CN116523853A - Chip detection system and method based on deep learning - Google Patents
Chip detection system and method based on deep learning Download PDFInfo
- Publication number
- CN116523853A CN116523853A CN202310409211.XA CN202310409211A CN116523853A CN 116523853 A CN116523853 A CN 116523853A CN 202310409211 A CN202310409211 A CN 202310409211A CN 116523853 A CN116523853 A CN 116523853A
- Authority
- CN
- China
- Prior art keywords
- chip
- detection
- deep learning
- main control
- module
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 65
- 238000013135 deep learning Methods 0.000 title claims abstract description 27
- 238000000034 method Methods 0.000 title claims abstract description 22
- 238000004891 communication Methods 0.000 claims abstract description 25
- 238000012549 training Methods 0.000 claims abstract description 10
- 238000012545 processing Methods 0.000 claims abstract description 8
- 230000011218 segmentation Effects 0.000 claims abstract description 3
- 230000003044 adaptive effect Effects 0.000 claims description 6
- 238000013528 artificial neural network Methods 0.000 claims description 5
- 238000004364 calculation method Methods 0.000 claims description 5
- 230000000877 morphologic effect Effects 0.000 claims description 5
- 238000012360 testing method Methods 0.000 claims description 4
- 238000005260 corrosion Methods 0.000 claims description 3
- 230000007797 corrosion Effects 0.000 claims description 3
- 238000013500 data storage Methods 0.000 claims description 3
- 239000004973 liquid crystal related substance Substances 0.000 claims description 3
- 238000007726 management method Methods 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 3
- 238000013145 classification model Methods 0.000 claims description 2
- 230000003993 interaction Effects 0.000 claims description 2
- 238000007689 inspection Methods 0.000 claims 4
- 230000001537 neural effect Effects 0.000 abstract description 2
- 238000013480 data collection Methods 0.000 abstract 1
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000011093 chipboard Substances 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/09—Supervised learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20036—Morphological image processing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30148—Semiconductor; IC; Wafer
Abstract
The invention discloses a chip detection system based on deep learning, which comprises a main control module, a detection module and a network communication module, wherein the main control module is used for processing detection data and generating a model, the detection module is used for detecting the current state of a chip, and the network communication module is used for interacting with a cloud; the main control module collects image information of the detection model and performs feature segmentation on the image information, and the new feature information is imported into the database for training of the model. The invention also discloses a chip detection method based on deep learning, which comprises the steps of establishing a model, running an intelligent recognition algorithm in an intelligent computing unit through a main control module, judging whether faults exist or not, and making a judgment result. According to the invention, data collection is carried out on the chip in the detection process, a training set is generated, and supervised learning is carried out through a deep network neural algorithm, so that a training model for detection comparison is derived, and the detection accuracy is further improved.
Description
Technical Field
The invention belongs to the technical field of chip detection, and particularly relates to a chip detection system and method based on deep learning.
Background
Currently, chip detection technology is relatively behind, and most scenes are still to check the fault condition of the chip through the eyes of workers. The manual detection is low in efficiency and high in error rate. In the prior art, with the rapid development of artificial intelligence, the production cost is reduced due to the consideration of improving the production efficiency; the chip is detected by combining artificial intelligence with a vision sensor, so that the accuracy and the efficiency of chip detection are improved, for example, the Chinese patent application number is 202210659472.2, and the name is a chip detection method and system based on machine vision; however, in this patent document, the detection accuracy is only recognized by the whole chip, and a training model itself cannot be formed, and the detection accuracy is not high.
Disclosure of Invention
The invention aims to: the invention aims to provide a chip detection system based on deep learning, which has high detection speed and high efficiency; another object of the present invention is to provide a chip detection method based on deep learning.
The technical scheme is as follows: the chip detection system based on deep learning comprises a main control module, a detection module and a network communication module, wherein the main control module is used for processing detection data and generating a model, the detection module is used for detecting the current state of a chip, and the network communication module is used for interacting with a cloud; the main control module collects image information of the detection model and performs feature segmentation on the image information, and the new feature information is imported into the database for training of the model.
