CN115393855A - License plate product quality detection method, system and equipment - Google Patents
License plate product quality detection method, system and equipment Download PDFInfo
- Publication number
- CN115393855A CN115393855A CN202210137467.5A CN202210137467A CN115393855A CN 115393855 A CN115393855 A CN 115393855A CN 202210137467 A CN202210137467 A CN 202210137467A CN 115393855 A CN115393855 A CN 115393855A
- Authority
- CN
- China
- Prior art keywords
- license plate
- image
- character
- detection
- target image
- 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 118
- 230000007547 defect Effects 0.000 claims abstract description 71
- 238000000034 method Methods 0.000 claims abstract description 47
- 230000009466 transformation Effects 0.000 claims description 15
- 238000013145 classification model Methods 0.000 claims description 11
- 230000011218 segmentation Effects 0.000 claims description 11
- 238000013528 artificial neural network Methods 0.000 claims description 9
- 238000000605 extraction Methods 0.000 claims description 9
- 238000004519 manufacturing process Methods 0.000 claims description 9
- 239000011159 matrix material Substances 0.000 claims description 9
- 238000010586 diagram Methods 0.000 claims description 8
- 230000004075 alteration Effects 0.000 claims description 6
- 230000006740 morphological transformation Effects 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 6
- 238000012937 correction Methods 0.000 claims description 4
- 238000010606 normalization Methods 0.000 abstract description 5
- 230000002349 favourable effect Effects 0.000 abstract 1
- 230000006870 function Effects 0.000 description 14
- 230000008569 process Effects 0.000 description 14
- 238000012545 processing Methods 0.000 description 9
- 230000008859 change Effects 0.000 description 8
- 238000011156 evaluation Methods 0.000 description 8
- 238000003860 storage Methods 0.000 description 8
- 238000006243 chemical reaction Methods 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 6
- 238000004364 calculation method Methods 0.000 description 5
- 230000009286 beneficial effect Effects 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 3
- 238000012216 screening Methods 0.000 description 3
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 description 2
- 230000002950 deficient Effects 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- XAGFODPZIPBFFR-UHFFFAOYSA-N aluminium Chemical compound [Al] XAGFODPZIPBFFR-UHFFFAOYSA-N 0.000 description 1
- 229910052782 aluminium Inorganic materials 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 239000012535 impurity Substances 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 230000000877 morphologic effect Effects 0.000 description 1
- 230000004660 morphological change Effects 0.000 description 1
- 239000003973 paint Substances 0.000 description 1
- 238000010422 painting Methods 0.000 description 1
- 238000003825 pressing Methods 0.000 description 1
- 238000001303 quality assessment method Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
- 238000011179 visual inspection Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06K—GRAPHICAL DATA READING; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
- G06K19/00—Record carriers for use with machines and with at least a part designed to carry digital markings
- G06K19/06—Record carriers for use with machines and with at least a part designed to carry digital markings characterised by the kind of the digital marking, e.g. shape, nature, code
- G06K19/06009—Record carriers for use with machines and with at least a part designed to carry digital markings characterised by the kind of the digital marking, e.g. shape, nature, code with optically detectable marking
- G06K19/06037—Record carriers for use with machines and with at least a part designed to carry digital markings characterised by the kind of the digital marking, e.g. shape, nature, code with optically detectable marking multi-dimensional coding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Biophysics (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
Abstract
The present disclosure relates to a method, a system and a device for detecting the quality of a license plate product, wherein the method comprises the following steps: acquiring an original image containing a license plate to be detected; extracting an image of a license plate region to be detected in an original image as a target image; correcting the target image; identifying a unique label in a target image, and acquiring unique identification information of the license plate to be detected; respectively carrying out character and frame integrity detection, defect detection and key size detection on the target image; and outputting a detection result according to the character and frame integrity detection, defect detection and key size detection conclusion. The system is used for realizing the method, and the equipment is used for executing the method. The method can improve the detection efficiency of the license plate quality, save labor, effectively avoid the influence of artificial subjective factors on the detection result, improve the reliability, accuracy and uniformity of the license plate quality detection result, contribute to the standardization and normalization of the license plate quality and be favorable for improving the delivery quality of the license plate products.
Description
Technical Field
The disclosure relates to the technical field of license plate product quality detection, in particular to a license plate product quality detection method, system and equipment.
