CN116879708A - Method for detecting faults of printed circuit board by utilizing infrared cloud picture recognition technology - Google Patents
Method for detecting faults of printed circuit board by utilizing infrared cloud picture recognition technology Download PDFInfo
- Publication number
- CN116879708A CN116879708A CN202310624217.9A CN202310624217A CN116879708A CN 116879708 A CN116879708 A CN 116879708A CN 202310624217 A CN202310624217 A CN 202310624217A CN 116879708 A CN116879708 A CN 116879708A
- Authority
- CN
- China
- Prior art keywords
- circuit board
- printed circuit
- infrared cloud
- state
- image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 33
- 238000005516 engineering process Methods 0.000 title claims abstract description 15
- 238000001514 detection method Methods 0.000 claims abstract description 24
- 230000020169 heat generation Effects 0.000 claims abstract description 12
- 238000013021 overheating Methods 0.000 claims abstract description 10
- 238000004781 supercooling Methods 0.000 claims abstract description 10
- 238000003745 diagnosis Methods 0.000 claims abstract description 8
- 238000007781 pre-processing Methods 0.000 claims abstract description 7
- 230000009466 transformation Effects 0.000 claims description 12
- 230000000873 masking effect Effects 0.000 claims description 9
- 238000010586 diagram Methods 0.000 claims description 7
- 238000003909 pattern recognition Methods 0.000 claims description 7
- 238000012937 correction Methods 0.000 claims description 5
- 238000000605 extraction Methods 0.000 claims description 5
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 claims description 4
- 238000009826 distribution Methods 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000003331 infrared imaging Methods 0.000 claims description 3
- 230000000007 visual effect Effects 0.000 claims description 3
- 230000003287 optical effect Effects 0.000 abstract description 11
- 230000007547 defect Effects 0.000 abstract description 9
- 238000003384 imaging method Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 238000005259 measurement Methods 0.000 description 3
- 238000013139 quantization Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 238000003709 image segmentation Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 230000032683 aging Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000011176 pooling Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000001931 thermography Methods 0.000 description 1
- 230000001131 transforming effect Effects 0.000 description 1
- 238000013024 troubleshooting Methods 0.000 description 1
- 238000003466 welding Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/28—Testing of electronic circuits, e.g. by signal tracer
- G01R31/2801—Testing of printed circuits, backplanes, motherboards, hybrid circuits or carriers for multichip packages [MCP]
- G01R31/281—Specific types of tests or tests for a specific type of fault, e.g. thermal mapping, shorts testing
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N25/00—Investigating or analyzing materials by the use of thermal means
- G01N25/72—Investigating presence of flaws
Abstract
The invention relates to a method for detecting faults of a printed circuit board by utilizing an infrared cloud picture identification technology, which comprises the following steps: an original infrared cloud picture of the printed circuit board; preprocessing an original infrared cloud image, and identifying the type and the position of an element in a multi-scale characteristic pyramid module by utilizing a single-channel image mode; extracting the element heat generation rate from the local infrared cloud image at the element as a state characteristic to generate a state characteristic list; and comparing the state characteristics of the tested printed circuit board with the state characteristics of the normal printed circuit board in the fault diagnosis module, marking the state of the element as normal, supercooling or overheating, and diagnosing faults. The method can detect the defects of invisible optical detection such as the occlusion on the printed circuit board, the smaller geometric dimension and the like, overcomes the defect of automatic optical detection, and improves the non-intrusive detection level of the printed circuit board.
Description
Technical Field
The invention relates to the technical field of printed circuit board testing, in particular to a method for detecting faults of a printed circuit board by utilizing an infrared cloud pattern recognition technology.
Background
Printed circuit boards are an important component of electronic information products. During the full life cycle of electronic information products, printed circuit boards may fail due to manufacturing process imperfections or aging. Traditional intervening printed circuit board detection methods require measurement of electrical parameters such as voltage, current, etc. at critical locations in the circuit. The technician determines the fault location and type of the printed circuit board empirically based on the circuit principle and the measurement of the electrical parameters. The whole test process is carried out on the basis of fully grasping the schematic diagram of the printed circuit board and the position information of the components, and a great deal of time and labor are required. With the increase of circuit integration, the size of the components is smaller and the layout is denser, so a non-intrusive, accurate and rapid fault detection method is urgently needed.
