CN114170612A - Recovery production method based on defect judgment of LED unit board - Google Patents
Recovery production method based on defect judgment of LED unit board Download PDFInfo
- Publication number
- CN114170612A CN114170612A CN202111535154.7A CN202111535154A CN114170612A CN 114170612 A CN114170612 A CN 114170612A CN 202111535154 A CN202111535154 A CN 202111535154A CN 114170612 A CN114170612 A CN 114170612A
- Authority
- CN
- China
- Prior art keywords
- display
- character
- screen
- font
- defect
- 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
- 230000007547 defect Effects 0.000 title claims abstract description 173
- 238000004519 manufacturing process Methods 0.000 title claims abstract description 23
- 238000011084 recovery Methods 0.000 title abstract description 7
- 238000012549 training Methods 0.000 claims abstract description 16
- 238000007635 classification algorithm Methods 0.000 claims abstract description 12
- 238000012423 maintenance Methods 0.000 claims abstract description 8
- 238000000034 method Methods 0.000 claims description 53
- 230000002159 abnormal effect Effects 0.000 claims description 31
- 230000000737 periodic effect Effects 0.000 claims description 21
- 238000012015 optical character recognition Methods 0.000 claims description 16
- 230000008439 repair process Effects 0.000 claims description 13
- 238000012545 processing Methods 0.000 claims description 12
- 238000004064 recycling Methods 0.000 claims description 11
- 238000004422 calculation algorithm Methods 0.000 claims description 10
- 230000008859 change Effects 0.000 claims description 10
- 238000012216 screening Methods 0.000 claims description 10
- 238000003672 processing method Methods 0.000 claims description 6
- 230000005856 abnormality Effects 0.000 claims description 5
- 238000005516 engineering process Methods 0.000 claims description 5
- 238000007781 pre-processing Methods 0.000 claims description 3
- 230000008569 process Effects 0.000 claims description 3
- 230000011218 segmentation Effects 0.000 claims description 3
- 238000012360 testing method Methods 0.000 claims description 3
- 238000001514 detection method Methods 0.000 description 3
- 239000000463 material Substances 0.000 description 2
- 238000013136 deep learning model Methods 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 239000011810 insulating material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- 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
- G06T2207/30108—Industrial image inspection
- G06T2207/30141—Printed circuit board [PCB]
-
- 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
- G06T2207/30176—Document
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Quality & Reliability (AREA)
- Control Of Indicators Other Than Cathode Ray Tubes (AREA)
- Controls And Circuits For Display Device (AREA)
Abstract
The invention discloses a recovery production method based on defect judgment of an LED unit board, which comprises the following steps: submitting text information to be displayed and a display mode through a product client; judging whether the unit board screen has a display defect; acquiring a display defect character list; solving the character display abnormity problem by adjusting a font display mode and/or resetting displayed text information; recovering and reproducing the unit board screen, obtaining maintenance work order data, carrying out multi-label text classification algorithm model training to obtain a display defect character prediction model, and predicting a total display defect character list through the trained display defect character prediction model; according to the full display defect character list, the screen of the production unit board needing to be recycled is sent to the scene meeting the display requirement for use.
Description
[ technical field ] A method for producing a semiconductor device
The invention relates to the field of LED unit board recovery production, in particular to a recovery production method based on LED unit board defect judgment.
[ background of the invention ]
A PCB, which is a wiring board, is generally referred to as a printed circuit board, in which a conductive pattern formed by forming a printed circuit, a printed component, or a combination of both on an insulating material according to a predetermined design is called a printed circuit.
The existing circuit board evolves various types along with the development of society and science and technology, is widely used in various aspects of life, and has greatly improved insulativity and high temperature resistance along with the progress of technology.
The existing circuit board is often difficult to repair after problems occur, various defects are inevitable in the circuit board production process, some defects are very obvious and influence the use, some defects exist but can still deal with the use of some simple scenes, and if the circuit board with slight defects can be put into the corresponding scene to play continuously, the circuit board has certain economic value. How to change the circuit board content needing to be displayed completely into the content which is displayed in a defect mode and has no semantic meaning or content loss is a problem which is researched by few people. For example, when a miss appears on the display, it is lady's, and in such a scenario, the board can be used further. After the LED light-emitting plate is assembled on the circuit board, the circuit board is usually returned to a factory for maintenance due to defect display, maintenance personnel can fill a maintenance work order according to the fault condition in the maintenance, and the defects of the circuit board and the corresponding light-emitting defects can be determined in the work order, so that the circuit board with the corresponding defects can avoid the requirement of displaying specific characters through data statistics, and the defective circuit board can be effectively recycled; therefore, a recycling production method based on the defect judgment of the LED unit board is provided for solving the problems.
[ summary of the invention ]
In order to overcome the defects of the prior art and solve the problem that the existing circuit board is difficult to repair after a problem occurs, the invention provides a recovery production method based on LED unit board defect judgment.
