CN116343213B - Model training and chip character recognition method, device, equipment and medium - Google Patents
Model training and chip character recognition method, device, equipment and medium Download PDFInfo
- Publication number
- CN116343213B CN116343213B CN202310633652.8A CN202310633652A CN116343213B CN 116343213 B CN116343213 B CN 116343213B CN 202310633652 A CN202310633652 A CN 202310633652A CN 116343213 B CN116343213 B CN 116343213B
- Authority
- CN
- China
- Prior art keywords
- image
- character
- images
- smd
- light source
- 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.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/14—Image acquisition
- G06V30/1431—Illumination control
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/14—Image acquisition
- G06V30/148—Segmentation of character regions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/19—Recognition using electronic means
- G06V30/191—Design or setup of recognition systems or techniques; Extraction of features in feature space; Clustering techniques; Blind source separation
- G06V30/19147—Obtaining sets of training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/19—Recognition using electronic means
- G06V30/191—Design or setup of recognition systems or techniques; Extraction of features in feature space; Clustering techniques; Blind source separation
- G06V30/1918—Fusion techniques, i.e. combining data from various sources, e.g. sensor fusion
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
The embodiment of the application discloses a method, a device, equipment and a medium for model training and chip character recognition, and relates to the field of chip character recognition. On the one hand, compared with the existing template matching method, the method for identifying the artificial intelligent model does not need to be used as a template in advance, so that the identification efficiency is improved; on the other hand, on the basis, considering that the SMD chips on the production line possibly cause uneven image quality of the shot SMD chips due to reflection of metal pins, when the model is trained, images of the SMD chips shot by different irradiation angles under the same light source and images of the SMD chips shot by different SMD chip pose under the same light source and the same irradiation angle are synthesized on a background image, so that sample enhancement is realized, and the effect of improving the accuracy of character recognition and character defect recognition of the SMD chips through model recognition is achieved.
Description
Technical Field
The present application relates to the field of chip character recognition technology, and in particular, to a method, an apparatus, a device, and a medium for model training and chip character recognition.
Background
The SMD chip refers to a chip of an SMD package, and the SMD (Surface Mount Device) package refers to a chip of a component directly soldered to a Surface of a circuit board, and pins of the component do not pass through the circuit board, but form a mesh-shaped pin on the Surface of the circuit board, and are soldered to the circuit board, and this package mode is also called Surface Mount (Surface Mount). The information formed by characters such as the manufacturer on the chip surface, the specification model parameters of the product and the like is very important for users, and in the existing method, the recognition efficiency is lower by adopting a template matching method for recognizing the character defects on the SMD chip.
It should be noted that the information disclosed in the above background section is only for enhancing understanding of the background of the application and thus may include information that does not form the prior art that is already known to those of ordinary skill in the art.
Disclosure of Invention
The embodiment of the application provides a method, a device, equipment and a medium for model training and chip character recognition, which solve the technical problem of low character defect recognition efficiency aiming at an SMD chip in the existing method.
In one aspect, an embodiment of the present application provides a method for recognizing characters of a chip, which is used for a computer device of an SMD chip production line, and includes:
Receiving a target image acquired on an SMD chip production line, wherein the target image comprises characters of a target SMD chip;
inputting the target image into a trained character recognition model to obtain a character recognition result of the target SMD chip; wherein the character recognition result comprises target character information and target character defect information; the training image of the character recognition model comprises an image synthesized by a background image and a plurality of character images, wherein the background image and the plurality of character images are obtained from a first image and a second image, the first image is an image of an SMD chip shot at different irradiation angles under the same light source, and the second image is an image of an SMD chip shot at different SMD chip poses under the same light source and the same irradiation angle; the character recognition model comprises character defect information;
and outputting the character recognition result.
Optionally, a light source for illuminating the SMD chip is arranged on the SMD chip production line, a vibration starting sheet is arranged at the position of the light source, and a polarizing sheet is arranged in front of an industrial camera lens for collecting images;
the step of receiving the target image acquired on the SMD chip production line further comprises:
Controlling the light source provided with the vibration starting sheet to irradiate a target SMD chip on the SMD chip production line;
and controlling the industrial camera lens provided with the polaroid to shoot the target SMD chip under the irradiation of the light source so as to obtain the target image.
