represents saturation and V represents brightness. For each laser powder block center point of each sub-region, if S i Alpha and delta are less than or equal to 1 ≤V i ≤δ 2 And the laser powder block corresponding to the center point is gray, namely the laser powder block is in a state of being out. If S i >Alpha and delta 1 ≤V i ≤δ 2 The laser powder block corresponding to the center point presents color, and the laser powder block is in a bright state. Wherein S is i Representing the saturation S channel value of the center point of the ith laser powder block, V i And the value of the brightness V channel of the center point of the ith laser powder block is represented, and i is the serial number of the center point of each laser powder block in each subarea. In the embodiment of the present invention, α is a saturation S channel threshold of gray in HSV color space, α=43.
And 5.3, generating a color histogram of each sub-area.
There are seven basic colors in the HSV color space, one for each range of H-channel values. And for each laser powder block in a bright state in each subarea, acquiring the basic color of each laser powder block according to the value of the color tone H channel of the central point of the laser powder block. And generating a color histogram of each subarea according to the counted number of laser powder blocks corresponding to each basic color in the subarea.
Step 5.4, judging whether the brightness state of all laser powder blocks in each sub-area meets tau or not 1 ≤γ t ≤τ 2 If yes, judging the on-off states of all laser powder blocks in the subarea as non-uniformity, otherwise, judging the on-off states of all laser powder blocks in the subarea as uniformity. Wherein, gamma t The ratio of the number of laser powder blocks in the bright state in the t th sub-area to the total number of laser powder blocks is represented. τ 1 ，τ 2 The two thresholds are set by counting the proportion gamma of the quantity of the laser powder blocks in the bright state of each subarea to the total quantity of the laser powder blocks under different angles of a real laser label t And based on the plurality of gamma obtained t Taking the minimum value of the values as tau 1 Taking the maximum value of τ 2 . In an embodiment of the invention, τ 1 ＝0.2，τ 2 ＝0.8。
Step 5.5, judging whether the tone state of all laser powder blocks in each sub-area meets ρ t <And xi, if yes, judging that the tone states of all the laser powder blocks in the subarea are inconsistent, otherwise, judging that the tone states of all the laser powder blocks in the subarea are consistent. ρ t The peak value of the color histogram representing the t-th sub-region is the proportion of the total number of laser powder blocks. And xi is a set threshold value, and under different angles, the ratio of the peak value of the color histogram of each sub-region to the total number of laser powder blocks of the real laser label is counted, wherein the maximum value is taken as xi. In the embodiment of the present invention, ζ=0.7.
And 5.6, judging whether the color distribution characteristics of each sub-area are inconsistent, namely judging whether the on-off state or the tone state of all laser powder blocks in the sub-area are inconsistent, if so, judging that the color distribution characteristics of the sub-area are inconsistent, otherwise, judging that the color distribution characteristics of the sub-area are consistent.
And 6, extracting color change characteristics of the anti-counterfeiting area.
And 6.1, detecting whether the laser powder block has a change of on and off.
2. Each laser powder block corresponds to a central point, and as the whole powder block changes when the laser powder block changes, whether the laser powder block changes in brightness or not can be judged by whether the central point of the laser powder block changes in brightness or not. And judging that the laser powder blocks corresponding to the center point have the on-off change when the two channel threshold conditions are met for the values of the saturation S channel and the brightness V channel of the center point of any laser powder block, and judging that the laser powder blocks corresponding to the center point have no on-off change when the two channel threshold conditions are not met. The two-channel threshold condition is that a third formula is satisfied and either one of the first formula and the second formula is satisfied:
wherein,,
s channel values of HSV images respectively representing the center points of the c-th laser powder blocks in t-th frame and t-x-th frame anti-counterfeiting areas are->
The V channel values of HSV images of the anti-counterfeiting areas of the nth frame and the nth-x frame of the center point of the nth laser powder block are respectively shown, and x is E (0, t). All sat_max, sat_min and val_range are set thresholds, and all radium in a deactivated state in a real laser label is takenThe maximum value in the S channel values of the center point of the powder injection block is sat_min, the minimum value in the S channel values of the center points of all the laser powder blocks in a bright state in the real laser label is sat_max, and the average value of the V channel variation amounts of the center points of all the laser powder blocks with the bright-dark variation in the real laser label is val_range. In the embodiment of the present invention, sat_max=200, sat_min=43, val_range=100.