The technical scheme adopted by the invention for solving the technical problems is as follows: the invention relates to a recovery production method based on LED unit board defect judgment, which is characterized by comprising the following steps:
submitting text information to be displayed and a display mode through a product client;
all the LED units on the screen of the unit board are set to emit light at the same time and are photographed and stored to obtain a first image group, a single character display picture is determined through the text information and the display mode, and a second image group is obtained through combination of photographing and storage of a camera;
performing display defect judgment processing on the first image group and the second image group, and judging whether a display defect exists in the unit board screen;
judging whether the display defect is a circuit fault problem or not, and simultaneously acquiring a display defect character list;
screening out periodic time points with display defects in a preset time period, acquiring character display abnormal types of each periodic time point, and redefining a character display mode to solve the character display abnormal problem;
when the character display abnormity problem cannot be solved by adjusting the font display mode, the character display abnormity problem is solved by resetting the displayed text information;
when the character display abnormity problem cannot be solved by adjusting a font display mode and/or resetting displayed text information, recovering and reproducing the cell board screen, obtaining maintenance work order data, performing multi-label text classification algorithm model training to obtain a display defect character prediction model, and predicting a full display defect character list through the trained display defect character prediction model;
and according to the full display defect character list, recovering the screen of the production unit board as required, and sending the screen to a scene meeting the display requirement for use.
Preferably, the submitting the text information to be displayed and the display mode through the product client includes: after a user installs a product client, the product client calls a corresponding text display scheme through a self-contained character database and controls a cell board screen to start working according to a display mode set by the user by submitting text information and a display mode which need to be displayed on the cell board screen on the product client, wherein the text information comprises various Chinese characters, English and symbolic characters; the display mode includes the setting of font quantity N, font color C, font height H, font width W, font interval D, font change mode and font display speed V to the demonstration, and wherein font height H will not be greater than the height of screen unit board, and font change mode includes that the setting of showing to the left side or showing to the right side from the screen left side to the font from the screen right side is preset and is shown to the left side from the screen right side.
Preferably, the setting of the unit board screen that all the LED units emit light simultaneously and perform photographing and storage to obtain a first image group, the determining of the single character display picture by the text information and the display mode, and the capturing and storage by the camera to obtain a second image group includes:
the method is characterized in that the steps of setting all the LED units of the screen of the unit board to emit light simultaneously, photographing and storing are specifically as follows: setting all the LED units on the screen of the unit board to emit light simultaneously, wherein the light emitting color is not limited, the light emitting intensity is kept above the minimum light emitting requirement, and using a camera to face the screen of the unit board and shoot and store images, and recording the images as a first image group; the method comprises the following steps of determining a single character display picture through text display information and a display mode, shooting and storing the single character display picture by combining a camera, and obtaining a second image group, wherein the specific steps are as follows: acquiring the unit board width L of the unit board screen, wherein the unit board width L comprises the number N of displayed fonts, the font width W, the font interval D, the font change mode and the font display speed V, acquiring a single character display picture from a first display character until the display picture data of all characters are acquired, and marking each corresponding frame image as a second image group;
in operation, the minimum luminous intensity is the minimum luminous intensity at which a camera is directed at the cell panel screen and a clear luminous area is viewed.
Preferably, the processing of judging display defects of the first image group and the second image group, judging whether a display defect exists in a cell panel screen, judging whether the display defect is a problem of a circuit fault, and acquiring a display defect character list includes:
the display defect judging method comprises two display defect judging methods, wherein the first method is to perform similarity contrast processing of traversing each picture on the first image group according to a first photographing scheme, and observe whether the luminous intensity of an LED unit is obviously different from that of other LED units, so as to judge whether the screen of the unit board has a display defect;
during working, the first processing method specifically comprises the steps of selecting a single normally displayed image of the LED display unit from the first image group as an image matching template, preprocessing the image through an OpenCV image processing library, extracting the area of each LED display unit by using an image threshold segmentation algorithm, matching and comparing the area image of each LED display unit with the image matching template through an image similarity algorithm, and judging that a unit panel screen has a display defect when the similarity is lower than a threshold value according to the matched similarity;
secondly, recognizing text information in the second image group according to OCR characters and matching the text information with the display text information according to the second image group obtained according to a second photographing scheme, and judging whether a display defect exists on a screen or not;
the second processing method specifically comprises the steps of extracting picture data of a shot video, namely the second image group, identifying characters of each image in the second image group through tesseract engine characters of an OCR (optical character recognition) engine, and comparing the characters with the displayed text information, wherein when the comparison result is inconsistent, the problem that a unit board screen has defects is shown;
according to the first display defect judgment method, when the screen is judged to have a display defect, the problem is embodied as that a part of LED units cannot be lighted to emit light or have low brightness, namely, the logic control circuit modules of the part of LED units have a circuit fault problem;
and according to a second display defect judgment method, acquiring a concrete character which does not correspond to the OCR text recognition result and the display text information, and obtaining the display defect character list.