The automatic optical detection (AutomatedOpticalInspection, AOI) can detect obvious visible defects such as welding and the like on the printed circuit board based on an optical principle, and can realize non-intrusive detection, classification and identification of various surface visible faults on the printed circuit board, but has limited detection capability on defects such as shielded, smaller geometric dimension and the like on the printed circuit board, and damage and parameter change inside elements.
The infrared thermal imaging technology is a mature non-interventional fault detection technology, and the working state and fault information of a detected object are analyzed by collecting the temperature distribution of the detected object. In the operation of printed circuit boards, a large number of information acquisition, processing and control processes have to be completed, with concomitant consumption and conversion of electrical energy. Defects and faults that are not readily detectable by AOI change the thermal profile of the printed circuit board during operation, which changes must be reflected in the infrared cloud of the printed circuit board. The information in the infrared cloud picture is fully mined, the element parameters and states can be obtained, and fault diagnosis is further carried out on the printed circuit board.
Disclosure of Invention
In order to solve the technical problems, the invention provides a method for detecting faults of a printed circuit board by utilizing an infrared cloud pattern recognition technology. The technical problems to be solved by the invention are realized by adopting the following technical scheme:
a method for detecting faults of a printed circuit board by utilizing an infrared cloud picture identification technology comprises the following steps:
the first step: obtaining an original infrared cloud image of the detected printed circuit board by using infrared imaging equipment;
and a second step of: preprocessing an original infrared cloud image, and correcting the image to ensure that the shapes and the sizes of all printed circuit board areas are consistent;
and a third step of: identifying element types and positions in a multi-scale feature pyramid module by utilizing single-channel image modes;
fourth step: inputting the type and the position of the element in the infrared cloud image identified in the third step into a state identification module, extracting the element heat generation rate from the local infrared cloud image at the element as a state characteristic, and generating a state characteristic list;
fifth step: and comparing the state characteristics of the tested printed circuit board with the state characteristics of the normal printed circuit board in the fault diagnosis module, marking the state of the element as normal, supercooling or overheating, and diagnosing faults.
The preprocessing of the original infrared cloud image comprises the following steps:
the first step: removing the background and removing the area outside the printed circuit board in the infrared cloud picture;
and a second step of: and (3) angle correction, namely correcting an inclined image in a three-dimensional coordinate into a plane image of the surface of the visual angle vertical circuit board by using a central projection mode and carrying out projection transformation and affine transformation, and carrying out size transformation into a uniform size to obtain the preprocessed infrared cloud image.
The identification element type and location in the third step comprises the steps of:
the first step: inputting the preprocessed infrared cloud image into a feature extraction backbone network, and obtaining three size feature images by using deep convolution calculation;
and a second step of: acquiring a suggestion frame from each size of feature graphs by using a regional suggestion network in a multi-scale feature pyramid module;
and a third step of: matching the obtained suggestion frame with the region of interest, intercepting a local public feature layer in the feature layer by using the suggestion frame, and carrying out category prediction and prediction frame regression;
fourth step: generating a masking layer on the feature map with the small primary size through a non-interested region masking network by utilizing a large primary size prediction frame, and inputting the feature layer filtered by the masking layer into a region suggestion network;
fifth step: sequentially identifying elements with different sizes according to the sequence from large to small by using a layering target detection method to obtain category prediction and prediction frame regression results;
sixth step: and dividing the original infrared cloud image to obtain a local infrared cloud image of each element and the adjacent area thereof according to the regression result of the element prediction frame calculated by the multi-scale characteristic pyramid module.
Generating the list of state features comprises the steps of:
the first step: inputting the local infrared cloud image generated in the third step into a state recognition network, wherein the output result of the network is the unit volume heat generation rate of the element;
and a second step of: taking the heat generated by the unit volume of the element as the state characteristics of the element, and identifying the state characteristics of all local infrared cloud pictures;
and a third step of: and summarizing the status features to obtain a printed circuit board status feature list.
Diagnosing a fault comprises the steps of:
the first step: calculating the proportional relation between the heat generation rate of each element unit volume on the board to be detected and the normal board by using the normal printed circuit board and the state characteristic list of the printed circuit board to be detected
And a second step of: judging the state of each element by using a threshold method, and marking as normal, supercooling or overheating;
and a third step of: according to the distribution rule of the supercooling or overheating elements, the fault elements are further positioned by combining the circuit schematic diagram.
