CN117395983A

CN117395983A - Automatic detection method and device for PCB (printed circuit board) patches

Info

Publication number: CN117395983A
Application number: CN202311526110.7A
Authority: CN
Inventors: 李泽锋; 朱磊
Original assignee: Shenzhen Chuangxin Zhihui Electronic Technology Co ltd
Current assignee: Shenzhen Chuangxin Zhihui Electronic Technology Co ltd
Priority date: 2023-11-16
Filing date: 2023-11-16
Publication date: 2024-01-12

Abstract

The invention is applicable to the technical field of PCB patch detection control, and provides an automatic PCB detection method which is applied to an automatic PCB detection control system and comprises an image acquisition module, a characteristic information module, a connection attribute module, a patch information module, a connection relation module, a wireless communication module, a memory, an alarm, a processing center and a smart phone; the image acquisition module, the characteristic information module, the connection attribute module, the patch information module, the connection relation module, the wireless communication module, the memory and the alarm are all connected with the processing center; the wireless communication module is provided with an internet of things unit, is responsible for receiving and transmitting wireless signals, and is connected with the intelligent mobile phone in an effective internet of things or internet range. The invention can carry out comprehensive and deep automatic detection on the PCB patch, solves the problem that part of detection items of the PCB patch are omitted or can not be identified, and improves the detection accuracy.

Description

Automatic detection method and device for PCB (printed circuit board) patches

Technical Field

The invention belongs to the technical field of detection and control of a PCB (printed circuit board) patch, and particularly relates to an automatic detection method and device for the PCB patch.

Background

The existing automatic detection of the PCB patch occasionally generates detection errors, which are caused by the fact that the PCB patch technology is not deeply and comprehensively detected, so that part of detection items are omitted or cannot be identified.

Disclosure of Invention

In order to solve the defects in the prior art, the invention aims to provide the automatic detection method and the device for the PCB patch, which can carry out comprehensive and deep automatic detection on the PCB patch by arranging the image acquisition module, the characteristic information module, the connection attribute module, the patch information module and the connection relation module, and solve the problem that part of detection items of the PCB patch are omitted or cannot be identified, thereby improving the detection accuracy, reducing the detection cost, reducing the repeated reworking and reducing the production cost.

In order to achieve the above purpose, the technical scheme adopted by the invention is as follows:

the automatic detection method of the PCB board is applied to an automatic detection control system of the PCB board and comprises an image acquisition module, a characteristic information module, a connection attribute module, a patch information module, a connection relation module, a wireless communication module, a memory, an alarm, a processing center and a smart phone; the image acquisition module, the characteristic information module, the connection attribute module, the patch information module, the connection relation module, the wireless communication module, the memory and the alarm are all connected with the processing center; the wireless communication module is provided with an internet of things unit, is responsible for receiving and transmitting wireless signals, and is connected with the intelligent mobile phone in an effective internet of things or internet range; the intelligent mobile phone is automatically connected with the wireless communication module in a networking way within the range of the Internet of things or the Internet;

The processing center is responsible for information transmission of an image acquisition module, a characteristic information module, a connection attribute module, a patch information module, a connection relation module, a wireless communication module, a memory and an alarm, is a hub center of the system, and compares the acquired first image definition with the image definition condition stored in the memory: if the information is consistent, the information is transmitted to the element characteristic unit, if the information is inconsistent, the information is transmitted to the alarm and the information reminding unit is informed to carry out rechecking; comparing the patch characteristic identification information with patch detection standards stored in a memory: if the detection result is consistent, a patch detection report is generated, if the detection result is inconsistent, the patch detection report is transmitted to an alarm and a reworking and rechecking is notified until the patch detection result is qualified;

the alarm automatically alarms and notifies the information reminding unit to recheck when the acquired first image definition is inconsistent with the image definition condition stored in the memory; when the patch characteristic identification information is inconsistent with the patch detection standard stored in the memory, automatically alarming and notifying reworking and reinspection until the patch characteristic identification information is qualified;

the image acquisition module obtains first high-definition image information of the positions of the soldered spots on the PCB through the high-definition camera and transmits the first high-definition image information to the characteristic information module;

The characteristic information module comprises a clear obtaining unit, an information reminding unit, a clear enhancing unit, an element characteristic unit, an identification characteristic unit, a characteristic information unit, a relation characteristic unit and a characteristic collection unit, and is responsible for identifying characteristics of the obtained first high-definition image information to obtain a characteristic information collection and transmitting the characteristic information collection to the connection attribute module; the clear enhancement unit comprises an inverse mapping chart subunit, a full resolution subunit, a crack capturing subunit, a multiple adjustment subunit, a function integration subunit and a mapping function subunit; the clear obtaining unit comprises a main body marking subunit, a main body characteristic subunit and an information saturation subunit;

the connection attribute module comprises a pixel selection unit, a world coordinate unit and a plane data unit, and obtains connection attribute information of each component in the PCB patch according to the characteristic information set and transmits the connection attribute information to the patch information module;

the patch information module is connected with attribute information according to the connection attribute information of each component in the patch to obtain patch detection standard information and transmits the patch detection standard information to the connection relation module;

the connection relation module comprises an equipotential discrete unit, a search area unit, a characteristic calculation unit, a target selection unit and a characteristic identification unit, and obtains patch characteristic identification information according to element information characteristics, element connecting piece information characteristics and element connection relation information characteristics and transmits the patch characteristic identification information to the processing center; the feature calculation unit comprises a transition probability subunit, a path parameter subunit and a feature coordinate subunit.