And 6.2, detecting whether the center point of each laser powder block has color change.
By means of
And (3) calculating the tone variation of the central point of each laser powder block in the HSV image in each frame of anti-counterfeiting area according to the formula. If->
And judging that the central point of the laser powder block has tone change in the frame anti-counterfeiting area image. If->
And judging that the central point of the laser powder block does not have tone change in the frame anti-counterfeiting area. Wherein (1)>
Represents the tone variation of the center point of the p-th laser powder block in the HSV image in the t-th frame anti-counterfeiting area,
the H channel values of the HSV images of the anti-counterfeiting areas of the p-th laser powder block center points in the t-th frame and the t-x-th frame are respectively shown, and x is E (0, t). And kappa is a set threshold value, and the threshold value is an average value of range values of H channels corresponding to seven basic colors in the HSV color space. In the embodiment of the present invention, κ=20.
And 6.3, counting the number of laser powder block central points with the brightness change and the tone change each time, recording the frame number of the image frame of the anti-counterfeiting area with the change, and judging whether the color change non-uniformity condition is met. If so, judging that the color change characteristics of the laser label are inconsistent, otherwise, judging that the color change characteristics of the laser label are consistent. The color change non-uniformity condition is that the following three conditions are simultaneously satisfied:
C i >m 3 or D i >m 3
Bc>m 4
wherein F represents a set formed by frame numbers of recorded anti-counterfeiting area image frames with color modulation changes at the center point of the laser powder block, and L represents a set formed by frame numbers of recorded anti-counterfeiting area image frames with light-on and light-off changes at the center point of the laser powder block. C (C) i Representing the total number of laser powder block center points with color tone change in an ith frame of anti-counterfeiting area image, D i And the total number of the central points of the laser powder blocks with the on-off change in the ith frame of anti-counterfeiting area image is represented. Bc represents the total number of frame numbers in set B, b=f D. m is m 1 ，m 2 ，m 3 ，m 4 All are set thresholds. m is m 1 And taking the average value of the total number of the center points of the laser powder blocks with the color tone change when the plurality of real laser labels respectively generate e color tone changes. m is m 2 And taking the average value of the total number of the center points of the laser powder blocks with brightness change when the brightness of the real laser labels is changed for e times. m is m 3 Taking the average value of the total number of the laser powder block central points of the real laser label with each tone change. m is m 4 When e brightness changes and e tone changes are respectively generated by a plurality of real laser labels, the average value of the frame numbers of the image frames of the anti-counterfeiting area with the tone changes and the brightness changes simultaneously occurs. In the embodiment of the invention, for example, f= {6, 10, 15}, it is indicated that the central point of the laser powder block is changed in color tone in the 6 th, 10 th and 15 th anti-fake area images. m is m 1 ＝30，m 2 ＝30，m 3 ＝5，m 4 ＝2，e＝8。
And 7, judging whether the laser anti-counterfeit label is true or false.
Judging whether the color characteristics of the anti-counterfeiting area of the laser anti-counterfeiting label meet the two non-consistency conditions, if so, judging that the laser anti-counterfeiting label is a true label, otherwise, judging that the laser anti-counterfeiting label is a false label. The duplicate non-uniformity condition is that the color distribution characteristics and the color change characteristics of the anti-counterfeiting area of the laser anti-counterfeiting label are non-uniform.
The effects of the present invention can be further illustrated by the following simulation experiments.
1. Simulation experiment conditions:
the hardware platform of the simulation experiment of the invention is: the processor is Intel i57500 CPU, the main frequency is 3.5GHz, and the memory is 16GB.
The software platform of the simulation experiment of the invention is: windows 10 operating system and python 3.6.7.