Preferably, the periodic time points at which the display defect occurs are screened out within a preset time period, the character display abnormal type of each periodic time point is obtained, and the character display abnormal problem is solved by redefining the character display mode, specifically comprising the following steps:
a1: screening a video time point with a display defect of the unit board screen by recording the display condition of the characters on the unit board screen within a preset time period, recording the display condition of the unit board screen within the preset time period in a camera video recording mode, judging whether the display defect problem exists in the video and recording the corresponding video time point;
a2: screening out existence of periodic or regular time points according to the video time points during defect display in the step A1, searching a next time point by a recorded time point in a first time point in a period T, marking the time point for 2 times if the next time point exists, increasing the period T if the next time point does not exist, searching the next time point again until all the time points are searched, recording all the time points with the marking times larger than 2 times and recording periods T1 and T2 … of the time points according to the operation, wherein the period T is smaller than the preset time period in the step A1, and the increase of the period T is the shortest interval time of two adjacent time points in all the time points;
a3: obtaining the image difference between a picture with a periodic display defect problem and a picture with the same character but without the display defect problem through an OpenCV image difference position calculation algorithm, thereby obtaining the character display abnormal type of each periodic time point, wherein the character display abnormal type comprises that font display has defects, font display has redundant bright points, and the character display has defects and the specific position of the font display has the redundant bright points;
a4: the character display method redefines the character display method to solve the problem that different types of characters in the step A3 are abnormal in display, redefines the font display format to avoid the defects in font display and the positions of redundant bright spots in font display in the step A3, and comprises the steps of adjusting the font display method to enable the characters to be completely displayed on a screen, wherein the font display adjustment method refers to adjusting the font height H, the font width W or/and changing the character size and the initial xy coordinate position until the positions of the defects in font display and the redundant bright spots in font display in the step A3 are avoided after adjustment; and the definite character adjustment time period is the period T stated in step A2, i.e. adjustment is made when the character has a problem.
Preferably, when the abnormal character display problem cannot be solved by adjusting the font display mode, the method for solving the abnormal character display problem by resetting the displayed text information includes:
according to the fact that display abnormality occurs to the total M characters in the display defect character list, starting from the first character with display abnormality, other characters capable of replacing the character are searched in the character replacement library of the product client side for replacement until all the characters are corrected and displayed normally, the method specifically comprises the following steps:
b1: constructing a character replacement library, wherein the character replacement comprises numeric replacement and replacement between Chinese and English, the numeric replacement comprises replacement of Arabic numerals, Roman numerals and Chinese numerals, and the storage content of the character replacement library is the screen display mode of each character, including font color C, font height H, font width W and font spacing D;
b2: identifying the current person and displaying a specific welcome word by the mode of the front camera of the screen of the unit board, comprising arranging the front camera of the screen of the unit board to monitor and detect the front of the screen in real time, detecting the specific situation of the person in front of the screen by a face recognition technology, and displaying the welcome word 'welcome mr'!when a man passes by the screen! ", said particular welcome word including a welcome word for a type of common person;
b3: the abnormal alpha characters are displayed by replacing the character replacement library, and a welcome word' welcome lady!is displayed on a screen! When the display defect of the Lady is detected, the Lady is replaced by the Lady to be displayed according to the character replacement library in the step B1 so as to avoid the problem of abnormal character display.
Preferably, the obtaining of the repair order data is performed with multi-label text classification algorithm model training to obtain a display defect character prediction model, and the display defect character prediction model after the training predicts a total display defect character list, which includes:
c1: the data set is made by collecting historical repair order data for a large number of screens, and specifically comprises: collecting historical repair work order data of a large number of screens of the same type and the same circuit structure, wherein the characteristic value of the data set comprises the following data in the historical repair work order: the method comprises the following steps of (1) using time of a screen, working environment of the screen, electrical characteristic data of a plurality of ports of a circuit board of the screen, a mini display unit with a damaged screen and a position of the mini display unit, wherein a predicted value of a data set is as follows: after manual testing, acquiring a total display defect character list of which the screen has display defects;
c2: and C1, performing model training by combining the data set with a multi-label text classification algorithm to obtain a display defect character prediction model, wherein the process of performing model training by combining the multi-label text classification algorithm is to use the characteristic value of the data set as an input parameter and the predicted value of the data set as an output parameter to perform model training to obtain the display defect character prediction model.
C3: and inputting the characteristic value of the screen of the unit board needing to be recycled and produced according to the display defect character prediction model to obtain a full-display defect character list of the screen of the unit board.
Preferably, the scene meeting the display requirement includes: and in the characters required to be displayed in the use scene, the scene of any character in the full display defect character list is not included.
The invention has the advantages that: the invention combines the fault detection and the text display adjustment method, improves the detection rate of display fault LED products, and the result deep learning model is used for producing recyclable LED unit boards again, thereby prolonging the service life of the products.
[ description of the drawings ]
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without inventive exercise.
Fig. 1 is a schematic flow chart of a recycling production method based on defect judgment of an LED unit board according to an embodiment of the present invention;
[ detailed description ] embodiments
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, a recycling method based on defect judgment of LED unit board includes the following steps:
s1: a user submits text information to be displayed and a display mode through a product client;
the S1 is specifically that after the user installs the product client, the product client submits the text information and the display mode to be displayed on the cell board screen, and the product client calls the corresponding text display scheme through the self-contained character database and controls the cell board screen to start working according to the display mode set by the user, where the text information includes various characters, english and symbols; the display mode includes the setting of font quantity N, font color C, font height H, font width W, font interval D, font change mode and font display speed V to the demonstration, and wherein font height H will not be greater than the height of screen unit board, and font change mode includes that the setting of showing to the left side or showing to the right side from the screen left side to the font from the screen right side is preset and is shown to the left side from the screen right side.