The invention provides an automatic detection method of a PCB, which comprises the following steps:

s10, the image acquisition module obtains first high-definition image information of a welding spot position of the PCB, which is pasted on the PCB after a reflow soldering process, through a set high-definition camera, and transmits the first high-definition image information to the characteristic information module;

s20, the characteristic information module identifies the characteristics of the obtained first high-definition image information to obtain a processed image of the characteristic information set, and transmits the processed image to the connection attribute module;

s30, the connection attribute module obtains plane point cloud data of connection attribute information of each component in the PCB patch according to the processed image and transmits the plane point cloud data to the patch information module;

s40, the patch information module obtains processed cloud data of patch detection standard information according to the planar point cloud data of the connection attribute information of each component in the patch, and transmits the processed cloud data to the connection relation module;

s50, the connection relation module obtains a plurality of patch characteristic identification information according to the processed cloud data of the patch detection standard information, marks the patch characteristic identification information to obtain a detection result, and transmits the detection result to the processing center;

s60, comparing the patch characteristic identification information with patch detection standards stored in a memory by the processing center: if the detection result is consistent, a patch detection report is generated, if the detection result is inconsistent, the patch detection report is transmitted to an alarm and a reworking and rechecking is notified until the patch detection result is qualified.

Further, the step S20 "the identification feature of the obtained first high definition image information to obtain the processed image of the feature information set", includes the steps of:

s21, the clear enhancement unit obtains a processed image with full resolution by adopting an automatic data enhancement method according to the resolution of the first high-definition image, and transmits the processed image to the element characteristic unit; the method also comprises the following steps:

s211, the inverse mapping image unit encodes the first image into a feature image, extracts various feature information, connects the feature information in series to obtain an image low-resolution inverse mapping image, and transmits the image low-resolution inverse mapping image to the full-resolution unit;

s212, the full resolution unit samples each pixel point in the low resolution inverse mapping chart in detail and multiplies the pixel points to obtain a full resolution mapping chart, and the full resolution mapping chart is transmitted to the crack capturing unit;

s213, the crack capturing unit multiplies each pixel point of the full resolution mapping diagram and the enhanced image thereof to obtain a full resolution depth image, and transmits the full resolution depth image to the multiple adjusting unit;

s214, the multiple adjusting unit performs the magnification or reduction of the designated multiple or maintains the original state of the full-resolution depth image according to the machine designated probability and the scaling ratio, and transmits the magnification or reduction to the area parameter unit;

s215, the function integration unit calculates the mapping function integration according to an area ratio parameter calculation formula' P=V/S, wherein P is an area ratio parameter, V is a mapping function integration, S is an image area after scaling or maintaining, and the mapping function integration is transmitted to the mapping function unit;

S216, the mapping function unit calculates a formula' according to the integral of the mapping function "V is the integral of the mapping function, H is the height of the image, W is the width of the image, alpha (x, y) is the Gaussian mapping function of the image, and d is the area' of each tiny bin on the curved surface in the image to calculate the size of the image enhancement area;

s22, the element characteristic unit obtains an element characteristic database according to the processed image with full resolution and transmits the element characteristic database to the identification characteristic unit;

s23, the identification feature unit identifies the first image information according to the element feature database to obtain element identification features, marks the element identification features and transmits the element identification features to the feature information unit;

s24, the characteristic information unit obtains element connection characteristic information according to the element identification characteristics and the element characteristic database and transmits the element connection characteristic information to the relation characteristic unit;

s25, the relation feature unit recognizes the element recognition features marked in the first image information according to the element connection feature information to obtain recognition features of the connection piece and the connection relation, and transmits the recognition features to the feature set unit;

s26, the feature set unit obtains an identification feature set according to the element identification feature, the connecting piece identification feature and the connection relation identification feature, and therefore efficiency of PCB patch detection is improved.

Further, before the step S20 "identifying features of the obtained first high-definition image information to obtain the processed image of the feature information set", it includes:

s201, a definition obtaining unit comprises a main body marking subunit, a main body characteristic subunit and an information saturation subunit, and obtains a first image definition of a PCB patch according to first high-definition image information and transmits the first image definition to a processing center;

s202, the processing center compares the acquired first image definition with the image definition conditions stored in the memory: if the information is consistent, the information is transmitted to the element characteristic unit, if the information is inconsistent, the information is transmitted to the alarm and the information reminding unit is informed to carry out rechecking;

and S203, the information reminding unit sends out reminding information to inform the image re-acquisition according to the fact that the definition of the first image is inconsistent with the definition condition of the stored image, and transmits the reminding information to the definition obtaining unit.

Further, before the step S201 "obtaining the first image definition of the PCB board patch according to the first high-definition image information", the method includes the following steps:

s2011, a main body marking subunit carries out main body identification of the element and the connecting piece on the PCB patch information according to the first image information and transmits the main body identification to a main body characteristic subunit;

S2012, the main body characteristic subunit extracts convolution kernel characteristics of the main body according to the identification, obtains main body convolution characteristics and transmits the main body convolution characteristics to the information saturation subunit;

and S2013, the information saturation subunit performs saturation processing on the first image based on the main body convolution characteristic so as to improve the information saturation of the first image.

Further, in the step S30, "the planar point cloud data of the connection attribute information of each component in the PCB patch is obtained according to the processed image", which includes the following steps:

s31, the pixel selection unit randomly selects one pixel point I from a plurality of pixel points in the second depth image and transmits the pixel point I to the world coordinate unit;

s32, the world coordinate unit converts a plurality of pixel points from an image coordinate system to a world coordinate system according to the pixel points I and transmits the pixel points to the plane data unit;

and S33, the plane data unit obtains plane point cloud data information in the image according to the world coordinate system of the high-definition camera image.