Background
The manufacturing process of the automobile license plate is completed by the steps of pasting a film on an aluminum plate, painting, pressing characters, erasing characters and the like. In the manufacturing process of the license plate, characters are incomplete, the size is not qualified, and defects such as scratches, scratches and stains are also formed on the surface of the license plate, so that the quality of the license plate product is not qualified. At present, before the automobile license plate is delivered from a factory after being produced, workers usually observe and measure the appearance and the size of the license plate manually to evaluate whether the license plate is qualified or not, and the manual evaluation mode has low efficiency and influences the overall production efficiency of the license plate on the one hand, and has larger subjective factor influence on the manual evaluation mode on the other hand, and qualified judgment standards of different workers are different, so that the reliability, the accuracy and the uniformity of a license plate quality detection result are poor, and the delivery quality of the license plate is influenced.
Disclosure of Invention
In order to solve the problems in the prior art, the present disclosure aims to provide a method, a system and a device for detecting the quality of a license plate product. The vehicle license plate quality detection method can replace manual vehicle license plate quality assessment, can improve vehicle license plate quality detection efficiency, saves manpower, can effectively avoid influence of human subjective factors on detection results, improves reliability, accuracy and uniformity of vehicle license plate quality detection results, is beneficial to standardization and normalization of vehicle license plate quality, and is beneficial to improvement of delivery quality of vehicle license plate products.
The invention discloses a license plate product quality detection method, which comprises the following steps:
s01, acquiring an original image containing a license plate to be detected;
s02, extracting an image of a license plate region to be detected in the original image as a target image;
s03, correcting the target image;
s04, identifying the unique label in the target image, and acquiring unique identification information of the license plate to be detected;
s05, respectively carrying out character and frame integrity detection, defect detection and key size detection on the target image;
and S06, outputting a detection result according to the character and frame integrity detection, the defect detection and the key size detection conclusion.
Preferably, after the step S06, the method further comprises:
and S07, constructing a database for storing the detection data.
Preferably, in step S02, extracting the image of the license plate region to be detected in the original image as the target image specifically includes:
and sequentially carrying out gray level transformation, threshold segmentation, morphological transformation and region segmentation processing on the original image to obtain an ROI (region of interest) serving as a target image.
Preferably, in step S03, the correcting the target image specifically includes:
s031, obtaining the coordinates of each vertex of the license plate in the standard license plate image and recording as the coordinates of the standard vertex;
s032, obtaining coordinates of each vertex of the license plate in the target image and recording the coordinates as coordinates of the target vertex;
and S033, sequentially performing matrix transformation operation and perspective transformation on the standard vertex coordinates and the target vertex coordinates, and correcting the target image according to the standard vertex coordinates and the target vertex coordinates.
Preferably, in step S04, the unique tag is a two-dimensional code.
Preferably, in the step S05, the detecting integrity of characters and frames of the target image includes:
s051a, obtaining a standard character frame image;
s052a, extracting a character frame image in the target image;
s053a, comparing the obtained character frame image with the standard character frame image to obtain a character frame image aberration;
s054a, calculating the integrity of the character frame according to the aberration of the character frame diagram, comparing the integrity of the character frame with a preset integrity threshold, and outputting the detection result of the integrity of the character and the frame according to the comparison result.
Preferably, in step S05, the performing defect detection on the target image includes:
s051b, collecting a plurality of license plate images respectively containing various production defects, and constructing a defect data set;
s052b, training a neural network by using the defect data set to construct a classification model about the license plate defects;
and S053b, pre-selecting a defect area in the target image, inputting the pre-selected target image into the classification model for defect classification, and outputting a defect detection result according to a defect classification result.
Preferably, in step S05, the detecting the key size of the target image includes:
s051c, obtaining a ratio value of the size of the characters in the standard license plate image to the size of the license plate, and recording the ratio value as a standard ratio value;
s052c, extracting the character images in the target image, and segmenting and positioning the character images;
s053c, acquiring the number of pixel points of each character image, calculating the character height, the character width and the adjacent character spacing of each character image according to the standard proportional value, comparing the obtained character height, character width and adjacent character spacing with the standard character size, and outputting a key size detection result according to the comparison result.
The present disclosure also provides a license plate product quality detection system, including:
the image acquisition module is used for acquiring an original image containing a license plate to be detected;
the extraction module is used for extracting an image of a license plate region to be detected in the original image as a target image;
the correction module is used for correcting the target image;
the identification module is used for identifying the unique label in the target image and acquiring the unique identification information of the license plate to be detected;
the detection module is used for respectively carrying out character and frame integrity detection, defect detection and key size detection on the target image;
and the output module is used for outputting a detection result according to the character and frame integrity detection, the defect detection and the key size detection conclusion.