S2: the method comprises the following steps of acquiring a first image group and a second image group through two photographing schemes, wherein the photographing schemes comprise: a first photographing scheme: setting all LED units of the screen of the unit board to emit light simultaneously and photographing and storing to obtain a first image group, and adopting a second photographing scheme: determining a single character display picture through text information and a display mode, and shooting and storing the single character display picture by combining a camera to obtain the second image group;
the method for setting all the LED units on the screen to emit light simultaneously and storing the pictures in the S2 specifically comprises the following steps: setting all the LED units on the screen of the unit board to emit light simultaneously, wherein the light emission color is not limited, the light emission intensity is kept above the minimum light emission requirement, and then using a camera to face the screen of the unit board and taking a picture to store the picture, and recording the picture as a first picture group; in S2, a single character display screen is determined by text display information and a display mode, and the second image group is obtained by shooting and storing with a camera, specifically: acquiring the cell board width L of the cell board screen, according to the step S1, acquiring a single character display picture from the first display character until the display picture data of all characters are acquired, taking the first display character as an example, when the character is completely displayed on the cell board screen from the right side of the cell board screen, i.e. the length of W counted from the right side of the cell board screen is occupied, starting to shoot a video by using a camera, recording the video until the first character just finishes the display of the complete character from the left side of the cell board screen, and then shooting the video by using the camera again when the second character is completely displayed on the cell board screen from the right side of the cell board screen, acquiring all characters according to the same processing, and acquiring N videos in total, recording the images as a first video group, and recording each corresponding frame image as a second image group;
in operation, the minimum luminous intensity is the minimum luminous intensity at which the camera is directed at the cell panel screen and a clear luminous area is viewed.
S3: performing display defect judgment processing on the first image group and the second image group obtained according to step S2 by two display defect judgment methods to judge whether or not a display defect exists in the cell panel screen;
the display defect determination method in S3 includes two specific methods, the first method is to perform similarity contrast processing of traversing each picture on the first image group according to the first photographing scheme in step S2, and observe whether there is an LED unit whose light emitting intensity is obviously different from that of other LED units, thereby determining whether there is a display defect on the screen of the cell panel.
During working, the first processing method specifically comprises the steps of selecting a single normally displayed image of an LED display unit in a first image group as an image matching template, preprocessing the image through an OpenCV image processing library, extracting the area of each LED display unit by using an image threshold segmentation algorithm, matching and comparing the area image of each LED display unit with the image matching template through an image similarity algorithm, and judging that a unit panel screen has a display defect when the similarity is lower than a threshold value according to the matched similarity;
secondly, according to the second image group obtained by the second photographing scheme in step S2, recognizing by OCR characters that the text information in the second image matches the display text information, and determining whether there is a display defect on the screen.
The second processing method specifically comprises the steps of extracting picture data of a shot video, namely the second image group, recognizing characters of each image in the second image group through tesseract engine characters of an OCR (optical character recognition), comparing the characters with the displayed text information, and judging whether a display defect exists, for example, in a first video shot by first character display, analyzing a video picture image, obtaining a recognition result through OCR text recognition, comparing the recognition result with the first character of the displayed text information, and when the text comparison is consistent, indicating that the display of a unit panel screen on the character is normal; when the comparison results are inconsistent, the problem that the display of the character on the unit board screen has a defect is shown until all the characters are judged;
the severity of the display defect can be judged by the OCR character recognition method, when some display units on the unit board screen have serious problems, the OCR character recognition can be failed or recognized into other characters, when some display units on the unit board screen have no serious problems, the characters are correctly recognized by the OCR character recognition method, and therefore the display problem degree of each character on the unit board screen can be effectively verified by the method.
S4: based on the two display defect judgment methods in step S3, further judging whether the display defect is a circuit failure problem, and simultaneously obtaining a display defect character list;
s4 is the first method for determining display defects according to step S3, when the screen is determined to have display defects, the problem is that some LED units cannot be lit or have low brightness, that is, the logic control circuit modules of the LED units have a circuit failure problem;
according to the second display defect judgment method in step S3, acquiring a specific character of which the OCR text recognition result does not correspond to the display text information, and obtaining the display defect character list;
in the above steps, it is determined whether a circuit fault problem exists in a part of logic control circuit modules of the LED units, and then the display defect character list is obtained, so that it is possible to avoid the situation that ocr character recognition itself is wrong, or the display text information is set wrong, and the like, which is erroneously determined as a character display defect.