It can be understood that, in the recognition process, the contrast between the character and the background is improved as much as possible during image acquisition, which is one of means for further improving the recognition accuracy. Although the sample image enhancement has been performed in the foregoing embodiment, on the basis of this, the quality of the target image is improved as much as possible when the target image is acquired, and the recognition accuracy is also improved from another point of view.
Optionally, before the step of receiving the target image acquired on the SMD chip production line, the method further includes:
acquiring an original sample set and an enhanced sample set; the enhanced sample set comprises a plurality of constructed sample images, wherein the constructed sample images are synthesized by the background images and the character images; the original sample set comprises a plurality of sample images, and the sample images and the constructed sample images comprise character defect information;
Training the constructed initial character recognition model by utilizing the original sample set and the enhanced sample set to obtain the character recognition model.
It can be appreciated that performing model recognition in advance can improve the efficiency of subsequent character recognition.
Optionally, before the step of obtaining the original sample set and the enhanced sample set, the method further includes:
acquiring first images of SMD chips shot at different irradiation angles under the same light source;
acquiring second images of the SMD chips shot by different SMD chip pose under the same light source and the same irradiation angle;
dividing the first image and the second image by using a vertical projection method to obtain a plurality of background images and a plurality of character images;
and fusing the background image and the character image by using a poisson fusion method to obtain the construction sample image.
It can be understood that when image shooting is performed on the SMD chip production line, an industrial light source is required to irradiate, the light source is generally arranged right above the chip, and the distance between the light source and the SMD chip can be adjusted through the threaded screw rod, so that a photo with clear characters is expected to be shot on the basis of clearly illuminating the SMD chip. However, because the metal pins of the SMD chip reflect light, and the pose of the SMD chip is different, or the surface shapes of the pins are slightly different, the light reflection effect and direction can be greatly different, and the shot image quality of the SMD chip may be uneven. Therefore, in this embodiment, by combining the images of the SMD chips photographed at different illumination angles under the same light source and the images of the SMD chips photographed at different positions and orientations of the SMD chips under the same light source, different reflective effects that may occur due to the deviation of the light source angle, the positions and orientations of the SMD chips, and the differences of the pins are fully simulated, so that the generalization capability of the model to character recognition of the SMD chips is improved, and the recognition accuracy is improved. Further, the character image is pasted onto the background image using the poisson fusion method to create a construction sample image. The poisson fusion method is a method of synthesizing an image into a background, and can simultaneously preserve the color and brightness characteristics of the image. Therefore, in this embodiment, the character image is pasted onto the background image by using the poisson fusion method, so that the original characteristics can be maintained, and the quality of the constructed sample image is ensured, thereby ensuring the training quality of the model.
Optionally, the step of fusing the background image and the character image by using a poisson fusion method to obtain the constructed sample image includes:
randomly screening N character images and a background image from a plurality of character images and a plurality of background images, wherein N is an integer greater than or equal to 2;
and synthesizing the N character images and one background image by using a poisson fusion method to obtain the constructed sample image.
It can be appreciated that multiple character images including different characters and/or different character defects can be pasted in one background image, so that model training efficiency and model generalization capability can be improved. In addition, under the condition that the types of character images and the types of character defects are large, the generalization capability of the character recognition model can be remarkably improved by adopting a random screening mode.
Optionally, after the step of outputting the character recognition result, the method further includes:
and if the character recognition result shows that the character defect exists, eliminating the target SMD chip.
It can be understood that when the real-time identification is performed on the SMD chip production line, if the character defect exists, the target SMD chip is removed by the vertical horse, so that the yield of products on the production line can be ensured.