The present disclosure further provides a computer device, which includes a processor and a memory, and is characterized in that at least one instruction or at least one program is stored in the memory, and when the at least one instruction or the at least one program is loaded by the processor, the method for detecting the quality of the license plate product as described above is performed.
The method, the system and the equipment for detecting the quality of the license plate product have the advantages that the method, the system and the equipment are combined with an image processing technology and a machine learning technology, the license plate product can be automatically identified, warehoused and subjected to quality detection, the manual labor is liberated, the efficiency of license plate quality detection is greatly improved, and the automation and intelligentization degree in the license plate production process are improved. Meanwhile, the influence of human subjective factors on the detection result can be effectively avoided by adopting a machine evaluation mode, so that each license plate product is detected under the same detection standard, the reliability, the accuracy and the uniformity of the license plate quality detection result can be improved, the standardization and the normalization of the license plate quality are facilitated, and the ex-factory quality of the license plate product is facilitated to be improved.
Drawings
FIG. 1 is a flow chart illustrating steps of a method for detecting the quality of a license plate product according to the present disclosure;
FIG. 2 is an interface diagram of the present disclosure identifying unique tags in a target image;
FIG. 3 is an interface diagram of the character and bezel integrity detection of the present disclosure;
FIG. 4 is an interface diagram of defect detection of the present disclosure;
FIG. 5 is an interface diagram of critical dimension inspection of the present disclosure;
FIG. 6 is a database management interface diagram of the present disclosure.
Detailed Description
As shown in fig. 1, the method for detecting the quality of a license plate product according to the present disclosure includes the following steps:
s01, acquiring an original image containing a license plate to be detected; specifically, its detection station is made the mode of visual inspection platform usually, and it fixes the license plate that awaits measuring from the mode that the back adsorbs through the sucking disc, sets up high definition camera in order to shoot the original image of obtaining the license plate that awaits measuring in the top of detecting the station, still need be equipped with auxiliary light source so that illumination is sufficient even usually, ensures that the formation of image effect is clear. The high-definition camera is usually in communication connection with an upper Computer, such as a PC (Personal Computer), so as to transmit the original image including the license plate to be detected, which is obtained by shooting, to the upper Computer, and the original image is processed and operated by the upper Computer.
S02, extracting an image of a license plate region to be detected in the original image as a target image; specifically, the original image includes other objects such as a visual detection table, which are not beneficial to subsequent license plate image processing, so that interfering objects in the image need to be removed, and the steps of gray level transformation, threshold segmentation (i.e., binarization processing), morphological transformation, and region segmentation extraction, which are commonly used in the image processing technology, are adopted here to extract an image of a license plate region to be detected in the original image, so as to obtain a region of interest (ROI), which is a target image to be obtained.
S03, correcting the target image; in this step, the size ratio of the target image needs to be corrected, so that the subsequent detection result is accurate. The method comprises the steps of firstly obtaining a license plate image in a standard license plate image, converting the license plate image into a gray image, then carrying out edge extraction on the obtained gray image, thinning and extracting a unique edge, wherein the license plate image is usually a regular rectangle, the process is convenient to carry out, traversing and searching coordinates of four vertexes of the image (the central point of the image can be used as an origin of a coordinate axis or the lower left vertex can be used as the origin of a coordinate) after the unique edge extraction is finished, recording the coordinates of the four vertexes obtained, recording the coordinates as standard vertex coordinates, and storing the standard vertex coordinates in a database.
And then obtaining the coordinates of each vertex of the license plate in the target image, specifically obtaining the coordinates of each vertex of the license plate in the target image in the same way of edge extraction and traversal searching, and marking the coordinates as the coordinates of the target vertex.
And calling a pre-stored standard vertex coordinate, carrying out matrix transformation operation according to a find homographic function by combining the obtained target vertex coordinate to obtain a transformation matrix H, finishing Perspective change of the transformation matrix H through a warp Perspective function, and finishing correction of a target image by referring to a standard license plate image. The specific calculation process is as follows:
an affine change matrix H (the matrix can be decomposed into a linear transformation part and a perspective transformation part) is obtained according to a find homograph function, the warpAffine function receives the H matrix to complete affine change on an image, the proportion of a transformed license plate ROI area and the standard license plate pixel can be different, and therefore the transformed image needs to be further corrected by combining the standard license plate (namely, the image is set to be the specific pixel proportion 4400 x 1400 according to a resize function).