Further, in the step S50, "the processed cloud data according to the patch detection standard information to obtain a plurality of patch feature identification information", the method includes the following steps:

s51, setting a plurality of parallel equipotential lines in the processed cloud data by the equipotential discrete unit, performing discrete processing to form a plurality of discrete points, and transmitting the discrete points to the search area unit;

S52, the searching area unit sets the coordinates of a certain discrete point P as (Px, py and Pz), the discrete point P as the sphere center and the surrounding sphere with the radius of R as the searching area, and transmits the searching area to the feature calculating unit;

s53, the feature calculation unit obtains coordinates (Ox, oy, oz) of a plurality of feature points O of the whole search area according to an ant colony algorithm, and transmits the coordinates to the target selection unit; the method also comprises the following steps:

s531, a transition probability unit obtains the probability of node transition according to a node i-j transition probability calculation formula 'ρij= (1- ρ): ij+Deltaρij, ρji is the probability of node i-j transition, ρ is the parameter of consistency of characteristic point search behaviors, ij is the distance of node i-j transition, deltaρji is the parameter of the passing number of node i-j paths', and the probability is transmitted to a path parameter unit;

s532, the path parameter unit obtains path passing number parameters according to a parameter calculation formula of the passing number of the node paths, wherein Deltaρij=q/Lij+ (1-q) Deltaρij, lij is the length of the node i to j paths, q is the parameter of consistency of characteristic point searching behaviors, deltaρji is the parameter of the passing number of the node i to j paths, and the parameters are transmitted to the characteristic coordinate unit;

s533, the feature coordinate unit obtains the probability of node transition according to a calculation formula of' tji = (1- ρ) = tij+delta tij, tji is the probability of node j to i transition, ρ is a parameter of consistency of feature point search behavior, delta tij is a parameter of the passing times of a path from node i to j;

S54, when the target selection unit is equal to (Ox, oy, oz) Px, py, pz, the characteristic point O is overlapped with the discrete point P, is the target point and is transmitted to the characteristic recognition unit;

s55, the feature recognition unit obtains a plurality of feature recognition information from all target point sets according to a least square method.

The invention provides an automatic detection control system for a PCB patch, which also comprises computer auxiliary equipment and a computer readable storage medium; the computer auxiliary equipment comprises a memory, a processing center and functional modules thereof, wherein the memory stores a computer program, and the functional modules realize the steps of the automatic detection method for the PCB patch, which is described in any one of the above, when executing the computer program; the computer readable storage medium stores a computer program, and the computer program realizes the steps of the automatic detection method for the PCB patch described in any one of the above when executed by each functional module.

The invention also provides an automatic detection device for the PCB patch, which is realized by the automatic detection method for the PCB patch.

Compared with the prior art, the invention has the beneficial effects that:

through setting up image acquisition module, characteristic information module, connection attribute module, paster information module, relation of connection module, can carry out comprehensive and deep automated inspection to the PCB board paster, solve the problem that the part detection item of PCB board paster was missed or can't be discerned to improve the rate of accuracy that detects, reduced detection cost, reduced repeated reworking, reduced manufacturing cost.

Drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings that are required for the description of the embodiments or exemplary techniques will be briefly described below, and it is apparent that the drawings in the following description are only some embodiments of the present application, and other drawings may be obtained according to these drawings without inventive effort for a person of ordinary skill in the art.

FIG. 1 is a schematic diagram of a system module of the present invention;

FIG. 2 is a schematic diagram of a feature information module according to the present invention;

FIG. 3 is a schematic diagram of a clear acquisition unit of the present invention;

FIG. 4 is a schematic diagram of a sharpness enhancement unit of the present invention;

FIG. 5 is a schematic diagram of a connection attribute module according to the present invention;

FIG. 6 is a schematic diagram of a connection module according to the present invention;

FIG. 7 is a schematic diagram of a feature calculation unit of the present invention;

FIG. 8 is a flow chart of the method of the present invention;

FIG. 9 is a schematic diagram of the process of step S20 in the method of the present invention;

FIG. 10 is a schematic diagram of the process of step S21 in the method of the present invention;

FIG. 11 is a schematic diagram illustrating the procedure before step S20 in the method flow of the present invention;

FIG. 12 is a schematic diagram showing the procedure before step S201 in the method flow of the present invention;

FIG. 13 is a schematic diagram of the process of step S30 in the method of the present invention;

FIG. 14 is a flowchart of the step S50 in the method of the present invention;

fig. 15 is a schematic diagram of the process of step S53 in the method flow of the present invention.

Detailed Description

The present invention will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.

The following describes in detail the implementation of the present invention in connection with specific embodiments:

in order to make the technical problems, technical schemes and beneficial effects to be solved by the present application more clear, the present application is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the present application.

It should be noted that when a module is referred to as being "disposed on" another module, it can be directly on the other module or be indirectly on the other module. When a module is referred to as being "connected to" another module, it can be directly connected to the other module or be indirectly connected to the other module.

In the description of the present application, the meaning of "a plurality" is two or more, unless explicitly defined otherwise. The meaning of "a number" is one or more than one unless specifically defined otherwise.

In the description of the present application, it should be noted that, unless explicitly specified and limited otherwise, the terms "connected," "connected," and "connected" are to be construed broadly, and may be either fixedly connected, detachably connected, or integrally connected, for example; can be mechanically or electrically connected; can be directly connected or indirectly connected through an intermediate medium, and can be communicated with the inside of two elements or the interaction relationship of the two elements. The specific meaning of the terms in this application will be understood by those of ordinary skill in the art as the case may be. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless expressly specified otherwise.

Referring to fig. 1, the invention provides an automatic detection method for a PCB board, which is applied to an automatic detection control system for a PCB board, and includes an image acquisition module, a feature information module, a connection attribute module, a patch information module, a connection relation module, a wireless communication module, a memory, an alarm, a processing center, and a smart phone; the image acquisition module, the characteristic information module, the connection attribute module, the patch information module, the connection relation module, the wireless communication module, the memory and the alarm are all connected with the processing center; the wireless communication module is provided with an internet of things unit, is responsible for receiving and transmitting wireless signals, and is connected with the intelligent mobile phone in an effective internet of things or internet range; the intelligent mobile phone is automatically connected with the wireless communication module in a networking mode within the range of the Internet of things or the Internet.