In still another aspect, an embodiment of the present application provides a model training method, including:
acquiring an original sample set and an enhanced sample set, wherein the enhanced sample set comprises a plurality of constructed sample images, the constructed sample images are synthesized by a background image and a plurality of character images, the background image and the plurality of character images are obtained from a first image and a second image, the first image is an image of an SMD chip shot at different irradiation angles under the same light source, and the second image is an image of an SMD chip shot at different SMD chip poses under the same light source and the same irradiation angle; the original sample set comprises a plurality of sample images, and the sample images and the constructed sample images comprise character defect information;
training the constructed initial character recognition model by utilizing the original sample set and the enhanced sample set to obtain the character recognition model.
Optionally, before the step of obtaining the original sample set and the enhanced sample set, the method further includes:
acquiring first images of SMD chips shot at different irradiation angles under the same light source;
acquiring second images of the SMD chips shot by different SMD chip pose under the same light source and the same irradiation angle;
Dividing the first image and the second image by using a vertical projection method to obtain a plurality of background images and a plurality of character images;
and fusing the background image and the character image by using a poisson fusion method to obtain the construction sample image.
It can be understood that when image shooting is performed on the SMD chip production line, an industrial light source is required to irradiate, the light source is generally arranged right above the chip, and the distance between the light source and the SMD chip can be adjusted through the threaded screw rod, so that a photo with clear characters is expected to be shot on the basis of clearly illuminating the SMD chip. However, because the metal pins of the SMD chip reflect light, and the pose of the SMD chip is different, or the surface shapes of the pins are slightly different, the light reflection effect and direction can be greatly different, and the shot image quality of the SMD chip may be uneven. Therefore, in this embodiment, by combining the images of the SMD chips photographed at different illumination angles under the same light source and the images of the SMD chips photographed at different positions and orientations of the SMD chips under the same light source, different reflective effects that may occur due to the deviation of the light source angle, the positions and orientations of the SMD chips, and the differences of the pins are fully simulated, so that the generalization capability of the model to character recognition of the SMD chips is improved, and the recognition accuracy is improved. Further, the character image is pasted onto the background image using the poisson fusion method to create a construction sample image. The poisson fusion method is a method of synthesizing an image into a background, and can simultaneously preserve the color and brightness characteristics of the image. Therefore, in this embodiment, the character image is pasted onto the background image by using the poisson fusion method, so that the original characteristics can be maintained, and the quality of the constructed sample image is ensured, thereby ensuring the training quality of the model.
Optionally, the step of fusing the background image and the character image by using a poisson fusion method to obtain the constructed sample image includes:
randomly screening N character images and a background image from a plurality of character images and a plurality of background images, wherein N is an integer greater than or equal to 2;
and synthesizing the N character images and one background image by using a poisson fusion method to obtain the constructed sample image.
It can be appreciated that multiple character images including different characters and/or different character defects can be pasted in one background image, so that model training efficiency and model generalization capability can be improved. In addition, under the condition that the types of character images and the types of character defects are large, the generalization capability of the character recognition model can be remarkably improved by adopting a random screening mode.
In still another aspect, an embodiment of the present application provides a character recognition apparatus for a chip, including:
the receiving module is used for receiving a target image acquired on the SMD chip production line, wherein the target image comprises characters of a target SMD chip;
the recognition module is used for inputting the target image into a trained character recognition model so as to obtain a character recognition result of the target SMD chip; wherein the character recognition result comprises target character information and target character defect information; the training image of the character recognition model comprises an image synthesized by a background image and a plurality of character images, wherein the background image and the plurality of character images are obtained from a first image and a second image, the first image is an image of an SMD chip shot at different irradiation angles under the same light source, and the second image is an image of an SMD chip shot at different SMD chip poses under the same light source and the same irradiation angle; the character recognition model comprises character defect information;
And the output module is used for outputting the character recognition result.