S5: screening out periodic time points with display defects in a preset time period, acquiring character display abnormal types of each periodic time point, and redefining a character display mode to solve the character display abnormal problem;
the step S5 specifically includes the following steps:
a1: screening a video time point with a display defect of the unit board screen by recording the display condition of the characters on the unit board screen within a preset time period, recording the display condition of the unit board screen within the preset time period in a camera video recording mode, judging whether the display defect problem exists in the video and recording the corresponding video time point;
a2: screening out existence of periodic or regular time points according to the video time points during character defect display in the step A1, searching a next time point by a recorded time point in a first time point in a period T, marking the time point for 2 times if the next time point exists, increasing the period T if the next time point does not exist, searching the next time point again until all the time points are searched, recording all the time points with the marking times larger than 2 times and recording periods T1 and T2 … of the time points according to the operation, wherein the period T is smaller than the preset time period in the step A1, and the amplitude of the period T is the shortest interval time of two adjacent time points in all the time points;
a3: the image difference between the picture with the periodic display defect problem and the picture with the same character but without the display defect problem is obtained through an OpenCV image difference position calculation algorithm, so that the character display abnormal type of each periodic time point is obtained, wherein the character display abnormal type comprises that font display has defects, font display has redundant bright spots, and the specific positions of the font display with defects and the font display with the redundant bright spots are also included, for example: and obtaining the image difference between a picture a with the problem of the display defect of the periodic character and a picture b with the same character but without the problem of the display defect by using an OpenCV image difference position calculation algorithm, wherein the obtained image difference point is that the display of a certain x and y coordinate of the font has a defect, and the obtained image difference point is used for judging that the display abnormal type of the character is that the display of the font has the defect, and the position where the display of the font has the defect is the (x, y) position.
A4: redefining a font display format for the condition that periodic or regular screen display defects exist to avoid the font display defects and the positions of redundant bright spots, wherein the font display method comprises the step of adjusting the font display mode to enable the characters to be completely displayed on a screen, and the adjustment of the font display mode refers to the adjustment of the font height H, the font width W or/and the change of the character size and the initial xy coordinate position in the step S1 until the adjustment of the font display defects and the positions of the redundant bright spots in the font display in the step A3 are avoided; and the definite character adjustment time period is the period T stated in step A2, i.e. adjustment is made when the character has a problem.
S6: according to the method of step S5, when the abnormal character display problem cannot be solved by adjusting the font display mode, the abnormal character display problem is solved by resetting the displayed text information, which includes that, according to step S4, M characters in the display defect character list are obtained to have abnormal display, and starting from the first character with abnormal display, other characters which can substitute the character are searched in the character substitution library of the product client, until all the characters are corrected and displayed normally, and specifically includes:
b1: constructing a character replacement library, wherein character replacement comprises replacement of numbers, replacement between Chinese and English words, and replacement between synonyms, such as 10, ten and pick; "Lady" and "Lady"; the character replacement library comprises a character color C, a character height H, a character width W and a character spacing D, wherein the character color C is a font color, the character height H is a font color, and the character width W is a font spacing D;
b2: identifying the current person and displaying a specific welcome word by means of a front camera of a cell panel screen, for example, arranging the front camera of the cell panel screen to monitor the front of a detection screen in real time, detecting the specific situation of the person in front of the screen by a face recognition technology, and displaying the welcome word 'welcome mr'!on the screen when a man passes by! ", the particular welcome word including a welcome word for a type of person in common, such as" welcome Mr.! "," welcome ladies! "and the like;
b3: based on the steps S2 and/or S3, the abnormal alpha characters are displayed instead by the character replacement library, for example, when the screen displays the welcome word "welcome ladies! "yes", it is detected through steps S2, S3 that "Lady" has display defect, then "Lady" is replaced by "Lady" to display according to the character replacement library described in step B1, so as to avoid the character display abnormal problem.
S7: according to the method of the steps S5-S6, when the problem of character display abnormality cannot be solved by adjusting a font display mode and/or resetting displayed text information, the cell board screen is recycled and reproduced, repair work order data is obtained to conduct multi-label text classification algorithm model training, a display defect character prediction model is obtained, and a possible total display defect character list is predicted through the trained display defect character prediction model.
The method for predicting the total display defect character list which may exist through the trained display defect character prediction model comprises the following steps:
c1: the data set is made by collecting historical repair order data for a large number of screens, and specifically comprises: collecting historical repair work order data of a large number of screens of the same type and the same circuit structure, wherein the characteristic value of the data set comprises the following data in the historical repair work order: the method comprises the following steps of (1) using time of a screen, working environment of the screen, electrical characteristic data of a plurality of ports of a circuit board of the screen, a mini display unit with a damaged screen and a position of the mini display unit, wherein a predicted value of a data set is as follows: after manual testing, acquiring a total display defect character list of which the screen has display defects;
c2: and C1, performing model training by combining the data set with a multi-label text classification algorithm to obtain a display defect character prediction model, wherein the process of performing model training by combining the multi-label text classification algorithm is to use the characteristic value of the data set as an input parameter and the predicted value of the data set as an output parameter to perform model training to obtain the display defect character prediction model.
C3: inputting a characteristic value of a screen of a unit board to be recycled according to the display defect character prediction model to obtain a full-display defect character list of the screen of the unit board;
s8: and according to the full-scale display defect character list in the step S7, sending the screen of the production cell board needing to be recovered to a scene meeting the display requirement for use.
The scene meeting the display requirement comprises the following steps: and in the characters required to be displayed in the use scene, the scene of any character in the full display defect character list is not included.
In the description herein, references to the description of "one embodiment," "an example," "a specific example" or the like are intended to mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
The foregoing shows and describes the general principles, essential features, and advantages of the invention. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, which are described in the specification and illustrated only to illustrate the principle of the present invention, but that various changes and modifications may be made therein without departing from the spirit and scope of the present invention, which fall within the scope of the invention as claimed.