The beneficial effects of the invention are as follows: on the basis of obtaining the infrared cloud image of the printed circuit board, the invention provides a method for applying the artificial intelligent target recognition technology to fault detection of the printed circuit board. The method is based on a FasterRCNN model in the target detection field, and aims at the characteristics of large number of printed circuit board elements, no overlapping and large size difference, a characteristic pyramid network suitable for high-density multi-scale target identification is provided, and the positioning of the printed circuit board elements from an infrared cloud picture is realized. On the basis of accurately extracting the position information of the components, the method takes the heat generation rate as the state characteristics of the components, designs a region extraction and state characteristic identification network, acquires the state parameters of each component, compares the state characteristics of the components corresponding to the tested printed circuit board and the normal working printed circuit board, and diagnoses faults according to the state parameters. The method can detect the defects of invisible optical detection such as the occlusion on the printed circuit board, the smaller geometric dimension and the like, overcomes the defect of automatic optical detection, and improves the non-intrusive detection level of the printed circuit board.
Drawings
The invention will be further described with reference to the drawings and examples.
FIG. 1 is a schematic flow chart of the method of the present invention;
FIG. 2 is a schematic diagram of a process flow of a multi-scale feature pyramid module of the present invention;
FIG. 3 is a schematic diagram of a state feature list generation flow of the present invention;
FIG. 4 is a schematic diagram of a diagnostic trouble shooting process according to the present invention.
Detailed Description
In order to make the technical solution of the present invention better understood by a person skilled in the art, the present invention will be more clearly and more fully described below with reference to the accompanying drawings in the embodiments, and of course, the described embodiments are only a part of, but not all of, the present invention, and other embodiments obtained by a person skilled in the art without making any inventive effort are within the scope of the present invention.
As shown in fig. 1 to 4, a method for detecting faults of a printed circuit board by using an infrared cloud pattern recognition technology, the method comprising the following steps:
the first step: obtaining an original infrared cloud image of the detected printed circuit board by using infrared imaging equipment; obtaining actual measurement temperature values of all positions on a circuit board which has the same structure and function as the tested printed circuit board and is in a known normal working state, and obtaining an original infrared cloud picture;
and a second step of: preprocessing an original infrared cloud image, and correcting the image to ensure that the shapes and the sizes of all printed circuit board areas are consistent;
and a third step of: identifying element types and positions in a multi-scale feature pyramid module by utilizing single-channel image modes;
fourth step: inputting the type and the position of the element in the infrared cloud image identified in the third step into a state identification module, extracting the element heat generation rate from the local infrared cloud image at the element as a state characteristic, and generating a state characteristic list;
fifth step: and comparing the state characteristics of the tested printed circuit board with the state characteristics of the normal printed circuit board in the fault diagnosis module, marking the state of the element as normal, supercooling or overheating, and diagnosing faults.
The preprocessing of the original infrared cloud image comprises the following steps:
the first step: removing the background and removing the area outside the printed circuit board in the infrared cloud picture; because the residual temperature of the surrounding environment of the printed circuit board is zero, and the circuit board self heats up due to the elements, temperature rise exists everywhere, and the residual temperature is not zero; therefore, removing the region with zero residual temperature in the infrared image, and keeping the region as the circuit board region;
and a second step of: angle correction, which uses a central projection mode to include projection transformation and affine transformation, so as to correct an inclined image under three-dimensional coordinates into a plane view of the surface of the visual angle vertical circuit board; the collected infrared cloud image is quadrilateral with any shape under the influence of imaging distance and angle, and the central projection transformation can correct the inclined image under the three-dimensional coordinate; the angle correction aims at transforming the printed circuit board area with any imaging angle into a regular rectangular area, so that the subsequent identification and detection are facilitated, the correction adopts central projection transformation, the projection transformation and affine transformation are included, the shapes and the sizes of all the transformed printed circuit board areas are consistent, and the transformed target size and angle can be given at will.