The simulation experiment of the invention uses the collected video of 69 true laser anti-counterfeit labels and the collected video of 17 false laser anti-counterfeit labels for simulation.
2. Simulation content and analysis of results thereof
The simulation experiment of the invention adopts the method of the invention to identify the authenticity of each laser anti-counterfeit label according to the collected genuine-fake laser anti-counterfeit label video, and the genuine-fake judgment results of 86 laser anti-counterfeit labels are obtained.
The recognition effect is evaluated by using an accuracy evaluation method:
where Acc represents the accuracy, TP represents the number of false tags correctly identified as false tags, TN represents the number of true tags correctly identified as true tags, FP represents the number of false tags incorrectly identified as true tags, and FN represents the number of false tags incorrectly identified as true tags. Tp=60, tn=17, fp=9, fn=0, acc=0.895, wherein fp=9, i.e. the reason why 9 true labels are erroneously judged as false labels is that the video image quality of the 9 true labels is poor, and the anti-counterfeit area cannot be extracted or the screening step of calculating the ambiguity and average brightness of the anti-counterfeit area in the method of the present invention is not passed. In 77 laser anti-counterfeit label videos with clear images and uniform brightness, the identification accuracy of the method is 100%.
The effects of the present invention are further described below in conjunction with the simulation diagram of fig. 2.
Fig. 2 (a) is an exemplary diagram of a video frame with too low brightness in a real laser anti-counterfeit label video in a simulation experiment of the present invention.
As can be seen from fig. 2 (a), when the brightness is too low, the position of the laser powder block in the video frame is blurred, the displayed color is disturbed, and is inconsistent with the actual displayed color of the laser powder block, and the color feature extraction process is disturbed and needs to be abandoned.
Fig. 2 (b) is an exemplary diagram of a video frame with too high brightness in a real laser anti-counterfeit label video in a simulation experiment of the present invention.
As can be seen from fig. 2 (b), when the brightness is too high, the position of the laser powder block in the video frame is not clear, the color of the laser powder block is disturbed, and the color is inconsistent with the color actually displayed by the laser powder block, which can interfere with the color feature extraction process and needs to be discarded.
Claims (10)
1. The quasi-dynamic laser anti-counterfeit label identification method based on feature non-uniformity is characterized by extracting color distribution features of all subareas in an anti-counterfeit area image, judging whether the color distribution features of all subareas are non-uniform, extracting color change features of the anti-counterfeit area, and judging whether the color change features are non-uniform; the method comprises the following steps:
step 1, collecting laser anti-counterfeiting label information:
step 1.1, acquiring a two-dimensional code image in a laser anti-counterfeiting label, analyzing the two-dimensional code image by utilizing a two-dimensional code analysis algorithm, uploading obtained two-dimensional code information to a server, and calling seed information corresponding to the two-dimensional code information stored at the server;
step 1.2, collecting laser anti-counterfeit label video;
step 2, extracting anti-counterfeiting areas in each frame of laser label image:
step 2.1, carrying out gray processing on each frame of laser anti-counterfeit label RGB image in the laser anti-counterfeit label video to obtain a gray image of each frame of laser anti-counterfeit label, extracting edge information in the gray image of each frame of laser anti-counterfeit label by using an edge detection method, outputting the edge image of each frame of laser anti-counterfeit label, analyzing the topological structure of the edge image of each frame of laser anti-counterfeit label by using a raster scanning and marking method, searching for contours, and forming a candidate contour set;
step 2.2, screening the anti-counterfeiting areas conforming to the shape characteristics and the area characteristics in each frame of laser anti-counterfeiting label image from the candidate contour set by utilizing the shape characteristics and the area characteristics of the anti-counterfeiting areas recorded in the seed information, and correcting the anti-counterfeiting areas of each frame of laser anti-counterfeiting label image by utilizing affine transformation;
step 2.3, extracting and counting local binary pattern LBP characteristics of the corrected gray level diagram of each frame of anti-counterfeiting area to obtain an LBP characteristic histogram of the gray level diagram of each frame of anti-counterfeiting area, calculating similarity with the LBP characteristic histogram recorded in the seed information by using a correlation comparison method, reserving the frame anti-counterfeiting area image with the similarity larger than a similarity threshold, and discarding the frame anti-counterfeiting area image with the similarity smaller than or equal to the threshold, wherein the similarity threshold is a value selected in a range of [0.7,0.8 ].