The main control module comprises a display unit and an intelligent computing unit, wherein the display unit is used for displaying the current detection state of the chip; the intelligent computing unit is used for running an image deep learning algorithm.
The network communication module comprises a wireless network communication module and a Bluetooth module, wherein the wireless network communication module is used for accessing a network and transmitting detection data to a cloud; the Bluetooth module is used for interaction with the field control device.
The display unit also comprises a liquid crystal display screen used for displaying the current detection state of the chip and the overall running state of the system.
The detection module comprises a CCD high-definition camera, and the CCD high-definition camera is used for obtaining chip image data and transmitting the chip image data back to the main control module.
The intelligent computing unit comprises at least 1 high-performance processor with a neural network processor, and the high-performance processor is used for running an image deep learning algorithm and inputting and outputting data and instructions.
A chip detection method based on deep learning at least comprises the following steps:
step 1, sending a chip to the lower part of a detection module, starting the detection module, shooting a chip picture, and sending data to a main control module;
step 2, the main control module runs an intelligent recognition algorithm in the intelligent computing unit, judges whether faults exist or not, and makes a judgment result;
step 3, the display unit of the main control module displays the judgment generated by the intelligent operation unit and gives a prompt to staff;
step 4, the network communication module sends the obtained chip state to a cloud management platform for data storage;
and 5, if the network communication module is normal, sending an instruction to enable the conveyor belt to be sent to the good product area, and if the conveyor belt has a fault, sending the conveyor belt to the fault area.
The intelligent computing unit in the step 2 judges whether the detecting chip has faults or not, and at least comprises the following steps:
step 21, obtaining a large number of image sets and labels in different states, performing supervised learning through a deep neural network algorithm, and deriving a trained model;
step 22, performing binarization and morphological image processing on the chip image acquired by the detection module, and then dividing the image and extracting the characteristics of each chip;
step 23, importing the characteristics of the chips into a pre-training model for calculation, classifying the chips, comparing the chips according to preset quantity, model and other data, and judging the chips as normal or fault;
step 24, outputting the test result to the display unit and the network communication module, and uploading the test result to the cloud end through the network communication module;
and step 25, adaptively adjusting the configuration of the super parameters according to user feedback, improving the recognition accuracy and achieving the output of the optimal classification model.
The binarization method in the step 22 adopts an adaptive parameter threshold method, and the specific calculation process meets the following conditions:
wherein L is i,j Is the brightness value of the binarized i, j pixel point, R i,j G i,j B i,j The RGB brightness values of i and j pixel points are respectively, K is an adaptive threshold value, and the K is changed along with the photo quality and the influence of ambient light.
The morphological image processing in step 22 includes preprocessing the input image, performing expansion and corrosion operations, deleting the part with low correlation with the target key information of the component, and unifying the target picture scale.
The beneficial effects are that: compared with the prior art, the invention has the following remarkable progress: according to the method based on deep learning, the camera is used for replacing human eyes to detect the chip, so that the efficiency of chip detection is improved; secondly, the invention collects data of the chip in the detection process, generates a training set, and supervises and learns through a deep network neural algorithm, thereby deriving a training model for detection comparison, and further improving the detection accuracy.
Drawings
FIG. 1 is a block diagram of the overall structure of the present invention;
FIG. 2 is a flow chart of the detection process of the present invention;
FIG. 3 is a flow chart of the intelligent computing process of the present invention.
Detailed Description
As shown in FIG. 1, the chip detection system based on deep learning in the invention comprises a main control module, a detection module and a network communication module. The main control module comprises a display unit and an intelligent computing unit. The display unit is used for displaying the current detection state of the chip; the intelligent computing unit is used for running an image deep learning algorithm. The detection module comprises a CCD high-definition camera and is used for obtaining chip image data and transmitting the chip image data back to the main control module. The network communication module comprises a wireless network communication unit and a Bluetooth unit, and can be used for accessing a network and transmitting detection data to a cloud. The display unit also comprises a chip liquid crystal display screen which is used for displaying the current detection state of the chip and the overall running state of the system. The intelligent computing unit comprises 1 or more high-performance processors with neural network processors, and is used for running an image deep learning algorithm and inputting and outputting data and instructions.