The processing center is responsible for information transmission of an image acquisition module, a characteristic information module, a connection attribute module, a patch information module, a connection relation module, a wireless communication module, a memory and an alarm, is a hub center of the system, and compares the acquired first image definition with the image definition condition stored in the memory: if the information is consistent, the information is transmitted to the element characteristic unit, if the information is inconsistent, the information is transmitted to the alarm and the information reminding unit is informed to carry out rechecking; comparing the patch characteristic identification information with patch detection standards stored in a memory: if the detection result is consistent, a patch detection report is generated, if the detection result is inconsistent, the patch detection report is transmitted to an alarm and a reworking and rechecking is notified until the patch detection result is qualified.

Further, the image definition condition refers to a condition that the pixel and frame rate of the image definition are lower, if the definition is lower than the condition, the error rate of the equipment recognition result can be increased, if the first image definition does not meet the image definition condition, the definition of the camera is qualified, and reminding information is generated and used for reminding to carry out image recognition and acquisition again, so that the accuracy of the PCB patch recognition result is improved.

Further, the characteristic relation information refers to the mutual relation of connection between the elements, and is obtained based on the element identification feature, the connecting piece identification feature and the connection relation identification feature, and the connection relation between the elements is known, so that the PCB patch can be detected, and the judgment is performed to see whether the connection relation is correct.

Further, the patch test report is a table for recording the result of the patch of the PCB to be detected, is automatically generated according to the information fed back by the system, and indicates that the patch of the PCB is correct and automatically generates a detection report if the patch information is consistent with the patch standard; if the patch information is inconsistent with the patch standard, the patch information indicates that the PCB is wrong, a feedback report is generated and transmitted to a large database, and the correct connection relation interpretation of the wrong part is provided for a detection user side, so that the patch detection is fully automatic, and the detection efficiency and accuracy are ensured.

And the alarm automatically alarms and notifies the information reminding unit to carry out rechecking when the acquired first image definition is inconsistent with the image definition condition stored in the memory, and automatically alarms and notifies reworking to recheck until the patch characteristic identification information is qualified when the patch characteristic identification information is inconsistent with the patch detection standard stored in the memory.

The memory is responsible for information storage of an image acquisition module, a characteristic information module, a connection attribute module, a patch information module, a connection relation module, a wireless communication module and an alarm, and storage of image definition conditions and patch detection standards.

The image acquisition module obtains first high-definition image information of the positions of the soldered spots on the PCB through the arranged high-definition camera and transmits the first high-definition image information to the characteristic information module.

Further stated, the high-definition camera is arranged on the PCB patch device, and is mainly used for collecting images of the PCB patches to be detected, analyzing and identifying the images, so that whether the PCB patches are correct, whether patch errors such as missing welding, error welding or polarity inversion exist or not can be detected more quickly and accurately.

Further, the first high-definition image information refers to a high-definition image which is obtained by shooting a welding spot position of the patch to be detected after a reflow soldering process by using a high-definition camera and can fully cover a welding spot area, and the high-definition image information comprises characteristic information such as element information, element connecting piece information, element connecting relation information and the like.

Further illustratively, the component information includes: whether foreign matters exist on the surface, whether the surface is scratched, whether the specification of the device is correct, whether the component is neglected to be assembled, wrought assembled and empty welded, whether the marks of the polar elements are correct, whether the multi-pin device or the adjacent element bonding pad is free of tin connection, tin beads and tin residues, whether the short circuit is bridged, and whether the polarity of the component is reversely attached.

Further illustratively, the component connection information includes: whether the solder pin shifts beyond the width of the bonding pad by 1/4, whether the component shifts beyond the width of the bonding pad by 1/4 in the parallel direction and the vertical direction, whether the shifting exceeds the bonding pad by 1/3, whether the positions of the red glue points and the two bonding pads deviate too much, whether the red glue amount is moderate and no glue shortage exists.

Further, the element connection relation information includes: whether the device falls off, whether the solder paste position and the bonding pad are centered, whether the solder paste amount is moderate, whether tin is less, tin is leaked, whether the tin point is formed poorly, tin is connected, uneven, whether the red glue position is centered, and whether the device is attached neatly and centered.

Referring to fig. 2, the feature information module includes a clear obtaining unit, an information reminding unit, a clear enhancing unit, an element feature unit, an identifying feature unit, a feature information unit, a relationship feature unit, and a feature set unit, and is responsible for identifying features of the obtained first high-definition image information to obtain a feature information set, and transmitting the feature information set to the connection attribute module.

Referring to fig. 3, the sharpness enhancement unit includes an inverse mapping chart subunit, a full resolution subunit, a crack capturing subunit, a multiple adjustment subunit, a function integration subunit, and a mapping function subunit, and obtains a processed image with full resolution by an automatic data enhancement method according to the resolution of the first high-resolution image, and transmits the processed image to the element feature unit; the element characteristic unit obtains an element characteristic database according to the processed image with full resolution and transmits the element characteristic database to the identification characteristic unit; the identification feature unit identifies the first image information according to the element feature database to obtain element identification features, marks the element identification features and transmits the element identification features to the feature information unit; the characteristic information unit obtains element connection characteristic information according to the element identification characteristic and the element characteristic database and transmits the element connection characteristic information to the relation characteristic unit; the relation feature unit recognizes the element recognition features marked in the first image information according to the element connection feature information to obtain recognition features of the connection piece and the connection relation, and transmits the recognition features to the feature set unit; the feature set unit obtains an identification feature set according to the element identification feature, the connector identification feature, and the connection relation identification feature.