The specific perspective change operation process and principle are as follows:
assuming that one coordinate of the original image is (u, v) and the coordinate after perspective change is (x, y), the perspective change can be expressed as formula:
in the change matrix
In
Representing a linear transformation, [ a ] 31 a 32 ]A translation is indicated and is indicated by,
a perspective is generated. The coordinate value after perspective change can be obtained by the transformation formula before rewriting, and the calculation formula is shown as the following formula:
through the calculation process, the target image can be corrected by referring to the standard vertex coordinates.
S04, identifying the unique label in the target image, and acquiring unique identification information of the license plate to be detected; when the license plates are produced, each license plate corresponds to a unique identification label which can be printed on the front surface of the license plate and is used for pointing to the unique identification information corresponding to the license plate when being identified. In a specific embodiment, the unique Identification tag may be a barcode, a two-dimensional Code (QR Code), an RFID (Radio Frequency Identification) tag, or other electronic tags that can be easily identified, the unique Identification information is usually a string of unique continuous numbers and/or characters, for example, a commonly used two-dimensional Code, and the two-dimensional Code image is printed at the upper left corner of the license plate. After the two-dimensional Code image is obtained, HSV (Hue, saturation, value brightness) conversion, HSV threshold segmentation, morphological transformation, and QR Code Detector (an open source module for two-dimensional Code recognition) recognition are sequentially performed on the two-dimensional Code image, unique recognition information pointed by the two-dimensional Code is obtained, the unique recognition information is used for recording the information of the license plate to be detected, and the information is correspondingly matched with a subsequent detection result, as shown in fig. 2, after the two-dimensional Code in the image is recognized, a number and a letter string shown in the figure are obtained as the unique recognition information. The specific two-dimensional code identification process is as follows:
because the license plate two-dimensional code is under blue paint background (the license plate background is blue) and is smaller, the license plate image needs to be preprocessed before the two-dimensional code is used for positioning and decoding, so that the two-dimensional code is more obvious, and the positioning and decoding of the two-dimensional code are facilitated.
The main pretreatment process is as follows:
and (3) sequentially carrying out HSV conversion, HSV threshold segmentation and morphological transformation on the original image. Wherein HSV color space is a color space representation of values in RGB in an inverted cone. The color is represented by the generatrix of a cone, the purity of the color is represented by the horizontal axis, and the shading is represented by the vertical axis. The following effect graph can be obtained through the maximum contrast and the brightness in the process of processing the two-dimensional code picture of the license plate, and the effect graph only containing black and white can be obtained through setting the saturation to be 0. Burrs and a few impurities existing in the processed two-dimensional code picture can be removed through morphological change.
After the OpenCV4.0 version, a two-dimensional code positioning decoding function is added. The QRCODDetector type two-dimensional code positioning and decoding module is included and provided to effectively solve the problem of two-dimensional code identification. The QRCODEDeDetector structure mainly comprises two core functions of detect () and decode (). The detect () function is mainly used for positioning, and decode () completes decoding the two-dimensional code and returns a character string containing the two-dimensional code content. The license plate two-dimensional code positioning and decoding process comprises the following steps:
two-dimensional code region picture obtained by image segmentation → detect () two-dimensional code positioning → decode () decoding → obtaining identification result
The two-dimensional code positioning principle is that the ratio of black and white intervals of the positioning pattern is fixed: 1. I.e., horizontal or vertical scanning, to obtain five line segments of similar scale, it is considered as part of the positioning pattern.
In the positioning process, three positioning marks are firstly found through vertical and horizontal scanning, then the sequence of the positioning marks is determined through the relative positions of the positioning marks, and the range and the direction of the two-dimensional code can be determined according to the three positioning representations after the sequence is known.
And S05, respectively carrying out character and frame integrity detection, defect detection and key size detection on the target image, wherein the three detection processes can be carried out synchronously or step by step according to any sequence.
The detection of the integrity of the characters and the frame of the target image specifically comprises the following steps:
s051a, obtaining a standard character frame image; specifically, an image of a standard license plate is obtained, the image is extracted through gray level conversion, threshold segmentation, morphological conversion and region segmentation in sequence, and then characters and frame images are extracted, wherein the specific extraction mode is that the area occupied by a single character in a standardized license plate image (4400 x 1400) is within a certain range, so that the image of the single character can be obtained through screening the area of the character region, and the image of each character in the license plate image can be extracted independently by extracting the image of each character one by one. The extraction mode of the frame image is similar to that of the character image.