Claims (8)
1. A recycling production method based on LED unit board defect judgment is characterized by comprising the following steps:
submitting text information to be displayed and a display mode through a product client;
all the LED units on the screen of the unit board are set to emit light at the same time and are photographed and stored to obtain a first image group, a single character display picture is determined through the text information and the display mode, and a second image group is obtained through combination of photographing and storage of a camera;
performing display defect judgment processing on the first image group and the second image group, and judging whether a display defect exists in the unit board screen;
judging whether the display defect is a circuit fault problem or not, and simultaneously acquiring a display defect character list;
screening out periodic time points with display defects in a preset time period, acquiring character display abnormal types of each periodic time point, and redefining a character display mode to solve the character display abnormal problem;
when the character display abnormity problem cannot be solved by adjusting the font display mode, the character display abnormity problem is solved by resetting the displayed text information;
when the character display abnormity problem cannot be solved by adjusting a font display mode and/or resetting displayed text information, recovering and reproducing the cell board screen, obtaining maintenance work order data, performing multi-label text classification algorithm model training to obtain a display defect character prediction model, and predicting a full display defect character list through the trained display defect character prediction model;
and according to the full display defect character list, recovering the screen of the production unit board as required, and sending the screen to a scene meeting the display requirement for use.
2. The recycling production method based on the defect judgment of the LED unit board as claimed in claim 1, wherein: the submitting of the text information to be displayed and the display mode through the product client comprises the following steps: after a user installs a product client, the product client calls a corresponding text display scheme through a self-contained character database and controls a cell board screen to start working according to a display mode set by the user by submitting text information and a display mode which need to be displayed on the cell board screen on the product client, wherein the text information comprises various Chinese characters, English and symbolic characters; the display mode includes the setting of font quantity N, font color C, font height H, font width W, font interval D, font change mode and font display speed V to the demonstration, and wherein font height H will not be greater than the height of screen unit board, and font change mode includes that the setting of showing to the left side or showing to the right side from the screen left side to the font from the screen right side is preset and is shown to the left side from the screen right side.
3. The recycling production method based on the defect judgment of the LED unit board as claimed in claim 2, wherein: all the LED units emit light simultaneously through the screen of the unit board and are photographed and stored to obtain a first image group, a single character display picture is determined through the text information and the display mode, and a second image group is obtained through combination of photographing and storage of a camera, and the method comprises the following steps:
the method is characterized in that the steps of setting all the LED units of the screen of the unit board to emit light simultaneously, photographing and storing are specifically as follows: setting all the LED units on the screen of the unit board to emit light simultaneously, wherein the light emitting color is not limited, the light emitting intensity is kept above the minimum light emitting requirement, and using a camera to face the screen of the unit board and shoot and store images, and recording the images as a first image group; the method comprises the following steps of determining a single character display picture through text display information and a display mode, shooting and storing the single character display picture by combining a camera, and obtaining a second image group, wherein the specific steps are as follows: acquiring the unit board width L of the unit board screen, wherein the unit board width L comprises the number N of displayed fonts, the font width W, the font interval D, the font change mode and the font display speed V, acquiring a single character display picture from a first display character until the display picture data of all characters are acquired, and marking each corresponding frame image as a second image group;
in operation, the minimum luminous intensity is the minimum luminous intensity at which a camera is directed at the cell panel screen and a clear luminous area is viewed.
4. The recycling production method based on the defect judgment of the LED unit board as claimed in claim 3, wherein: the processing of judging the display defects of the first image group and the second image group, judging whether the display defects exist on the screen of the unit board, judging whether the display defects are the problem of circuit faults, and simultaneously acquiring a display defect character list comprises the following steps:
the display defect judging method comprises two display defect judging methods, wherein the first method is to perform similarity contrast processing of traversing each picture on the first image group according to a first photographing scheme, and observe whether the luminous intensity of an LED unit is obviously different from that of other LED units, so as to judge whether the screen of the unit board has a display defect;
during working, the first processing method specifically comprises the steps of selecting a single normally displayed image of the LED display unit from the first image group as an image matching template, preprocessing the image through an OpenCV image processing library, extracting the area of each LED display unit by using an image threshold segmentation algorithm, matching and comparing the area image of each LED display unit with the image matching template through an image similarity algorithm, and judging that a unit panel screen has a display defect when the similarity is lower than a threshold value according to the matched similarity;
secondly, recognizing text information in the second image group according to OCR characters and matching the text information with the display text information according to the second image group obtained according to a second photographing scheme, and judging whether a display defect exists on a screen or not;
the second processing method specifically comprises the steps of extracting picture data of a shot video, namely the second image group, identifying characters of each image in the second image group through tesseract engine characters of an OCR (optical character recognition) engine, and comparing the characters with the displayed text information, wherein when the comparison result is inconsistent, the problem that a unit board screen has defects is shown;
according to the first display defect judgment method, when the screen is judged to have a display defect, the problem is embodied as that a part of LED units cannot be lighted to emit light or have low brightness, namely, the logic control circuit modules of the part of LED units have a circuit fault problem;
and according to a second display defect judgment method, acquiring a concrete character which does not correspond to the OCR text recognition result and the display text information, and obtaining the display defect character list.