Step 3, calculating the ambiguity and average brightness of each frame of anti-counterfeiting area image:
step 3.1, performing binarization processing on the gray level map of each frame of anti-counterfeiting area by using a maximum inter-class variance method to obtain a black-and-white map of the anti-counterfeiting area, taking a white area in the black-and-white map of the anti-counterfeiting area as each laser powder block after segmentation, and calculating the area proportion of all the laser powder blocks;
calculating the area proportion of all laser powder blocks refers to calculating the proportion of the area of all the laser powder blocks divided in each frame of black-and-white image to the total area of the black-and-white image, obtaining the proportion of the number of white pixels relative to the number of the total pixels in the black-and-white image, comparing the proportion with an area proportion value recorded in seed information, and expressing the difference between the calculated area proportion of the laser powder blocks and the recorded area proportion value of each frame of anti-counterfeiting area black-and-white image as epsilon, if the difference between the two proportions is larger than a threshold mu, namely epsilon| > mu, discarding the image of the frame of anti-counterfeiting area, if the difference between the proportions is smaller than or equal to the threshold mu, retaining the image of the frame of anti-counterfeiting area, and mu taking the absolute value of the difference between the area proportion of all the laser powder blocks obtained when all the powder block areas are correctly divided in the anti-counterfeiting area black-and-white image and the seed information;
step 3.2, calculating the blurring degree of the gray level map of each frame of anti-counterfeiting area;
step 3.3, calculating the average brightness of each frame of anti-counterfeiting area image;
step 4, determining the center point position of each laser powder block according to each laser powder block area;
step 5, extracting color distribution characteristics of each subarea of the anti-counterfeiting area image based on the color state of each laser powder block in each frame of the anti-counterfeiting area image:
step 5.1, dividing each frame of anti-counterfeiting area image into four subareas with the same size according to the horizontal direction and the vertical direction;
step 5.2, judging the on-off state of each laser powder block in each subarea according to the value of the tone S channel of the central point of each laser powder block and the range of the value of the brightness V channel;
step 5.3, generating a color histogram of each sub-area according to the colors of all the laser powder blocks in the bright state in each sub-area;
step 5.4, judging whether the on-off state of all laser powder blocks in each sub-area meets tau or not 1 ≤γ k ≤τ 2 If yes, judging the on-off states of all laser powder blocks in the subarea as non-uniformity, otherwise, judging the on-off states of all laser powder blocks in the subarea as uniformity, wherein gamma is calculated as the uniformity k Representing the proportion of the number of laser powder blocks in a bright state in the kth sub-region to the total number of laser powder blocks, τ 1 For the luminance non-uniformity lower threshold, τ 2 The brightness non-uniformity upper threshold is the radium in a bright state of each subarea under different angles by a statistically true laser labelThe quantity of the powder injection blocks accounts for the proportion gamma of the total quantity of the laser powder blocks k And based on the plurality of gamma obtained k Taking the minimum value of the values as tau 1 Taking the maximum value of τ 2 ；
Step 5.5, judging whether the tone state of all laser powder blocks in each sub-area meets ρ k <ζ, if yes, determining that the tone states of all laser powder blocks in the subarea are non-uniform, otherwise, determining that the tone states of all laser powder blocks in the subarea are uniform, and ρ k The peak value of the color histogram of the kth sub-area accounts for the proportion of the total number of laser powder blocks, and xi is a tone non-consistency threshold value which is obtained by taking the maximum value from the proportion of the peak value of the color histogram of each sub-area to the total number of laser powder blocks under different angles of the statistically true laser label;
step 5.6, judging whether the on-off state or the tone state of all laser powder blocks in each sub-area is judged to be inconsistent, if so, judging that the color distribution characteristics of the sub-area are inconsistent, otherwise, judging that the color distribution characteristics of the sub-area are consistent;