S052a, extracting the character frame image in the target image by the method in the previous step,
s053a, comparing the obtained character frame image with the standard character frame image to obtain a character frame image aberration; the specific calculation process is as follows:
in OpenCV, the absdiff function provides a method for calculating an image difference map, and the principle of the method is to calculate an absolute value of a difference of each corresponding pixel in two pictures (because the input picture is a black-and-white binary image at this time, a complete border is a black background and a white border, and a defective border is discontinuous at a defect, i.e., the function can be used to complete detection of a difference pixel), and a pixel point calculation formula is as follows:
dst[i]＝|src1[i]-src2[i]|
and finally, evaluating the defects by counting the proportion of the number of white pixel points (namely difference pixels) of the difference image in the standard character side frame image pixel points (white) with the same pixel size (by correcting the image after resize before, the counted total number of the pixel number and the occupied area of the part of pixels are in a linear relation, namely the counted number of the pixels and the area have the same function).
And S054a, calculating the integrity of the character frame according to the aberration of the character frame diagram, comparing the integrity of the character frame with a preset integrity threshold value, if 95%, and outputting the integrity detection results of the character and the frame according to the comparison result, namely outputting that the integrity of the character frame is greater than or equal to 95% as that the integrity detection results of the character and the frame are qualified, otherwise, as that the integrity detection results of the character and the frame are unqualified.
The defect detection of the target image comprises the following steps:
s051b, collecting a plurality of license plate images respectively containing various production defects, and constructing a defect data set; a plurality of license plate images containing different production defects, such as scratches, scratches and stains, and the same defects are distributed at different positions of a license plate are collected and taken as a defect data set.
S052b, training a neural network by using the defect data set to construct a classification model about the license plate defects; specifically, the defect data set is used as input of a neural network, the neural network is trained to construct a classification model about the license plate defect, the neural network can specifically select a Mobile Net V3 (a lightweight network), a ResNet (a residual error network), a VGGNet (a convolutional neural network) or other neural networks with training learning capacity, and the output result of the classification model is continuously close to the real defect classification by continuously inputting the defective license plate image, so that the classification accuracy of the classification model is continuously improved. It should be noted that, the classification model with the image classification capability is obtained by adopting neural network training, which belongs to the existing machine learning technology, and the specific model training method is not improved in the present disclosure.
S053b, preselecting a defect region in the target image, inputting the preselected target image into the classification model for defect classification, and outputting a defect detection result according to the defect classification result, specifically, as shown in fig. 4, the classification model outputs a defect type and a defect evaluation result corresponding to the target image after processing the target image according to the input target image, and outputs whether the defect detection result is qualified according to the defect evaluation result.
The defect area of the target image is preselected by utilizing opencv to perform pre-detection and screen a license plate area possibly with random defects, and the specific process is as follows:
random defects mainly include scratches, stains and scratches. The main detection mode is realized by Blob analysis.
The main detection method of the scratch defects is Hall line transformation, and the main principle is that one line in an image can be detected by the number of sinusoids intersected at one point in a theta-r plane, and the more sinusoids are intersected at one point, the straight line represented by the point can be shown to exist in the image, so that the straight line in the image is detected.
The main detection method of the stain defects is to obtain the stain defects by threshold segmentation and then directly screening according to the area.
The main detection method of the scratch defects is to adopt a closed operation (MORPH _ CLOSE) in morphological transformation, eliminate and combine blanks in scratches into a single area, and then extract scratches through area screening.
Outputting the defect type and defect evaluation result corresponding to the target image specifically comprises:
the defect types are obtained by classifying the neural network mobilenet v3, specifically, in defect assessment, the scratch type defects take the long edge of a scratch external minimum rectangle as a judgment basis, the width larger than the standard character I is unqualified, stain and scratch assessment take occupied pixel points as a reference, and the number of the pixel points is in direct proportion to the occupied area after the license plate is corrected. The standard is that the area of the single defect is larger than half of the area of the license plate separator, namely the area is unqualified.
And finally, the quality of the license plate takes the ratio of the total area with the detected defects as output, and if unqualified defects occur, the license plate is directly output to be unqualified.