As shown in fig. 2, the intelligent detection flow includes the following steps: the chip is sent to the lower part of the detection module by using a conveyor belt, the detection module is started, a chip picture is shot, and data is sent to the main control module; the main control module runs an intelligent recognition algorithm in the intelligent computing unit, judges whether a fault condition exists, and makes a judgment; the display unit of the main control module displays the judgment generated by the intelligent operation unit and gives a prompt to staff; the network communication module sends the obtained chip state to the cloud management platform for data storage; if normal, the network communication module sends an instruction to the conveyor belt to be sent to the good product area, and if a fault exists, the conveyor belt is sent to the fault area.
As shown in fig. 3, the binarization method adopts an adaptive parameter threshold method, and the specific calculation process meets the following conditions:
wherein L is i,j Is the brightness value of the binarized i, j pixel point, R i,j G i,j B i,j The RGB brightness values of i and j pixel points are respectively, K is an adaptive threshold value, and the K is changed along with the photo quality and the influence of ambient light. Preferably, when the resolution of the photo is 1920×1080 and the detection chip board is located in the black box, only the camera is illuminated by the self-contained light source, the value of K is 504. The morphological image processing comprises preprocessing an input image, adopting expansion and corrosion operations, deleting a part with low correlation with the target key information of the component, and unifying the target picture scale.
Claims (10)
1. The chip detection system based on deep learning is characterized in that: the cloud computing system comprises a main control module, a detection module and a network communication module, wherein the main control module is used for processing detection data and generating a model, the detection module is used for detecting the current state of a chip, and the network communication module is used for interacting with a cloud; the main control module collects image information of the detection model and performs feature segmentation on the image information, and the new feature information is imported into the database for training of the model.
2. The chip detection system based on deep learning of claim 1, wherein the main control module comprises a display unit and an intelligent computing unit, and the display unit is used for displaying the current detection state of the chip; the intelligent computing unit is used for running an image deep learning algorithm.
3. The deep learning-based chip detection system of claim 1, wherein the network communication module comprises a wireless network communication module and a bluetooth module, the wireless network communication module is used for accessing a network and transmitting detection data to a cloud; the Bluetooth module is used for interaction with the field control device.
4. The deep learning-based chip inspection system of claim 1, wherein the display unit further comprises a liquid crystal display for displaying the current inspection status of the chip and the overall operation status of the system.
5. The chip detection system based on deep learning of claim 1, wherein the detection module comprises a CCD high-definition camera, and the CCD high-definition camera is configured to obtain chip image data and transmit the chip image data back to the main control module.
6. The deep learning based chip inspection system of claim 1, wherein the intelligent computing unit includes at least 1 high performance processor with a neural network processor for running image deep learning algorithms and inputting and outputting data and instructions.
7. The method for detecting a chip based on deep learning according to any one of claims 1 to 6, comprising at least the steps of:
step 1, sending a chip to the lower part of a detection module, starting the detection module, shooting a chip picture, and sending data to a main control module;
step 2, the main control module runs an intelligent recognition algorithm in the intelligent computing unit, judges whether faults exist or not, and makes a judgment result;
step 3, the display unit of the main control module displays the judgment generated by the intelligent operation unit and gives a prompt to staff;
step 4, the network communication module sends the obtained chip state to a cloud management platform for data storage;
and 5, if the network communication module is normal, sending an instruction to enable the conveyor belt to be sent to the good product area, and if the conveyor belt has a fault, sending the conveyor belt to the fault area.
8. The method for detecting a chip based on deep learning as claimed in claim 7, wherein the intelligent computing unit in step 2 determines whether the chip has a fault, and the method at least comprises the following steps:
step 21, obtaining a large number of image sets and labels in different states, performing supervised learning through a deep neural network algorithm, and deriving a trained model;
step 22, performing binarization and morphological image processing on the chip image acquired by the detection module, and then dividing the image and extracting the characteristics of each chip;
step 23, importing the characteristics of the chips into a pre-training model for calculation, classifying the chips, comparing the chips according to preset quantity, model and other data, and judging the chips as normal or fault;
step 24, outputting the test result to the display unit and the network communication module, and uploading the test result to the cloud end through the network communication module;
and step 25, adaptively adjusting the configuration of the super parameters according to user feedback, improving the recognition accuracy and achieving the output of the optimal classification model.