step 6, extracting color change characteristics of the anti-counterfeiting area based on the color change of each laser powder block in the multi-frame anti-counterfeiting area image:
step 6.1, judging whether the saturation S channel value and the saturation V channel value of the central point of the laser powder block meet the channel threshold condition, if so, judging that the central point of the laser powder block has the on-off change, otherwise, judging that the central point of the laser powder block has no on-off change;
step 6.2, utilize
Calculating the tone variation of the central point of each laser powder block in the HSV image formed by a tone H channel, a saturation S channel and a brightness V channel in each frame of anti-counterfeiting area according to a formula, wherein +_>
Representing the t frame anti-counterfeitingTone variation of the center point of the p-th laser powder block in HSV image in the region, ++>
Respectively representing H channel values of HSV images of the p-th laser powder block center points in t-th frames and t-x-th frame anti-counterfeiting areas, wherein x is E (0, t);
step 6.3, judging the tone variation of the center point of each laser powder block
Whether the color change threshold value K is larger than the color change threshold value K is judged, if yes, the central point of the laser powder block is judged to have color change in the frame anti-counterfeiting area image, otherwise, the central point of the laser powder block is judged to have no color change in the frame anti-counterfeiting area, wherein the color change threshold value K is an average value of range values of H channels corresponding to seven basic colors in HSV color space;
step 6.3, counting the number of laser powder block center points with the brightness change and the tone change each time, and recording the frame number of the image frame of the anti-counterfeiting area with the change;
step 6.4, judging whether the number of the central points of the laser powder blocks with the brightness change and the tone change each time and the frame number of the image frame of the anti-counterfeiting area with the change meet the color change non-uniformity condition, if so, judging the color change characteristics of the laser label as non-uniformity, otherwise, judging the color change characteristics of the laser label as uniformity;
step 7, determining the authenticity of the laser anti-counterfeit label:
and judging the laser anti-counterfeit label meeting the duplicate non-uniformity condition as a true label, otherwise, judging the laser anti-counterfeit label as a false label.
2. The method for identifying the quasi-dynamic laser anti-counterfeit label based on the characteristic non-uniformity according to claim 1, wherein in the step 3.1.
3. The method for identifying the quasi-dynamic laser anti-counterfeit label based on the feature inconsistency according to claim 1, wherein the calculating of the blurring degree of the gray level map of each anti-counterfeit area in the step 3.2 means that a Laplacian operator is used for convolution with each pixel in the gray level map of the anti-counterfeit area to obtain a second derivative response map, the variance of the second derivative response map is calculated according to each pixel value of the second derivative response map, the image of the anti-counterfeit area with the variance being greater than a threshold value is reserved, the image of the anti-counterfeit area with the variance being less than or equal to the threshold value is discarded, and the threshold value is taken as an average value of variances of a plurality of blurred gray level maps of the anti-counterfeit area.
4. The method for identifying the quasi-dynamic laser anti-counterfeit label based on the feature non-uniformity according to claim 1, wherein the step 3.3 of calculating the average brightness of each frame of the anti-counterfeit area image means that the RGB image of the anti-counterfeit area is converted into HSV color space to obtain the HSV image of the anti-counterfeit area, the average value of the V channel values of all pixel points of the HSV image is calculated, and if the average value of the V channel is greater than or equal to a threshold delta 1 Reserving the frame anti-counterfeiting area image; if the average value is too low, less than or equal to the threshold value delta 2 Discarding the image of the anti-counterfeiting area of the frame, delta 1 And delta 2 Is an empirical value for detecting brightness anomalies in engineering practice.