The detecting the key size of the target image comprises the following steps:
s051c, obtaining a ratio value of the size of the characters in the standard license plate image to the size of the license plate, and recording the ratio value as a standard ratio value; the method specifically comprises the following steps:
through the image processing process, extracting and obtaining each character image of the standard license plate, calculating and obtaining the height and the width of the standard character, and respectively comparing the height of the standard character with the height and the width of the standard license plate to obtain a standard height proportion value and a standard width proportion value;
s052c, extracting the character images in the target image, and segmenting and positioning the character images;
s053c, obtaining the pixel number of each character image, calculating the character height, the character width and the adjacent character interval of each character image according to the standard proportional value, comparing the obtained character height, character width and adjacent character interval with the standard character size, and outputting a key size detection result according to the comparison result.
Specifically, a proportion value is obtained according to the size of the standard license plate and the obtained pixel size of the image. A boundingselect function is provided in OpenCV to extract a minimum positive rectangle of license plate characters, the positions of the width, height and central point of the characters (the width, height and relative positions are pixel coordinate values) can be returned by utilizing the Rect type obtained by the boundingselect function, and the actual width, height and position coordinates of the characters are calculated by combining the obtained proportions.
For example, the actual size of the character width and height is calculated as follows:
character actual height (width) = (140 × pixel value of character height (width)/pixel value of license plate height (width) (where 140mm is license plate standard height).
Wherein, the character width and height size (after conversion) is qualified if the difference between the character width and height size and the corresponding standard character width and height size is less than 5%, and the character spacing (after conversion) error is less than 2mm, namely the character spacing is qualified (the standard character spacing is 12 mm).
And S06, outputting a detection result according to the character and frame integrity detection, the defect detection and the key size detection conclusion, specifically, outputting the detection result to be qualified only when the character and frame integrity detection, the defect detection and the key size detection pass, and otherwise, outputting the detection result to be unqualified.
The license plate product identification and storage method combines the image processing technology and the machine learning technology, can automatically identify and store license plate products and detect the quality of the license plate products, relieves the manpower, greatly improves the efficiency of license plate quality detection, and improves the automation and intelligence degree in the license plate production process. Meanwhile, the influence of human subjective factors on the detection result can be effectively avoided by adopting a machine evaluation mode, so that each license plate product is detected under the same detection standard, the reliability, the accuracy and the uniformity of the license plate quality detection result can be improved, the standardization and the normalization of the license plate quality are facilitated, and the factory quality of the license plate product is improved.
Further, in this embodiment, after the step S06, the method further includes:
and S07, constructing a database for storing the detection data. In a specific embodiment, the database may adopt an access database, and the operations including querying, modifying, adding, deleting, and the like are completed on the database through the Ole Db Data Adapter database operation class of C #, so that a user can process Data in the database conveniently.
This embodiment still provides a license plate product quality detection system, includes:
the image acquisition module is used for acquiring an original image containing a license plate to be detected;
the extraction module is used for extracting an image of a license plate region to be detected in the original image as a target image;
the correction module is used for correcting the target image;
the identification module is used for identifying the unique label in the target image and acquiring the unique identification information of the license plate to be detected;
the detection module is used for respectively carrying out character and frame integrity detection, defect detection and key size detection on the target image;
and the output module is used for outputting a detection result according to the character and frame integrity detection, the defect detection and the key size detection conclusion.
The operations respectively executed by the modules correspond to the steps of the license plate product quality detection method of the foregoing embodiment one to one, and can be understood with reference to the above description, which is not repeated herein.
The license plate product quality detection system can replace manual evaluation of license plate quality, can improve license plate quality detection efficiency, save manpower, effectively avoid influence of human subjective factors on detection results, improve reliability, accuracy and uniformity of the license plate quality detection results, contribute to standardization and normalization of license plate quality, and contribute to improvement of ex-factory quality of license plate products.
The embodiment of the present disclosure further provides a computer device, which includes a processor and a memory, where at least one instruction or at least one program is stored in the memory, and when the at least one instruction or the at least one program is loaded by the processor, the method for detecting the quality of the license plate product is performed. The memory may be used to store software programs and modules, and the processor may execute various functional applications by executing the software programs and modules stored in the memory. The memory can mainly comprise a program storage area and a data storage area, wherein the program storage area can store an operating system, application programs needed by functions and the like; the storage data area may store data created according to use of the apparatus, and the like. Further, the memory may include high speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other volatile solid state storage device. Accordingly, the memory may also include a memory controller to provide the processor access to the memory.
The method embodiments provided by the embodiments of the present disclosure may be executed in a computer terminal, a server or a similar computing device, that is, the computer device may include a computer terminal, a server or a similar computing device. The internal structure of the computer device may include, but is not limited to: a processor, a network interface, and a memory. The processor, the network interface and the memory in the computer device may be connected by a bus or other means.