In still another aspect, an embodiment of the present application provides a model training apparatus, including:
the acquisition module is used for acquiring an original sample set and an enhanced sample set, wherein the enhanced sample set comprises a plurality of constructed sample images, the constructed sample images are synthesized by background images and a plurality of character images, the background images and the plurality of character images are obtained from a first image and a second image, the first image is an image of an SMD chip shot at different irradiation angles under the same light source, and the second image is an image of an SMD chip shot at different SMD chip pose under the same light source and the same irradiation angle; the original sample set comprises a plurality of sample images, and the sample images and the constructed sample images comprise character defect information;
and the training module is used for training the constructed initial character recognition model by utilizing the original sample set and the enhanced sample set so as to obtain the character recognition model.
In yet another aspect, an embodiment of the present application provides a computer apparatus, including: the device comprises a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to realize the method.
In yet another aspect, an embodiment of the present application provides a computer readable storage medium, where a computer program is stored, and the processor executes the computer program to implement the foregoing method.
The embodiment of the application provides a method, a device, equipment and a medium for model training and chip character recognition, wherein the method is used for computer equipment of an SMD chip production line and comprises the following steps: receiving a target image acquired on an SMD chip production line, wherein the target image comprises characters of a target SMD chip; inputting the target image into a trained character recognition model to obtain a character recognition result of the target SMD chip; wherein the character recognition result comprises target character information and target character defect information; the training image of the character recognition model comprises an image synthesized by a background image and a plurality of character images, wherein the background image and the plurality of character images are obtained from a first image and a second image, the first image is an image of an SMD chip shot at different irradiation angles under the same light source, and the second image is an image of an SMD chip shot at different SMD chip poses under the same light source and the same irradiation angle; the character recognition model comprises character defect information; and outputting the character recognition result. That is, on the one hand, compared with the existing template matching method, the method for identifying the artificial intelligent model does not need to be used as a template in advance, so that the identification efficiency is improved; on the other hand, on the basis, considering that the SMD chips on the production line possibly cause uneven image quality of the shot SMD chips due to reflection of metal pins, when the model is trained, images of the SMD chips shot by different irradiation angles under the same light source and images of the SMD chips shot by different SMD chip pose under the same light source and the same irradiation angle are synthesized on a background image, so that sample enhancement is realized, and the effect of improving the accuracy of character recognition and character defect recognition of the SMD chips through model recognition is achieved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for the description of the embodiments will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present application, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic diagram of a computer device according to an embodiment of the present application;
FIG. 2 is a schematic flow chart of a method for recognizing characters of a chip according to an embodiment of the present application;
FIG. 3 is a flowchart of another method for recognizing characters of a chip according to an embodiment of the present application;
FIG. 4 is a schematic flow chart of a model training method according to an embodiment of the present application;
fig. 5 is a schematic structural diagram of a character recognition device of a chip according to an embodiment of the present application;
fig. 6 is a schematic structural diagram of a model training device according to an embodiment of the present application.
Detailed Description
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 application.
The main solutions of the embodiments of the present application are: provided are a method, apparatus, device and medium for model training and chip character recognition, the method is used for computer equipment of SMD chip production line, and comprises: receiving a target image acquired on an SMD chip production line, wherein the target image comprises characters of a target SMD chip; inputting the target image into a trained character recognition model to obtain a character recognition result of the target SMD chip; wherein the character recognition result comprises target character information and target character defect information; the training image of the character recognition model comprises an image synthesized by a background image and a plurality of character images, wherein the background image and the plurality of character images are obtained from a first image and a second image, the first image is an image of an SMD chip shot at different irradiation angles under the same light source, and the second image is an image of an SMD chip shot at different SMD chip poses under the same light source and the same irradiation angle; the character recognition model comprises character defect information; and outputting the character recognition result.
In the prior art, SMD (Surface Mount Device) packages refer to soldering a chip of a component directly to a surface of a circuit board, and pins of the component do not pass through the circuit board, but form a mesh-shaped pin on the surface of the circuit board. In addition, in the existing character recognition, it is necessary to perform character segmentation first, and then perform gray-scale-based template matching or edge-feature-based template matching for the segmented characters. The template matching method needs to make templates in advance in a targeted manner.