5. The method for identifying the quasi-dynamic laser anti-counterfeit label based on the feature non-uniformity according to claim 1, wherein the step 4 of determining the center point position of each laser powder block according to each laser powder block area divided refers to performing a pixel-by-pixel phase operation on the anti-counterfeit area black-and-white image and a binary image in corresponding seed information to obtain a common area black-and-white image, analyzing the topological structure of the common area black-and-white image by using a grating scanning and marking method, extracting the outline of each white area, and taking the center point position of each outline as the center point position of each laser powder block.
6. The identification method of the quasi-dynamic laser anti-counterfeit label based on the characteristic non-uniformity according to claim 4, wherein the identification method is characterized by comprising the following step 52, determining the on-off state of each laser powder block in each sub-region refers to converting the RGB image of each frame of anti-counterfeit region into HSV color space, where H represents hue, S represents saturation, and V represents brightness, for each laser powder block center point of each sub-region, if S i Alpha and delta are less than or equal to 1 ≤V i ≤δ 2 The laser powder block corresponding to the center point is gray, i.e. the laser powder block is in a state of being off, if S i >Alpha and delta 1 ≤V i ≤δ 2 The laser powder block corresponding to the center point presents color, and the laser powder block is in a bright state at the moment, wherein S i Representing the saturation S channel value of the center point of the ith laser powder block, V i And (3) representing the value of the brightness V channel of the central point of the ith laser powder block, wherein i is the serial number of the central point of each laser powder block in each subarea, and alpha is the threshold value of the saturation S channel of gray in HSV color space.
7. The method for identifying the quasi-dynamic laser anti-counterfeit label based on the feature non-uniformity according to claim 1, wherein the step 5.3 of generating the color histogram of each sub-region refers to generating the color histogram of each sub-region by acquiring the basic color of each laser powder block according to the value of the color tone H channel of the center point of the laser powder block for each laser powder block in a bright state and according to the counted number of the laser powder blocks corresponding to each basic color in each sub-region.
8. The method for identifying a quasi-dynamic laser anti-counterfeit label based on feature non-uniformity according to claim 1, wherein the channel threshold condition in step 6.1 means that a third formula is satisfied and any one of the first formula and the second formula is satisfied in the following three formulas:
Wherein,,
s channel values of HSV images respectively representing the center points of the c-th laser powder blocks in t-th frame and t-x-th frame anti-counterfeiting areas are->
The method comprises the steps of respectively representing the V channel values of HSV images of the center point of a c-th laser powder block in t frames and t-x frame anti-counterfeiting areas, x epsilon (0, t), sat_min being an off-state threshold value, wherein the threshold value is the maximum value of the S channel values of the center points of all laser powder blocks in an off-state in a real laser label, sat_max is a bright-state threshold value, the threshold value is the minimum value of the S channel values of the center points of all laser powder blocks in a bright state in the real laser label, val_range is a brightness threshold value, and the threshold value is the average value of the V channel variation of the center points of all laser powder blocks in the real laser label, wherein the brightness variation occurs.
9. The method for identifying the quasi-dynamic laser anti-counterfeit label based on the feature non-uniformity according to claim 1, wherein the color change non-uniformity condition in the step 6.4 is that the following three conditions are satisfied at the same time:
C j >m 3 Or D j >m 3
Bc>m 4
Wherein F represents a set formed by frame numbers of recorded anti-counterfeiting area image frames with color modulation changes at the center point of the laser powder block, L represents a set formed by frame numbers of recorded anti-counterfeiting area image frames with light-on and light-off changes at the center point of the laser powder block, C j Representing the total number of laser powder block center points with color tone change in the jth frame of anti-counterfeiting area image, D j Represents the total number of laser powder block central points with the change of on and off in the jth frame anti-counterfeiting area image, bc represents the total number of frame numbers in the set B, and B=F n L, m 1 The threshold value of the number of the center points of the tone change is taken as the average value, m, of the total number of the center points of the laser powder blocks with the tone change when a plurality of real laser labels respectively generate e tone changes 2 The threshold value of the number of the central points of the on-off change is taken as the average value, m, of the total number of the central points of the laser powder blocks with brightness change when a plurality of real laser labels respectively change the brightness for e times 3 The threshold value of the number of the central points of single tone change is taken as the average value of the total number of the central points of the laser powder blocks, m, of each time the actual laser label generates tone change 4 Taking the average value of the frame numbers of image frames of an anti-counterfeiting area with color tone change and brightness change simultaneously when a plurality of real laser labels respectively generate e brightness change and e color tone change as a frame number threshold value, wherein e is a change number threshold value, and the threshold value is in [6,12]Is selected from a range of values.