The processor (or CPU) is a computing core and a control core of the computer device. The network interface may optionally include a standard wired interface, a wireless interface (e.g., WI-FI, mobile communication interface, etc.). Memory (Memory) is a Memory device in a computer device used to store programs and data. It is understood that the memory herein may be a high-speed RAM storage device, or may be a non-volatile storage device (non-volatile memory), such as at least one magnetic disk storage device; optionally, at least one memory device located remotely from the processor. The memory provides storage space that stores an operating system of the electronic device, which may include, but is not limited to: a Windows system (an operating system), linux (an operating system), android (Android, a mobile operating system) system, IOS (a mobile operating system) system, and the like, which are not limited by the present disclosure; also, one or more instructions, which may be one or more computer programs (including program code), are stored in the memory space and are adapted to be loaded and executed by the processor. In the embodiment of the present specification, the processor loads and executes one or more instructions stored in the memory to implement the method for detecting the quality of the license plate product described in the embodiment of the method.
In the description of the present disclosure, it is to be understood that the directions or positional relationships indicated by the directional terms such as "front, rear, upper, lower, left, right", "lateral, vertical, horizontal" and "top, bottom", etc., are generally based on the directions or positional relationships shown in the drawings for the convenience of description and simplicity of description, and in the case of not being described to the contrary, these directional terms are not intended to indicate and imply that the device or element so referred to must have a particular orientation or be constructed and operated in a particular orientation, and therefore, should not be considered as limiting the scope of the present disclosure.
Various other modifications and changes may be made by those skilled in the art based on the above-described technical solutions and concepts, and all such modifications and changes should fall within the scope of the claims of the present disclosure.
Claims (10)
1. A method for detecting the quality of a license plate product is characterized by comprising the following steps:
s01, acquiring an original image containing a license plate to be detected;
s02, extracting an image of a license plate region to be detected in the original image as a target image;
s03, correcting the target image;
s04, identifying the unique label in the target image, and acquiring unique identification information of the license plate to be detected;
s05, respectively carrying out character and frame integrity detection, defect detection and key size detection on the target image;
and S06, outputting a detection result according to the character and frame integrity detection, the defect detection and the key size detection conclusion.
2. The method for detecting the quality of the license plate product of claim 1, further comprising, after the step S06:
and S07, constructing a database for storing the detection data.
3. The method for detecting the quality of the license plate product of claim 1, wherein in the step S02, the step of extracting the image of the license plate region to be detected in the original image as the target image specifically comprises:
and sequentially carrying out gray level transformation, threshold segmentation, morphological transformation and region segmentation on the original image to obtain an ROI (region of interest) region as a target image.
4. The method for detecting the quality of the license plate product of claim 1, wherein the step S03 of correcting the target image specifically comprises:
s031, obtain the license plate in the standard license plate picture each vertex coordinate, record as the standard vertex coordinate;
s032, obtaining coordinates of each vertex of the license plate in the target image and recording the coordinates as coordinates of the target vertex;
and S033, sequentially performing matrix transformation operation and perspective transformation on the standard vertex coordinates and the target vertex coordinates, and correcting the target image according to the standard vertex coordinates and the target vertex coordinates.
5. The method for detecting the quality of the license plate product of claim 1, wherein in the step S04, the unique tag is a two-dimensional code.
6. The method for detecting the quality of the license plate product of claim 1, wherein in the step S05, the step of detecting the integrity of the characters and the frame of the target image comprises the following steps:
s051a, obtaining a standard character frame image;
s052a, extracting a character frame image in the target image;
s053a, comparing the obtained character frame image with the standard character frame image to obtain the aberration of the character frame image;
s054a, calculating the integrity of the character frame according to the aberration of the character frame diagram, comparing the integrity of the character frame with a preset integrity threshold, and outputting the detection result of the integrity of the character and the frame according to the comparison result.
7. The method for detecting the quality of the license plate product of claim 1, wherein in the step S05, the detecting the defect of the target image comprises:
s051b, collecting a plurality of license plate images respectively containing various production defects, and constructing a defect data set;
s052b, training a neural network by using the defect data set to construct a classification model about the license plate defects;
and S053b, preselecting a defect area in the target image, inputting the preselected target image into the classification model for defect classification, and outputting a defect detection result according to a defect classification result.