Further, the inverse mapping map subunit encodes the first image into a feature map, extracts various feature information, connects each feature information in series to obtain an image low-resolution inverse mapping map, and transmits the image low-resolution inverse mapping map to the full-resolution subunit; the full resolution subunit samples each pixel point in the low resolution inverse mapping chart in detail and multiplies the pixel points to obtain a full resolution mapping chart, and transmits the full resolution mapping chart to the crack capturing subunit; the crack capturing subunit performs multiplication on each pixel point of the full resolution mapping chart and the enhanced image thereof to obtain a full resolution depth image, and transmits the full resolution depth image to the multiple adjusting subunit, so that the difficulty in extracting the characteristic points can be reduced, even if tiny cracks appear at the welding point position can be clearly captured, and the scaling of the welding point position and the region enhancement are facilitated; the multiple adjustment subunit performs amplification or reduction or maintenance of the designated multiple on the full-resolution depth image according to the machine designated probability and the scaling ratio, and transmits the amplification or reduction or maintenance of the designated multiple to the area parameter subunit; the function integration subunit calculates the mapping function integration according to an area ratio parameter calculation formula 'P=V/S, wherein P is an area ratio parameter, V is a mapping function integration, S is an image area after scaling or maintaining the original state', and the mapping function integration is transmitted to the mapping function subunit; the mapping function subunit calculates a formula V= ≡according to the mapping function integral ^H ∫ ^W Alpha (x, y) dxdy, V is the integral of the mapping function, H is the height of the image, W is the width of the image, alpha (x, y) is the Gaussian mapping function of the image, and d is the area of each micro-bin on the curved surface in the image.

Further, the element characteristic database records characteristic data of all elements, integrates the characteristic data according to big data sharing and document records, and updates the characteristic data in real time to ensure the integrity of the content of the database; the component connection characteristic information is acquired based on a component characteristic database according to component identification characteristics, and corresponding connector identification characteristics in the database can be obtained through the component identification characteristics and the component connection characteristics.

Further, the relation feature unit recognizes the element recognition feature marked in the first image information according to the element connection feature information to obtain recognition features of the connection piece and the connection relation, and transmits the recognition features to the feature set unit.

Further, the feature set unit obtains the feature set according to the element identification feature, the connecting piece identification feature and the connection relation identification feature, so that the efficiency of PCB patch detection is improved.

Further, the identification feature means that the elements of the patch are identified according to the features of the image information, the patch information is analyzed, the coordinates of the elements are positioned, element information, element connecting piece information, element connection relations and the like are obtained, feature identification is performed on the elements, and a feature set is formed by arrangement, so that the patch detection efficiency is improved.

Further, the feature information set refers to the above element information, element connector information, element connection relation information and other various feature information already pasted on the PCB board.

Referring to fig. 4, the definition obtaining unit includes a main body marking subunit, a main body feature subunit, and an information saturation subunit, and obtains a first image definition of the PCB patch according to the first high-definition image information, and transmits the first image definition to the processing center.

Further, before the "obtaining the first image definition of the PCB patch according to the first high-definition image information" includes: the main body marking subunit carries out main body identification of the element and the connecting piece on the PCB patch information according to the first image information and transmits the main body identification to the main body characteristic subunit; the main body characteristic subunit performs convolution kernel characteristic extraction on the main body according to the identification, obtains main body convolution characteristics and transmits the main body convolution characteristics to the information saturation subunit; the information saturation subunit performs saturation processing on the first image based on the subject convolution feature to increase the first image information saturation.

Further, the "obtaining the first image definition of the PCB patch" is to detect the definition of the high-definition camera before the image acquisition of the PCB patch to be detected by the high-definition camera, the higher the first image definition is, the more accurate the recognition result obtained when the device analyzes the first image definition is, so as to improve the patch detection efficiency; if the first image has low definition, a certain element, a connecting piece and a connecting relation may not be identified, or the identification is wrong, which is important for the final detection result.

And the information reminding unit sends out reminding information to inform the image acquisition unit to acquire the image again according to the fact that the definition of the first image is inconsistent with the definition condition of the stored image, and transmits the acquired image to the definition acquisition unit.

Referring to fig. 5, the connection attribute module includes a pixel selection unit, a world coordinate unit, and a plane data unit, and obtains connection attribute information of each component in the PCB patch according to the feature information set, and transmits the connection attribute information to the patch information module.

Further, the pixel selection unit randomly selects one pixel point I from a plurality of pixel points in the second depth image and transmits the pixel point I to the world coordinate unit; the world coordinate unit converts a plurality of pixel points from an image coordinate system to a world coordinate system according to the pixel points I and transmits the pixel points to the plane data unit; and the plane data unit obtains plane point cloud data information in the image according to the world coordinate system of the high-definition camera image.

And the patch information module is connected with the attribute information according to the connection attribute information of each component in the patch to obtain patch detection standard information and transmits the patch detection standard information to the connection relation module.

Further, the connection attribute information refers to what elements are, how elements are connected, what relationships the elements are, such as element feature database information; the patch detection standard information refers to a correct connection mode of the connection relation between the elements, is derived from the connection attribute information and is used for judging with the electronic components to be detected, and the correctness of the connection relation between the elements is identified, so that the accuracy of patch detection is improved.

Referring to fig. 6, the connection relationship module includes an equipotential discrete unit, a search area unit, a feature calculation unit, a target selection unit, and a feature recognition unit, and obtains patch feature recognition information according to the element information feature, the element connector information feature, and the element connection relationship information feature, and transmits the patch feature recognition information to the processing center.