5. The recycling production method based on the defect judgment of the LED unit board as claimed in claim 4, wherein: the method comprises the following steps of screening out periodic time points with display defects in a preset time period, acquiring character display abnormal types of the periodic time points, redefining a character display mode to solve the character display abnormal problem, and specifically comprises the following steps:
a1: screening a video time point with a display defect of the unit board screen by recording the display condition of the characters on the unit board screen within a preset time period, recording the display condition of the unit board screen within the preset time period in a camera video recording mode, judging whether the display defect problem exists in the video and recording the corresponding video time point;
a2: screening out existence of periodic or regular time points according to the video time points during defect display in the step A1, searching a next time point by a recorded time point in a first time point in a period T, marking the time point for 2 times if the next time point exists, increasing the period T if the next time point does not exist, searching the next time point again until all the time points are searched, recording all the time points with the marking times larger than 2 times and recording periods T1 and T2 … of the time points according to the operation, wherein the period T is smaller than the preset time period in the step A1, and the increase of the period T is the shortest interval time of two adjacent time points in all the time points;
a3: obtaining the image difference between a picture with a periodic display defect problem and a picture with the same character but without the display defect problem through an OpenCV image difference position calculation algorithm, thereby obtaining the character display abnormal type of each periodic time point, wherein the character display abnormal type comprises that font display has defects, font display has redundant bright points, and the character display has defects and the specific position of the font display has the redundant bright points;
a4: the character display method redefines the character display method to solve the problem that different types of characters in the step A3 are abnormal in display, redefines the font display format to avoid the defects in font display and the positions of redundant bright spots in font display in the step A3, and comprises the steps of adjusting the font display method to enable the characters to be completely displayed on a screen, wherein the font display adjustment method refers to adjusting the font height H, the font width W or/and changing the character size and the initial xy coordinate position until the positions of the defects in font display and the redundant bright spots in font display in the step A3 are avoided after adjustment; and the definite character adjustment time period is the period T stated in step A2, i.e. adjustment is made when the character has a problem.
6. The recycling production method based on the defect judgment of the LED unit board as claimed in claim 5, wherein: when the character display abnormity problem cannot be solved by adjusting the font display mode, the character display abnormity problem is solved by resetting the displayed text information, and the method comprises the following steps:
according to the fact that display abnormality occurs to the total M characters in the display defect character list, starting from the first character with display abnormality, other characters capable of replacing the character are searched in the character replacement library of the product client side for replacement until all the characters are corrected and displayed normally, the method specifically comprises the following steps:
b1: constructing a character replacement library, wherein the character replacement comprises numeric replacement and replacement between Chinese and English, the numeric replacement comprises replacement of Arabic numerals, Roman numerals and Chinese numerals, and the storage content of the character replacement library is the screen display mode of each character, including font color C, font height H, font width W and font spacing D;
b2: identifying the current person and displaying a specific welcome word by the mode of the front camera of the screen of the unit board, comprising arranging the front camera of the screen of the unit board to monitor and detect the front of the screen in real time, detecting the specific situation of the person in front of the screen by a face recognition technology, and displaying the welcome word 'welcome mr'!when a man passes by the screen! ", said particular welcome word including a welcome word for a type of common person;
b3: the abnormal alpha characters are displayed by replacing the character replacement library, and a welcome word' welcome lady!is displayed on a screen! When the display defect of the Lady is detected, the Lady is replaced by the Lady to be displayed according to the character replacement library in the step B1 so as to avoid the problem of abnormal character display.
7. The recycling production method based on the defect judgment of the LED unit board as claimed in claim 6, wherein: the method for obtaining the maintenance work order data to train the multi-label text classification algorithm model to obtain a display defect character prediction model, and predicting a total display defect character list by the display defect character prediction model after the training comprises the following steps:
c1: the data set is made by collecting historical repair order data for a large number of screens, and specifically comprises: collecting historical repair work order data of a large number of screens of the same type and the same circuit structure, wherein the characteristic value of the data set comprises the following data in the historical repair work order: the method comprises the following steps of (1) using time of a screen, working environment of the screen, electrical characteristic data of a plurality of ports of a circuit board of the screen, a mini display unit with a damaged screen and a position of the mini display unit, wherein a predicted value of a data set is as follows: after manual testing, acquiring a total display defect character list of which the screen has display defects;
c2: and C1, performing model training by combining the data set with a multi-label text classification algorithm to obtain a display defect character prediction model, wherein the process of performing model training by combining the multi-label text classification algorithm is to use the characteristic value of the data set as an input parameter and the predicted value of the data set as an output parameter to perform model training to obtain the display defect character prediction model.
C3: and inputting the characteristic value of the screen of the unit board needing to be recycled and produced according to the display defect character prediction model to obtain a full-display defect character list of the screen of the unit board.