8. The method for detecting the quality of the license plate product of claim 1, wherein in the step S05, the detecting the key size of the target image comprises:
s051c, obtaining a ratio value of the size of the characters in the standard license plate image to the size of the license plate, and recording the ratio value as a standard ratio value;
s052c, extracting the character images in the target image, and segmenting and positioning the character images;
s053c, obtaining the pixel number of each character image, calculating the character height, the character width and the adjacent character interval of each character image according to the standard proportional value, comparing the obtained character height, character width and adjacent character interval with the standard character size, and outputting a key size detection result according to the comparison result.
9. A license plate product quality detection system, comprising:
the image acquisition module is used for acquiring an original image containing a license plate to be detected;
the extraction module is used for extracting an image of a license plate region to be detected in the original image as a target image;
the correction module is used for correcting the target image;
the identification module is used for identifying the unique label in the target image and acquiring the unique identification information of the license plate to be detected;
the detection module is used for respectively carrying out character and frame integrity detection, defect detection and key size detection on the target image;
and the output module is used for outputting a detection result according to the character and frame integrity detection, the defect detection and the key size detection conclusion.
10. A computer device comprising a processor and a memory, wherein at least one instruction or at least one program is stored in the memory, and when the at least one instruction or the at least one program is loaded by the processor, the method for detecting the quality of a license plate product according to any one of claims 1 to 8 is performed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210137467.5A CN115393855A (en) | 2022-02-15 | 2022-02-15 | License plate product quality detection method, system and equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210137467.5A CN115393855A (en) | 2022-02-15 | 2022-02-15 | License plate product quality detection method, system and equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115393855A true CN115393855A (en) | 2022-11-25 |
Family
ID=84115329
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210137467.5A Pending CN115393855A (en) | 2022-02-15 | 2022-02-15 | License plate product quality detection method, system and equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115393855A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116704513A (en) * | 2023-08-04 | 2023-09-05 | 深圳思谋信息科技有限公司 | Text quality detection method, device, computer equipment and storage medium |
-
2022
- 2022-02-15 CN CN202210137467.5A patent/CN115393855A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116704513A (en) * | 2023-08-04 | 2023-09-05 | 深圳思谋信息科技有限公司 | Text quality detection method, device, computer equipment and storage medium |
| CN116704513B (en) * | 2023-08-04 | 2023-12-15 | 深圳思谋信息科技有限公司 | Text quality detection method, device, computer equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110826416B (en) | Bathroom ceramic surface defect detection method and device based on deep learning | |
| CN105913093B (en) | A kind of template matching method for Text region processing | |
| CN111259889A (en) | Image text recognition method and device, computer equipment and computer storage medium | |
| CN113240626B (en) | Glass cover plate concave-convex type flaw detection and classification method based on neural network | |
| CN108090511B (en) | Image classification method and device, electronic equipment and readable storage medium | |
| CN109409355B (en) | Novel transformer nameplate identification method and device | |
| CN113160192A (en) | Visual sense-based snow pressing vehicle appearance defect detection method and device under complex background | |
| CN111767780B (en) | AI and vision combined intelligent integrated card positioning method and system | |
| CN113903024A (en) | Handwritten bill numerical value information identification method, system, medium and device | |
| CN111767754A (en) | Identification code identification method and device, electronic equipment and storage medium | |
| CN113033558A (en) | Text detection method and device for natural scene and storage medium | |
| CN113392819B (en) | Batch academic image automatic segmentation and labeling device and method | |
| CN115908774A (en) | Quality detection method and device of deformed material based on machine vision | |
| CN116704516A (en) | Visual inspection method for water-soluble fertilizer package | |
| CN115393855A (en) | License plate product quality detection method, system and equipment | |
| CN110596118A (en) | Print pattern detection method and print pattern detection device | |
| CN111507177B (en) | Identification method and device for metering turnover cabinet | |
| CN112308848A (en) | Method and system for identifying state of baffle plate of scrap steel truck, electronic equipment and medium | |
| CN115619783B (en) | Method and device for detecting product processing defects, storage medium and terminal | |
| CN110298347B (en) | Method for identifying automobile exhaust analyzer screen based on GrayWorld and PCA-CNN | |
| CN116797602A (en) | Surface defect identification method and device for industrial product detection | |
| Bala et al. | Image simulation for automatic license plate recognition | |
| CN115546141A (en) | Small sample Mini LED defect detection method and system based on multi-dimensional measurement | |
| CN116485992A (en) | Composite three-dimensional scanning method and device and three-dimensional scanner | |
| CN115909151A (en) | Method for identifying serial number of motion container under complex working condition |
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 |