Referring to fig. 7, the feature calculation unit includes a transition probability subunit, a path parameter subunit, and a feature coordinate subunit, obtains coordinates (Ox, oy, oz) of a plurality of feature points O of the entire search area according to an ant colony algorithm, and transmits the coordinates to the target selection unit; the equipotential discrete unit sets a plurality of parallel equipotential lines in the processed cloud data, performs discrete processing on the parallel equipotential lines into a plurality of discrete points, and transmits the discrete points to the search area unit; the searching area unit sets coordinates of a certain discrete point P as (Px, py and Pz), the discrete point P as a sphere center and a surrounding sphere with a radius of R as a searching area, and transmits the searching area to the feature calculating unit; the target selecting unit is used for overlapping the characteristic point O with the discrete point P when the (Ox, oy, oz) is equal to the (Px, py, PZ), and transmitting the characteristic point O to the characteristic identifying unit, wherein the characteristic point O is a target point; the feature recognition unit fits all target points according to a least square method to obtain a plurality of feature recognition information, so that fine welding spot cracks are recognized by collecting all determined target points into a smooth curve, and the detection accuracy of fine line defects is improved.

Further described, the transition probability subunit calculates the formula "ρij= (1- ρ) according to the node i to j transition probability

) ij+Deltaρij, ρji are the probability of node i to j transition, ρ is the parameter of consistency of the feature point search behavior, ij is the distance of node i to j transition, deltaρji is the parameter of the passing number of node i to j path, the probability of node transition is obtained, and the probability is transmitted to the path parameter subunit; the path parameter subunit obtains path passing number parameters according to a parameter calculation formula of the passing number of the node paths, wherein ' Deltaρij=q/Lij+ (1-q) ' Deltaρij, lij is the length of the node i to j paths, q is the parameter of consistency of characteristic point searching behaviors, deltaρji is the parameter of the passing number of the node i to j paths ', and the parameters are transmitted to the characteristic coordinate subunit; the feature coordinate subunit obtains the probability of node transition according to a calculation formula of 'tji = (1- ρ): tij +:. DELTAtij of the probability of node j to i transition, tji is the probability of node j to i transition, ρ is a parameter of consistency of feature point search behavior, and DELTAtij is a parameter of the passing times of the path of node i to j'.

System operation theory:

after the PCB is pasted, the image acquisition module is controlled to obtain first high-definition image information of the welding spot position of the pasted PCB after the reflow soldering process through the high-definition camera, and the first high-definition image information is transmitted to the characteristic information module; identifying the characteristics of the obtained first high-definition image information through the characteristic information module to obtain a processed image of the characteristic information set, and transmitting the processed image to the connection attribute module; the control connection attribute module obtains the plane point cloud data of the connection attribute information of each component in the PCB patch according to the processed image and transmits the plane point cloud data to the patch information module; the patch information module is used for obtaining processed cloud data of patch detection standard information according to the planar point cloud data of the connection attribute information of each component in the patch and transmitting the processed cloud data to the connection relation module; the control connection relation module obtains a plurality of patch characteristic identification information according to the processed cloud data of the patch detection standard information, marks the patch characteristic identification information to obtain a detection result, and transmits the detection result to the processing center; comparing the patch characteristic identification information with patch detection standards stored in a memory through a processing center: if the detection result is consistent, a patch detection report is generated, if the detection result is inconsistent, the patch detection report is transmitted to an alarm and a reworking and rechecking is notified until the patch detection result is qualified.

When the operator or manager is outdoors or out of the business, the intelligent mobile phone can be used for automatically networking connection in the network range of the Internet or the Internet of things through the wireless communication module, so that the detection device can be controlled in a networking and intelligent mode.

Referring to fig. 8, the method for automatically detecting a PCB board provided by the present invention includes the following steps:

further, the high-definition image refers to the definition of shooting the first depth image, the welding spot area can be covered comprehensively, the depth image is converted into cloud data, the characteristic identification information features appearing in the welding spot position are obtained through selecting and fitting of equipotential lines, fine characteristic identification information appearing in the welding spot position can be accurately detected, the technical problems of missing detection and false detection caused by fuzzy characteristic expression of the fine welding spot characteristic identification information in detection are solved, and the accuracy of detecting the defects of the welding spot characteristic identification information is improved.

further, the patch characteristic identification information includes element information characteristics, element connector information characteristics, element connection relation information characteristics.

Further, the "obtaining a plurality of patch feature identification information and marking to obtain a detection result" refers to that feature identification information exists at a patch welding spot position to be detected, marking the welding spot position, outputting a smooth curve target feature picture, and recording in detail an area or a position where each target feature exists, where the welding spot feature identification information may exist, and the welding spot position where the feature identification information is not found after detection, so that the welding spot position where the feature identification information is not found after detection needs to be marked, the position is marked as OK or other forms capable of representing good products, and a specific area or position is attached to the position, so that the inspection of a plurality of welding spot positions in later period is facilitated, and the phenomenon of repeated detection is avoided.

S60, comparing the patch characteristic identification information with patch detection standards stored in a memory by the processing center: if the detection results are consistent, a patch detection report is generated, if the detection results are inconsistent, the patch detection report is transmitted to an alarm and is notified to reworking and rechecking until the patch is qualified, so that the defects of the tiny feature identification information existing after reflow soldering of the patch are accurately detected, the missing detection rate of the detection of the feature identification information of the welding spot is further reduced, and the precision and the high efficiency of the detection of the tiny defects are improved.

Referring to fig. 9, the step S20 "the identification feature of the obtained first high definition image information to obtain the processed image of the feature information set" includes the following steps:

s21, the clear enhancement unit obtains a processed image with full resolution by adopting an automatic data enhancement method according to the resolution of the first high-definition image, and transmits the processed image to the element characteristic unit;

s26, the feature set unit obtains an identification feature set according to the element identification feature, the connecting piece identification feature and the connection relation identification feature so as to improve the efficiency of PCB patch detection.