9. The method for chip detection based on deep learning according to claim 8, wherein the binarization method in step 22 uses an adaptive parameter threshold method, and the specific calculation process satisfies the following conditions:
wherein L is i,j Is the brightness value of the binarized i, j pixel point, R i,j G i,j B i,j The RGB brightness values of i and j pixel points are respectively, K is an adaptive threshold value, and the K is changed along with the photo quality and the influence of ambient light.
10. The method for chip inspection based on deep learning according to claim 8, wherein the morphological image processing in step 22 includes preprocessing the input image with expansion and corrosion operations, deleting the part with low correlation with the critical information of the component object, and unifying the object picture scale.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310409211.XA CN116523853A (en) | 2023-04-17 | 2023-04-17 | Chip detection system and method based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310409211.XA CN116523853A (en) | 2023-04-17 | 2023-04-17 | Chip detection system and method based on deep learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116523853A true CN116523853A (en) | 2023-08-01 |
Family
ID=87396785
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310409211.XA Pending CN116523853A (en) | 2023-04-17 | 2023-04-17 | Chip detection system and method based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116523853A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116933608A (en) * | 2023-09-15 | 2023-10-24 | 深圳市正和兴电子有限公司 | Storage chip management method, system and storage medium |
-
2023
- 2023-04-17 CN CN202310409211.XA patent/CN116523853A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116933608A (en) * | 2023-09-15 | 2023-10-24 | 深圳市正和兴电子有限公司 | Storage chip management method, system and storage medium |
| CN116933608B (en) * | 2023-09-15 | 2023-12-22 | 深圳市正和兴电子有限公司 | Storage chip management method, system and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110390691B (en) | Ore dimension measuring method based on deep learning and application system | |
| US20200349875A1 (en) | Display screen quality detection method, apparatus, electronic device and storage medium | |
| CN106875373B (en) | Mobile phone screen MURA defect detection method based on convolutional neural network pruning algorithm | |
| CN111091109B (en) | Method, system and equipment for predicting age and gender based on face image | |
| CN109886335B (en) | Classification model training method and device | |
| CN110287963B (en) | OCR recognition method for comprehensive performance test | |
| US20230119593A1 (en) | Method and apparatus for training facial feature extraction model, method and apparatus for extracting facial features, device, and storage medium | |
| CN109801265B (en) | Real-time transmission equipment foreign matter detection system based on convolutional neural network | |
| Wang et al. | Investigation into recognition algorithm of helmet violation based on YOLOv5-CBAM-DCN | |
| CN111783620A (en) | Expression recognition method, device, equipment and storage medium | |
| CN113838034B (en) | Quick detection method for surface defects of candy package based on machine vision | |
| GB2595558A (en) | Exposure defects classification of images using a neural network | |
| CN110082821A (en) | A kind of no label frame microseism signal detecting method and device | |
| CN110096945B (en) | Indoor monitoring video key frame real-time extraction method based on machine learning | |
| CN113642474A (en) | Hazardous area personnel monitoring method based on YOLOV5 | |
| CN112561879B (en) | Ambiguity evaluation model training method, image ambiguity evaluation method and image ambiguity evaluation device | |
| CN116523853A (en) | Chip detection system and method based on deep learning | |
| CN111524113A (en) | Lifting chain abnormity identification method, system, equipment and medium | |
| CN111507325A (en) | Industrial visual OCR recognition system and method based on deep learning | |
| CN113916899B (en) | Method, system and device for detecting large transfusion soft bag product based on visual identification | |
| CN114972246A (en) | Die-cutting product surface defect detection method based on deep learning | |
| CN114049676A (en) | Fatigue state detection method, device, equipment and storage medium | |
| CN110855467B (en) | Network comprehensive situation prediction method based on computer vision technology | |
| CN113642473A (en) | Mining coal machine state identification method based on computer vision | |
| CN111242047A (en) | Image processing method and apparatus, electronic device, and computer-readable storage medium |
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 |