The identification element type and location in the third step comprises the steps of:
the first step: inputting the preprocessed infrared cloud image into a feature extraction backbone network, and obtaining three size feature images by using deep convolution calculation;
and a second step of: acquiring a suggestion frame from each size of feature graphs by using a regional suggestion network in a multi-scale feature pyramid module;
and a third step of: matching the obtained suggestion frame with the region of interest, intercepting a local public feature layer in the feature layer by using the suggestion frame, and carrying out category prediction and prediction frame regression;
fourth step: generating a masking layer on the feature map with the small primary size through a non-interested region masking network by utilizing a large primary size prediction frame, and inputting the feature layer filtered by the masking layer into a region suggestion network;
fifth step: sequentially identifying elements with different sizes according to the sequence from large to small by using a layering target detection method to obtain category prediction and prediction frame regression results;
sixth step: and dividing the original infrared cloud image to obtain a local infrared cloud image of each element and the adjacent area thereof according to the regression result of the element prediction frame calculated by the multi-scale characteristic pyramid module.
As shown in fig. 2, the preprocessed infrared cloud image I is input into a backbone network B based on a res net-50 to extract features, wherein a small-size feature layer C1 is an output of a conv4_x layer in the res net-50, a medium-size feature layer C2 is an output of a conv5_x layer in the res net-50, that is, an output of the backbone network of the res net-50, and a large-size feature layer C3 is a result of downsampling the medium-size feature layer by 2×2.
The large-size feature layer C3 is input into the regional suggestion network, and a large-size suggestion box P3 is generated.
The large-size suggestion frame P3 is matched through the region of interest to obtain a large-size local public feature layer F3, and a category prediction result C3 and a prediction frame regression result R3 are further obtained.
And mapping a prediction frame R3 with the element region in the large-size feature recognition into a middle-size feature layer C2, wherein grid points of a region corresponding to the middle-size feature layer C2 are covered by a non-interested region covering network to form a covering layer M3, the region does not generate a priori frame in RPN, and other regions generate a middle-size suggestion frame P2 and pass RoIAlign to obtain a middle-size local public feature layer F2, so that a category prediction result C2 and a prediction frame regression result R2 are further obtained.
Similarly, a prediction frame R2 with a class of element regions in the medium-size feature recognition is mapped into a small-size feature layer C1, grid points of regions corresponding to the small-size feature layer C1 form a mask layer M2 in an RPN through RnIMask, the regions do not generate prior frames, other regions generate small-size suggestion frames P1 and pass RoIAlign to obtain a small-size local public feature layer F1, and a class prediction result C1 and a prediction frame regression result R1 are further obtained.
Because the multi-scale feature pyramid realizes the separate identification of different scales, 9 prior frames of each feature layer grid point in the RPN are reduced to 3, and only 3 prior frames with medium size are reserved, so that the network complexity is reduced.
And RoIALign is used for replacing ROI pooling, so that the position accuracy of the feature map mapped onto the infrared cloud map is improved. RoIAlign cancels quantization operation, obtains pixel point values with coordinates of floating point numbers by using a bilinear interpolation method, and eliminates position quantization errors accumulated by two operations of quantization of the boundary of a candidate frame and average division of a quantized region.
The infrared cloud image segmentation is based on category prediction results C1, C2 and C3 and prediction frames R1, R2 and R3 obtained by the multi-scale feature pyramid, and local infrared cloud images I1 and I2 of each element of the normal printed circuit board and the printed circuit board to be detected and the adjacent area of each element are obtained by segmentation from the original infrared cloud image I.
Generating the list of state features comprises the steps of:
the first step: inputting the local infrared cloud image generated in the third step into a state recognition network, wherein the output result of the network is the unit volume heat generation rate of the element;
and a second step of: taking the heat generated by the unit volume of the element as the state characteristics of the element, and identifying the state characteristics of all local infrared cloud pictures;
and a third step of: and summarizing the status features to obtain a printed circuit board status feature list.
The local infrared cloud charts I1 and I2 are input into a state recognition network, and the output result of the network is the unit volume heat generation rate of the element. And identifying the state characteristics of all the local infrared cloud pictures to obtain a printed circuit board state characteristic list.
The infrared cloud patterns of the normal working state printed circuit board and the printed circuit board to be detected respectively pass through a state identification module to obtain state characteristic lists L1 and L2 which are used as the basis of subsequent fault diagnosis
Diagnosing a fault comprises the steps of:
the first step: calculating the proportional relation between the heat generation rate of each element unit volume on the board to be detected and the normal board by using the normal printed circuit board and the state feature lists L1 and L2 of the printed circuit board to be detected;
and a second step of: judging the state of each element by using a threshold method, marking as normal, supercooling or overheating, and using the state as a state comparison and fault diagnosis result O;
and a third step of: according to the distribution rule of the supercooling or overheating elements, the fault elements are further positioned by combining the circuit schematic diagram.