10. The method for identifying the quasi-dynamic laser anti-counterfeit label based on the feature inconsistency according to claim 1, wherein the duplicate inconsistency condition in the step 7 is that the number of subareas, which are judged by the color distribution feature in the anti-counterfeit area of the laser anti-counterfeit label through the inconsistency, exceeds 2, and the color change feature of the anti-counterfeit area of the laser anti-counterfeit label is judged by the inconsistency.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211061606.7A CN115423771B (en) | 2022-09-01 | 2022-09-01 | Quasi-dynamic laser anti-counterfeit label identification method based on feature non-uniformity |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211061606.7A CN115423771B (en) | 2022-09-01 | 2022-09-01 | Quasi-dynamic laser anti-counterfeit label identification method based on feature non-uniformity |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115423771A CN115423771A (en) | 2022-12-02 |
| CN115423771B true CN115423771B (en) | 2023-06-30 |
Family
ID=84200445
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211061606.7A Active CN115423771B (en) | 2022-09-01 | 2022-09-01 | Quasi-dynamic laser anti-counterfeit label identification method based on feature non-uniformity |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115423771B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116092061B (en) * | 2023-04-12 | 2023-09-22 | 深圳企业云科技股份有限公司 | Label verification method and device |
| CN116758528B (en) * | 2023-08-18 | 2023-11-03 | 山东罗斯夫新材料科技有限公司 | Acrylic emulsion color change identification method based on artificial intelligence |
| CN117475356B (en) * | 2023-12-25 | 2024-03-29 | 华南理工大学 | Control method and system for preparing invisible double anti-counterfeit label by corrugated case |
| CN118190954B (en) * | 2024-05-15 | 2024-08-16 | 青岛澳科顺诚包装有限公司 | Laser film anti-counterfeiting pattern monitoring quality inspection system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111222612A (en) * | 2020-03-18 | 2020-06-02 | 杭州沃朴物联科技有限公司 | Laser anti-counterfeit label and manufacturing method thereof and identification and counterfeit verification method |
| CN113808472A (en) * | 2021-09-14 | 2021-12-17 | 山东泰宝信息科技集团有限公司 | Holographic laser variable verification code anti-counterfeit label and manufacturing method and using method thereof |
| CN114203016A (en) * | 2021-11-23 | 2022-03-18 | 中山国安火炬科技发展有限公司 | Anti-counterfeit label and application thereof |
| CN114663271A (en) * | 2022-03-28 | 2022-06-24 | 美盈森集团股份有限公司 | Processing method for anti-counterfeit label image symbolization before printing |
Family Cites Families (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103712919B (en) * | 2013-11-02 | 2015-12-30 | 深圳市科彩印务有限公司 | A kind of sign of native-surface laser paper color characteristics and detection method |
| CN106650669A (en) * | 2016-12-27 | 2017-05-10 | 重庆邮电大学 | Face recognition method for identifying counterfeit photo deception |
| CN210284932U (en) * | 2017-12-30 | 2020-04-10 | 山东泰宝防伪制品有限公司 | SV code anti-fake packaging box |
| CN110533704B (en) * | 2019-07-22 | 2022-11-11 | 杭州沃朴物联科技有限公司 | Method, device, equipment and medium for identifying and verifying ink label |
| CN110428028B (en) * | 2019-07-22 | 2023-01-10 | 杭州沃朴物联科技有限公司 | Identification and verification method, device, equipment and medium based on quasi-dynamic laser label |
| CN110796221B (en) * | 2019-10-18 | 2022-09-02 | 周晓明 | Anti-counterfeit label generation method, anti-counterfeit label verification method and system and anti-counterfeit label |
| CN111160215B (en) * | 2019-12-25 | 2024-01-12 | Tcl华星光电技术有限公司 | Brightness regulator and method for image mark |
| CN111242259B (en) * | 2020-01-08 | 2021-03-16 | 浙江码尚科技股份有限公司 | Intelligent anti-counterfeiting code encoding and decoding method and system |