Referring to fig. 10, in the step S21, "a full-resolution processed image is obtained by using an automatic data enhancement method", the method includes the following steps:

s211, the inverse mapping map subunit codes the first image into a feature map, extracts various feature information, connects each feature information in series to obtain an image low-resolution inverse mapping map, and transmits the image low-resolution inverse mapping map to the full-resolution subunit;

s212, the full resolution subunit samples each pixel point in the low resolution inverse mapping chart in detail and multiplies the pixel points to obtain a full resolution mapping chart, and transmits the full resolution mapping chart to the crack capturing subunit;

s213, the crack capturing subunit multiplies each pixel point of the full resolution mapping chart and the enhanced image thereof to obtain a full resolution depth image, and transmits the full resolution depth image to the multiple adjusting subunit, so that the difficulty in extracting the characteristic points can be reduced, even if tiny cracks appear at the welding point position can be clearly captured, and the scaling of the welding point position and the region enhancement are facilitated;

S214, the multiple adjustment subunit performs the magnification or reduction of the designated multiple or maintains the original state on the full-resolution depth image according to the machine designated probability and the scaling ratio, and transmits the magnification or reduction to the area parameter subunit;

further, the specified probability and the scaling ratio are respectively inversely proportional to the image magnification, and if the random specified probability is smaller than 0.15 and the scaling ratio is smaller than 0.3, the first depth image with full resolution is magnified by 2-6 times; if the random specified probability is equal to 0.15 and the scaling ratio is equal to 0.3, maintaining the first depth image with full resolution as it is; if the random assignment probability is larger than 0.15 and the scaling ratio is larger than 0.3, the first depth image with full resolution is scaled down by 3-5 times.

S215, the function integration subunit calculates the mapping function integration according to an area ratio parameter calculation formula 'P=V/S, wherein P is an area ratio parameter, V is a mapping function integration, S is an image area after scaling or maintaining the original state', and the mapping function integration is transmitted to the mapping function subunit;

s216, the mapping function subunit calculates a formula 'according to the integral calculation of the mapping function'V is the mapping function productThe size of an image enhancement region is calculated for the area of each tiny bin on the curved surface in the image, so that the size of the enhancement region in the depth image with full resolution can be automatically amplified by adjusting the area ratio parameter P, the size of the welding spot position in the image can be automatically amplified, the problem that the cracks in the image change is solved, and the definition of pixel points in the first depth image is improved.

Furthermore, the first depth image is preprocessed, the welding spot position to be detected can be scaled according to multiple, meanwhile, the enhancement area is adjusted according to the area ratio parameter P in the scaling process, so that the characteristics of the welding spot position in the first depth image are unchanged, the resolution is improved, meanwhile, the definition of all the characteristics of the welding spot position is enhanced, the integrity of converting the depth image into cloud data is improved, and the phenomenon that characteristic points are lost is avoided.

Referring to fig. 11, before the step S20 "the identification feature of the obtained first high definition image information to obtain the processed image of the feature information set", it includes:

Referring to fig. 12, before the step S201 "obtaining the first image definition of the PCB patch according to the first high-definition image information", the method includes the following steps:

Referring to fig. 13, in the step S30, the step of obtaining planar point cloud data of connection attribute information of each component in the PCB patch according to the processed image includes the following steps:

further, the image coordinate system refers to a coordinate system used on an imaging plane when the image is shot by a high-definition camera, an intersection point of the imaging plane and an optical axis of the high-definition camera is an origin, and the coordinates are expressed by (X, Y); the world coordinate system refers to the real coordinates representing an object in the three-dimensional world, the coordinates are expressed by (Xw, yw, zw), and w is a world abbreviation; the four-coordinate system of the high-definition camera image is formed by the four-coordinate system, the high-definition camera coordinate system and the pixel coordinate system, the high-definition camera coordinate system is a coordinate representing an origin with the optical center of the high-definition camera, the Z axis coincides with the optical axis, the coordinate is expressed by (Xc, yc, zc), and c is camera abbreviation; the pixel coordinate system is an image on the imaging plane of the high-definition camera, and cannot be directly used under normal conditions, a new coordinate system is defined to represent the image displayed on the electronic device, the upper left corner of the image is the origin, and the coordinate system is represented by (u, v).

Referring to fig. 14, in the step S50, "the processed cloud data according to the patch detection standard information to obtain a plurality of patch feature identification information" includes the following steps:

further, the bounding sphere is a set search area, and the target points are selected from all points in the area, so that enough target points can be obtained, all the target points form a track, and then all the target points are fitted.

S53, the feature calculation unit obtains coordinates (Ox, oy, oz) of a plurality of feature points O of the whole search area according to an ant colony algorithm, and transmits the coordinates to the target selection unit;

furthermore, the ant colony algorithm is a probability-based search algorithm, and a multi-dimensional and interrelated search space is formed by adopting a mutual learning mode and through information sharing among populations.

further, the feature point overlapping with the discrete point includes that two points are completely overlapped, two points are partially overlapped or two points are adjacent, wherein the two points are adjacent to all points located on adjacent sides of the discrete point.

And S55, fitting all target points by the feature recognition unit according to a least square method to obtain a plurality of feature recognition information, so that fine welding spot cracks are recognized by collecting all determined target points into a smooth curve, and the detection accuracy of the fine line defects is improved.

In this embodiment, a plurality of equipotential lines are constructed in the second planar point cloud data, the points which are coincident with or partially coincident with or adjacent to each discrete point are selected as target points, and finally, a least square method is adopted to fit all the target points into a smooth curve as a target feature, so that the welding point cracks existing in the welding point positions of the patches are obtained.