The printed circuit board detection method provided by the scheme can accurately and rapidly acquire the type and position information of all elements from the infrared cloud picture of the printed circuit board, automatically realize image segmentation and feature extraction, and provide sufficient and reliable basis for fault diagnosis. Compared with the existing automatic optical detection method, the method has the advantages that optical images are not required to be acquired, the defect that the imaging quality of the optical images depends on imaging conditions is overcome, meanwhile, the detection of optical invisible defects is realized, the detection range of automatic optical detection is expanded, and the non-intrusive detection level of a printed circuit board is improved.
The foregoing has shown and described the basic principles, principal 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, and that the above embodiments and descriptions are merely illustrative of the principles of the present invention, and various changes and modifications may be made without departing from the spirit and scope of the invention, which is defined in the appended claims. The scope of the invention is defined by the appended claims and equivalents thereof.
Claims (5)
1. A method for detecting faults of a printed circuit board by utilizing an infrared cloud picture recognition technology is characterized by comprising the following steps of: the method comprises the following steps:
the first step: obtaining an original infrared cloud image of the detected printed circuit board by using infrared imaging equipment;
and a second step of: preprocessing an original infrared cloud image, and correcting the image to ensure that the shapes and the sizes of all printed circuit board areas are consistent;
and a third step of: identifying element types and positions in a multi-scale feature pyramid module by utilizing single-channel image modes;
fourth step: inputting the type and the position of the element in the infrared cloud image identified in the third step into a state identification module, extracting the element heat generation rate from the local infrared cloud image at the element as a state characteristic, and generating a state characteristic list;
fifth step: and comparing the state characteristics of the tested printed circuit board with the state characteristics of the normal printed circuit board in the fault diagnosis module, marking the state of the element as normal, supercooling or overheating, and diagnosing faults.
2. The method for detecting faults of a printed circuit board by utilizing an infrared cloud pattern recognition technology as claimed in claim 1, wherein the method comprises the following steps of: the preprocessing of the original infrared cloud image comprises the following steps:
the first step: removing the background and removing the area outside the printed circuit board in the infrared cloud picture;
and a second step of: and (3) angle correction, namely correcting an inclined image in a three-dimensional coordinate into a plane image of the surface of the visual angle vertical circuit board by using a central projection mode and carrying out projection transformation and affine transformation, and carrying out size transformation into a uniform size to obtain the preprocessed infrared cloud image.
3. The method for detecting faults of a printed circuit board by utilizing an infrared cloud pattern recognition technology as claimed in claim 1, wherein the method comprises the following steps of: the identification element type and location in the third step comprises the steps of:
the first step: inputting the preprocessed infrared cloud image into a feature extraction backbone network, and obtaining three size feature images by using deep convolution calculation;
and a second step of: acquiring a suggestion frame from each size of feature graphs by using a regional suggestion network in a multi-scale feature pyramid module;
and a third step of: matching the obtained suggestion frame with the region of interest, intercepting a local public feature layer in the feature layer by using the suggestion frame, and carrying out category prediction and prediction frame regression;
fourth step: generating a masking layer on the feature map with the small primary size through a non-interested region masking network by utilizing a large primary size prediction frame, and inputting the feature layer filtered by the masking layer into a region suggestion network;
fifth step: sequentially identifying elements with different sizes according to the sequence from large to small by using a layering target detection method to obtain category prediction and prediction frame regression results;
sixth step: and dividing the original infrared cloud image to obtain a local infrared cloud image of each element and the adjacent area thereof according to the regression result of the element prediction frame calculated by the multi-scale characteristic pyramid module.
4. The method for detecting faults of a printed circuit board by utilizing an infrared cloud pattern recognition technology as claimed in claim 1, wherein the method comprises the following steps of: generating the list of state features comprises the steps of:
the first step: inputting the local infrared cloud image generated in the third step into a state recognition network, wherein the output result of the network is the unit volume heat generation rate of the element;
and a second step of: taking the heat generated by the unit volume of the element as the state characteristics of the element, and identifying the state characteristics of all local infrared cloud pictures;
and a third step of: and summarizing the status features to obtain a printed circuit board status feature list.