Therefore, the application provides a solution, on one hand, compared with the existing template matching method, the method for identifying the artificial intelligent model does not need to make templates in advance, so that the identification efficiency is improved; on the other hand, on the basis, considering that the SMD chips on the production line possibly cause uneven image quality of the shot SMD chips due to reflection of metal pins, when the model is trained, images of the SMD chips shot by different irradiation angles under the same light source and images of the SMD chips shot by different SMD chip pose under the same light source and the same irradiation angle are synthesized on a background image, so that sample enhancement is realized, and the effect of improving the accuracy of character recognition and character defect recognition of the SMD chips through model recognition is achieved.
Referring to fig. 1, fig. 1 is a schematic diagram of a computer device structure of a hardware running environment according to an embodiment of the present application.
As shown in fig. 1, the computer device may include: a processor 1001, such as a central processing unit (Central Processing Unit, CPU), a communication bus 1002, a user interface 1003, a network interface 1004, a memory 1005. Wherein the communication bus 1002 is used to enable connected communication between these components. The user interface 1003 may include a Display, an input unit such as a Keyboard (Keyboard), and the optional user interface 1003 may further include a standard wired interface, a wireless interface. The network interface 1004 may optionally include a standard wired interface, a WIreless interface (e.g., a WIreless-FIdelity (WI-FI) interface). The Memory 1005 may be a high-speed random access Memory (Random Access Memory, RAM) Memory or a stable nonvolatile Memory (NVM), such as a disk Memory. The memory 1005 may also optionally be a storage device separate from the processor 1001 described above.
Those skilled in the art will appreciate that the architecture shown in fig. 1 is not limiting of a computer device and may include more or fewer components than shown, or may combine certain components, or a different arrangement of components.
As shown in fig. 1, the memory 1005, which is a storage medium, may include an operating system, a network communication module, a user interface module, and a character recognition device or a model training device of a chip.
In the computer device shown in fig. 1, the network interface 1004 is mainly used for data communication with a network server; the user interface 1003 is mainly used for data interaction with a user; the processor 1001 and the memory 1005 in the computer device of the present application may be provided in the computer device, and the computer device calls the character recognition device or the model training device of the chip stored in the memory 1005 through the processor 1001, and executes the character recognition method or the model training method of the chip provided by the embodiment of the present application.
Referring to fig. 2, an embodiment of the present application provides a method for recognizing characters of a chip, which is used in the computer device of the foregoing embodiment, and the computer device may be a computer device used on an SMD chip production line, and is electrically connected to an image acquisition device on the SMD chip production line, and receives an image of the SMD chip on the SMD chip production line acquired by the image acquisition device, so as to recognize characters of the chip through the image. The image acquisition device may be an industrial camera.
The method comprises the following steps:
s20, receiving a target image acquired on an SMD chip production line, wherein the target image comprises characters of a target SMD chip;
in the implementation process, the SMD chip production line refers to a production line for producing or detecting SMD chips. The target image refers to a front image on the target SMD chip, which is acquired on the SMD chip production line using an industrial camera, and thus, includes characters of the target SMD chip.
It will be appreciated that these characters may include a date of manufacture, a product model number, a number, etc., and thus, the characters may include numbers, letters, symbols, etc., in particular.
S40, inputting the target image into a trained character recognition model to obtain a character recognition result of the target SMD chip; wherein the character recognition result comprises target character information and target character defect information; the training image of the character recognition model comprises an image synthesized by a background image and a plurality of character images, wherein the background image and the plurality of character images are obtained from a first image and a second image, the first image is an image of an SMD chip shot at different irradiation angles under the same light source, and the second image is an image of an SMD chip shot at different SMD chip poses under the same light source and the same irradiation angle; the character recognition model comprises character defect information;
In the implementation process, the target character information refers to information of the recognized character itself, and the target character defect information refers to defect information of the recognized character, for example, character smudging, missing printing, pinholes, and the like. In particular, both the character and the character defect may be in the form of a label in the image.
The training image of the character recognition model not only comprises an originally collected sample image, but also comprises an image synthesized by a background image and a plurality of character images, and the sample is enhanced by synthesizing the images of the SMD chips shot at different irradiation angles under the same light source and the images of the SMD chips shot at different SMD chip pose under the same light source and the same irradiation angle, so that the effect of improving the accuracy of character recognition and character defect recognition of the SMD chips through model recognition is achieved.