| CN111414779B (en) * | 2020-05-22 | 2022-07-26 | 杭州沃朴物联科技有限公司 | Anti-counterfeit label identification method and device |
| CN112329673B (en) * | 2020-11-12 | 2024-05-17 | 北京文通科技有限公司 | Anti-fake fiber identification method, anti-fake identification method, device and storage medium |
-
2022
- 2022-09-01 CN CN202211061606.7A patent/CN115423771B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111222612A (en) * | 2020-03-18 | 2020-06-02 | 杭州沃朴物联科技有限公司 | Laser anti-counterfeit label and manufacturing method thereof and identification and counterfeit verification method |
| CN113808472A (en) * | 2021-09-14 | 2021-12-17 | 山东泰宝信息科技集团有限公司 | Holographic laser variable verification code anti-counterfeit label and manufacturing method and using method thereof |
| CN114203016A (en) * | 2021-11-23 | 2022-03-18 | 中山国安火炬科技发展有限公司 | Anti-counterfeit label and application thereof |
| CN114663271A (en) * | 2022-03-28 | 2022-06-24 | 美盈森集团股份有限公司 | Processing method for anti-counterfeit label image symbolization before printing |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115423771A (en) | 2022-12-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115423771B (en) | Quasi-dynamic laser anti-counterfeit label identification method based on feature non-uniformity | |
| CN110119741B (en) | Card image information identification method with background | |
| CN107545239B (en) | Fake plate detection method based on license plate recognition and vehicle characteristic matching | |
| CN107301405A (en) | Method for traffic sign detection under natural scene | |
| Wen et al. | An algorithm for license plate recognition applied to intelligent transportation system | |
| CN102426649B (en) | Simple steel seal digital automatic identification method with high accuracy rate | |
| CN109034145B (en) | Bank card number identification method based on OpenCV | |
| CN111753692B (en) | Target object extraction method, product detection method, device, computer and medium | |
| CN110838119A (en) | Human face image quality evaluation method, computer device and computer readable storage medium | |
| CN115244542A (en) | Method and device for verifying authenticity of product | |
| Jagannathan et al. | License plate character segmentation using horizontal and vertical projection with dynamic thresholding | |
| CN110533704B (en) | Method, device, equipment and medium for identifying and verifying ink label | |
| CN113392664B (en) | Anti-counterfeiting two-dimensional code generation method, anti-counterfeiting method and device | |
| CN110163152A (en) | Antiforge recognizing method, method for anti-counterfeit, system, device end and storage medium | |
| CN110689003A (en) | Low-illumination imaging license plate recognition method and system, computer equipment and storage medium | |
| CN116416268B (en) | Method and device for detecting edge position of lithium battery pole piece based on recursion dichotomy | |
| CN106407959A (en) | Low-illumination complicated background license plate positioning method based on wavelet transform and SVM | |
| Agarwal et al. | An efficient algorithm for automatic car plate detection & recognition | |
| CN118275449A (en) | Copper strip surface defect detection method, device and equipment | |
| CN113435219B (en) | Anti-counterfeiting detection method and device, electronic equipment and storage medium | |
| Julliand et al. | Automated image splicing detection from noise estimation in raw images | |
| CN114241463A (en) | Signature verification method and device, computer equipment and storage medium | |
| Vargas et al. | A shadow removal algorithm for vehicle detection based on reflectance ratio and edge density | |
| JP3416058B2 (en) | Character extraction method of gray image and recording medium recording the program | |
| CN111275687B (en) | Fine-grained image stitching detection method based on connected region marks |
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 |