5. The method for detecting faults of a printed circuit board by utilizing an infrared cloud pattern recognition technology as claimed in claim 1, wherein the method comprises the following steps of: diagnosing a fault comprises the steps of:
the first step: calculating the proportional relation between the heat generation rate of each element unit volume on the board to be detected and the normal board by using the normal printed circuit board and the state characteristic list of the printed circuit board to be detected
And a second step of: judging the state of each element by using a threshold method, and marking as normal, supercooling or overheating;
and a third step of: according to the distribution rule of the supercooling or overheating elements, the fault elements are further positioned by combining the circuit schematic diagram.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310624217.9A CN116879708A (en) | 2023-05-30 | 2023-05-30 | Method for detecting faults of printed circuit board by utilizing infrared cloud picture recognition technology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310624217.9A CN116879708A (en) | 2023-05-30 | 2023-05-30 | Method for detecting faults of printed circuit board by utilizing infrared cloud picture recognition technology |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116879708A true CN116879708A (en) | 2023-10-13 |
Family
ID=88261120
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310624217.9A Pending CN116879708A (en) | 2023-05-30 | 2023-05-30 | Method for detecting faults of printed circuit board by utilizing infrared cloud picture recognition technology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116879708A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117150368A (en) * | 2023-10-26 | 2023-12-01 | 珠海智锐科技有限公司 | Printed circuit board fault diagnosis method, system, equipment and storage medium |
-
2023
- 2023-05-30 CN CN202310624217.9A patent/CN116879708A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117150368A (en) * | 2023-10-26 | 2023-12-01 | 珠海智锐科技有限公司 | Printed circuit board fault diagnosis method, system, equipment and storage medium |
| CN117150368B (en) * | 2023-10-26 | 2024-02-02 | 珠海智锐科技有限公司 | Printed circuit board fault diagnosis method, system, equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112734692B (en) | Defect identification method and device for power transformation equipment | |
| CN107483014B (en) | A kind of photovoltaic panel failure automatic detection method | |
| KR100830523B1 (en) | Image processing system for use with inspection systems | |
| CN112053318B (en) | Two-dimensional PCB defect real-time automatic detection and classification device based on deep learning | |
| US20010028732A1 (en) | Method and system for detecting defects on a printed circuit board | |
| Dubey et al. | Maximally stable extremal region marking-based railway track surface defect sensing | |
| CN107895362B (en) | Machine vision method for detecting quality of miniature wiring terminal | |
| CN114216915B (en) | Method for detecting and classifying class levels based on stains and defect types of circuit board | |
| CN107239742A (en) | A kind of gauge pointer scale value calculating method | |
| CN111899241A (en) | Quantitative on-line detection method and system for defects of PCB (printed Circuit Board) patches in front of furnace | |
| CN111369516A (en) | Transformer bushing heating defect detection method based on infrared image recognition | |
| CN111539927B (en) | Detection method of automobile plastic assembly fastening buckle missing detection device | |
| CN116879708A (en) | Method for detecting faults of printed circuit board by utilizing infrared cloud picture recognition technology | |
| CN110186375A (en) | Intelligent high-speed rail white body assemble welding feature detection device and detection method | |
| WO2017071406A1 (en) | Method and system for detecting pin of gold needle element | |
| JP2022027473A (en) | Generation of training data usable for inspection of semiconductor sample | |
| CN116152697A (en) | Three-dimensional model measuring method and related device for concrete structure cracks | |
| CN113705564B (en) | Pointer type instrument identification reading method | |
| CN110321869A (en) | Personnel's detection and extracting method based on Multiscale Fusion network | |
| CN117274258A (en) | Method, system, equipment and storage medium for detecting defects of main board image | |
| CN115375608A (en) | Detection method and device, detection equipment and storage medium | |
| CN108898187A (en) | A kind of method and device of automatic identification power distribution room indicating equipment image | |
| Caliskan et al. | Design and realization of an automatic optical inspection system for PCB solder joints | |
| WO2017080295A1 (en) | Element positioning method and system | |
| CN114529543B (en) | Installation detection method and device for peripheral screw gasket of aero-engine |
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 |