8. The recycling production method based on the defect judgment of the LED unit board as claimed in claim 7, wherein: the scene meeting the display requirement comprises the following steps: and in the characters required to be displayed in the use scene, the scene of any character in the full display defect character list is not included.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111535154.7A CN114170612A (en) | 2021-12-15 | 2021-12-15 | Recovery production method based on defect judgment of LED unit board |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111535154.7A CN114170612A (en) | 2021-12-15 | 2021-12-15 | Recovery production method based on defect judgment of LED unit board |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114170612A true CN114170612A (en) | 2022-03-11 |
Family
ID=80486730
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111535154.7A Pending CN114170612A (en) | 2021-12-15 | 2021-12-15 | Recovery production method based on defect judgment of LED unit board |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114170612A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114419624A (en) * | 2022-03-28 | 2022-04-29 | 天津市北海通信技术有限公司 | Image character checking method and system based on image visual algorithm |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2003100405A1 (en) * | 2002-05-23 | 2003-12-04 | Hitachi High-Technologies Corporation | Defect classification device generation method and automatic defect classification method |
| CN104634792A (en) * | 2013-11-07 | 2015-05-20 | 沈阳芯源微电子设备有限公司 | PSS yield online monitoring system and detection method thereof |
| CN112718552A (en) * | 2020-12-11 | 2021-04-30 | 江苏烨明光电有限公司 | Automatic quality defect inspection device for LED circuit board and working method thereof |
| CN112881430A (en) * | 2021-03-10 | 2021-06-01 | 深圳汇义科技有限公司 | Display screen defect monitoring system based on image recognition and detection method thereof |
| WO2021179679A1 (en) * | 2020-03-13 | 2021-09-16 | 苏州鑫睿益荣信息技术有限公司 | Automobile dashboard testing system and testing method based on machine vision |
-
2021
- 2021-12-15 CN CN202111535154.7A patent/CN114170612A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2003100405A1 (en) * | 2002-05-23 | 2003-12-04 | Hitachi High-Technologies Corporation | Defect classification device generation method and automatic defect classification method |
| CN104634792A (en) * | 2013-11-07 | 2015-05-20 | 沈阳芯源微电子设备有限公司 | PSS yield online monitoring system and detection method thereof |
| WO2021179679A1 (en) * | 2020-03-13 | 2021-09-16 | 苏州鑫睿益荣信息技术有限公司 | Automobile dashboard testing system and testing method based on machine vision |
| CN112718552A (en) * | 2020-12-11 | 2021-04-30 | 江苏烨明光电有限公司 | Automatic quality defect inspection device for LED circuit board and working method thereof |
| CN112881430A (en) * | 2021-03-10 | 2021-06-01 | 深圳汇义科技有限公司 | Display screen defect monitoring system based on image recognition and detection method thereof |
Non-Patent Citations (2)
| Title |
|---|
| WEI-CHENG CHANG AUTHOR PICTUREWEI-CHENG CHANG CARNEGIE MELLON UNIVERSITY, PITTSBURGH, PA, USAVIEW PROFILE: "Deep learning model for imbalanced multi-label surface defect classification", TAMING PRETRAINED TRANSFORMERS FOR EXTREME MULTI-LABEL TEXT CLASSIFICATION, 31 August 2020 (2020-08-31) * |
| 陶显: "基于深度学习的表面缺陷检测方法综述", 《自动化学报》, 2 April 2020 (2020-04-02), pages 1017 - 1034 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114419624A (en) * | 2022-03-28 | 2022-04-29 | 天津市北海通信技术有限公司 | Image character checking method and system based on image visual algorithm |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105808416B (en) | A kind of automated testing method and system of man-machine figure interactive interface | |
| US6226407B1 (en) | Method and apparatus for analyzing computer screens | |
| US20120070086A1 (en) | Information reading apparatus and storage medium | |
| CN110909897A (en) | Method, system, storage medium and computer equipment for repairing equipment failure | |
| CN114170612A (en) | Recovery production method based on defect judgment of LED unit board | |
| KR102602689B1 (en) | Methods and system for vision-based automatic fault notification and classification of system lighting | |
| CN113111903A (en) | Intelligent production line monitoring system and monitoring method | |
| CN115187597B (en) | Textile defect detection method | |
| CN114255232B (en) | LED display screen display defect detection method, electronic equipment and device | |
| CN111768398A (en) | Defect detection method and device for photovoltaic module, storage medium and computing equipment | |
| CN111967490A (en) | Model training method for map detection and map detection method | |
| JP2014137284A (en) | Teacher data creation support device, teacher data creation device, image classification device, teacher data creation support method, teacher data creation method and image classification method | |
| US20170193330A1 (en) | Method and apparatus for detecting and classifying active matrix organic light emitting diode panel | |
| JP4993678B2 (en) | Interactive moving image monitoring method, interactive moving image monitoring apparatus, and interactive moving image monitoring program | |
| CN112991438B (en) | Coal gangue detection and identification system, intelligent coal discharging system and training method of model | |
| US20110122247A1 (en) | Surveillance system | |
| US20230386188A1 (en) | Image analysis system an update method for machine learning model | |
| CN116778837A (en) | Multifunctional display fault detection platform | |
| CN116935766A (en) | LED screen anomaly detection and data analysis method | |
| KR20230126163A (en) | Training data generation apparatus, training data generation method, and program recorded on recording medium | |
| CN112330583A (en) | Product defect detection method, device, equipment and storage medium | |
| JPH07175894A (en) | Neural network, character recognition method, electronic parts mounting inspecting device and managing method using the same | |
| CN114782426A (en) | Knitted fabric broken yarn defect detection method based on artificial intelligence system | |
| CN110675397B (en) | Transformer substation protection pressing plate state checking method | |
| CN112380970A (en) | Video target detection method based on local area search |
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 |