Referring to fig. 15, the step S53 "obtaining coordinates (Ox, oy, oz) of a plurality of feature points O of the entire search area according to the ant colony algorithm" includes the following steps:

S531, a transition probability subunit obtains the probability of node transition according to a node i-j transition probability calculation formula 'ρij= (1- ρ) = ij+Δρij, ρji is the probability of node i-j transition, ρ is the parameter of consistency of characteristic point search behaviors, ij is the distance of node i-j transition, Δρji is the parameter of the passing number of node i-j paths', and the probability is transmitted to a path parameter subunit;

s532, the path parameter subunit obtains path passing number parameters according to a parameter calculation formula of the passing number of the node paths, wherein Deltaρij=q/Lij+ (1-q) Deltaρij, lij is the length of the node i to j paths, q is the parameter of consistency of characteristic point searching behaviors, deltaρji is the parameter of the passing number of the node i to j paths, and the parameters are transmitted to the characteristic coordinate subunit;

s533, the feature coordinate subunit obtains the probability of node transition according to a calculation formula of' tji = (1- ρ) = tij +. DELTAtij, tji is the probability of node j to i transition, ρ is a parameter of consistency of feature point search behaviors, and DELTAtij is a parameter of the passing times of the paths of the nodes i to j.

Further, by the above iterative method, the best path is found, and the length of the path is determined by the probability of transition between nodes, and the change of the path is determined by the number of passes between nodes.

Further, the invention is described in terms of the implementation of a software program of a PCB patch automatic detection control system, and for each PCB patch automatic detection method, the software program instructions generated by each step are equally divided into a plurality of modules or units, and the software program instructions include a PCB patch automatic detection method as described above.

Further, the content described above is written according to the content implemented by the software copyright "create core intelligent sink printed circuit board quality data acquisition tracking system (version number: V1.0, registration date: 2019, 3 month, 29 days)", "create core intelligent sink PCBA board ICT test interaction operation and maintenance system (version number: V1.0, registration date: 2022, 5 month, 31 days)", "create core intelligent sink PCBA board high-efficiency detection operation and maintenance interaction system (version number: V1.0, registration date: 2022, 6 month, 17 days)".

It will be evident to those skilled in the art that the present application is not limited to the details of the foregoing illustrative embodiments, and that the present application may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive, the scope of the application being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein.

Finally, it should be noted that the foregoing embodiments are merely for illustrating the technical solution of the present application and not for limiting the same, and although the present application has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications and equivalents may be made to the technical solution of the present application, and are included in the protection scope of the present application.

Claims

1. An automatic detection method for a PCB board is characterized in that: the system comprises an image acquisition module, a characteristic information module, a connection attribute module, a patch information module, a connection relation module, a wireless communication module, a memory, an alarm, a processing center and a smart phone; the image acquisition module, the characteristic information module, the connection attribute module, the patch information module, the connection relation module, the wireless communication module, the memory and the alarm are all connected with the processing center; the wireless communication module is provided with an internet of things unit, is responsible for receiving and transmitting wireless signals, and is connected with the intelligent mobile phone in an effective internet of things or internet range; the intelligent mobile phone is automatically connected with the wireless communication module in a networking way within the range of the Internet of things or the Internet;

2. The automatic detection method of a PCB board according to claim 1, wherein: the method comprises the following steps:

s10, the image acquisition module obtains first high-definition image information of a welding spot position of the PCB after the surface mount is subjected to a reflow soldering process through a high-definition camera, and transmits the first high-definition image information to the characteristic information module;

3. The automatic detection method of a PCB board according to claim 2, wherein: the step S20 "identifying features of the obtained first high-definition image information to obtain a processed image of the feature information set", includes the steps of:

s216, the mapping function unit calculates a formula according to the integral of the mapping functionV is the integral of the mapping function, H is the height of the image, W is the width of the image, alpha (x, y) is the Gaussian mapping function of the image, and d is the area' of each tiny bin on the curved surface in the image to calculate the size of the image enhancement area;

4. The automatic detection method for a PCB according to claim 3, wherein: the step S20 "before identifying features of the obtained first high-definition image information to obtain a processed image of the feature information set" includes:

5. The automatic detection method for the PCB according to claim 4, wherein: before the step S201 "obtaining the first image definition of the PCB patch according to the first high-definition image information", the method includes the following steps:

6. The automatic detection method of a PCB board according to claim 2, wherein: in the step S30, "the planar point cloud data of the connection attribute information of each component in the PCB patch is obtained according to the processed image", which includes the following steps:

s33, the plane data unit obtains plane point cloud data information in the image according to the world coordinate system of the camera image.

7. The automatic detection method of a PCB board according to claim 2, wherein: in the step S50, "the processed cloud data according to the patch detection standard information to obtain a plurality of patch feature identification information", the method includes the following steps:

S532, the path parameter unit calculates a formula of 'Deltaρij=q/Lij' according to the parameter of the passing number of the node paths

A + (1-q) Δρij, wherein Lij is the length of the paths from node i to j, q is the parameter of the consistency of the characteristic point search behavior, Δρji is the parameter of the passing number of the paths from node i to j, and the path passing number parameter is obtained and transmitted to the characteristic coordinate unit;

s55, the feature recognition unit gathers all target points according to a least square method to obtain a plurality of feature recognition information.

8. The automatic detection method of a PCB board according to claim 1, wherein: also included is a computer-assisted device, a computer-readable storage medium; the computer auxiliary equipment comprises a memory, a processing center and functional modules thereof, wherein the memory stores a computer program, and the functional modules realize the steps of the automatic detection method for the PCB patch according to any one of the claims 1-7 when executing the computer program; the computer readable storage medium stores a computer program which, when executed by each functional module, implements the steps of a method for automatically detecting a PCB board patch according to any one of claims 1 to 7.

9. Automatic detection device of PCB board paster, its characterized in that: the method is realized by the automatic detection method for the PCB patch according to the claims 1-8.