Specifically, it can be understood that when the image capturing is performed on the SMD chip production line, the industrial light source is required to be irradiated, the light source is generally arranged right above the chip, and the distance between the light source and the SMD chip can be adjusted through the threaded screw rod, so that on the basis of clearly illuminating the SMD chip, a photo with clear characters is expected to be captured. However, because the metal pins of the SMD chip reflect light, and the pose of the SMD chip is different, or the surface shapes of the pins are slightly different, the light reflection effect and direction can be greatly different, and the shot image quality of the SMD chip may be uneven. Therefore, in this embodiment, by combining the images of the SMD chips photographed at different illumination angles under the same light source and the images of the SMD chips photographed at different positions and orientations of the SMD chips under the same light source, different reflective effects that may occur due to the deviation of the light source angle, the positions and orientations of the SMD chips, and the differences of the pins are fully simulated, so that the generalization capability of the model to character recognition of the SMD chips is improved, and the recognition accuracy is improved.
In addition, it should be noted that in the existing template matching process, in order to enable the subsequent template matching, pose correction of the SMD chip is performed before character segmentation is performed. By the sample enhancement method in the embodiment, the image of the SMD chip shot under various conditions can be identified without correcting the pose of the SMD chip, so that the identification efficiency is further improved.
S60, outputting the character recognition result;
in the implementation process, the mode of outputting the character recognition result can be that the character recognition result is displayed through a display screen, or that the related character and the coordinate data of the defect are output, and the like.
Therefore, according to the character recognition method of the SMD chip, on one hand, compared with the existing template matching method, the character recognition method of the SMD chip does not need to be used for making templates in advance or correcting the pose by using the artificial intelligent model recognition method, and therefore recognition efficiency is improved; on the other hand, on the basis, considering that the SMD chips on the production line possibly cause uneven image quality of the shot SMD chips due to reflection of metal pins, when the model is trained, images of the SMD chips shot by different irradiation angles under the same light source and images of the SMD chips shot by different SMD chip pose under the same light source and the same irradiation angle are synthesized on a background image, so that sample enhancement is realized, and the effect of improving the accuracy of character recognition and character defect recognition of the SMD chips through model recognition is achieved.
As an alternative embodiment, the SMD chip production line is provided with a light source for illuminating the SMD chip, the light source is provided with a vibration starting sheet, and a polarizing sheet is arranged in front of an industrial camera lens for collecting images; the step of receiving the target image acquired on the SMD chip production line further comprises:
controlling the light source provided with the vibration starting sheet to irradiate a target SMD chip on the SMD chip production line;
and controlling the industrial camera lens provided with the polaroid to shoot the target SMD chip under the irradiation of the light source so as to obtain the target image.
In the specific implementation process, the polarizing plate and the polarizing plate are common knowledge in the optical field, and are not described in detail herein.
A diffuse reflection bar-shaped combined light source may be used in this embodiment. In order to improve the contrast between the character and the background and reduce the irradiation and radiation intensity, a polarizing plate and a polarizing plate are adopted. The vibration starting sheet can change light emitted by the light source into polarized light and is placed at the position of the light source; the polarizing plate is arranged in front of the industrial camera lens, certain shielding is carried out on polarized light, the polarized light becomes natural light after diffuse reflection, and the polarized light is still polarized light after mirror surface emission. The lighting mode adopts a front bright field. According to the limitation of the station conditions of the test handler, the light source is arranged right above the SMD chip, and the distance between the light source and the SMD chip is adjusted through the threaded screw rod.
It can be understood that, in the recognition process, the contrast between the character and the background is improved as much as possible during image acquisition, which is one of means for further improving the recognition accuracy. Although the sample image enhancement has been performed in the foregoing embodiment, on the basis of this, the quality of the target image is improved as much as possible when the target image is acquired, and the recognition accuracy is also improved from another point of view.
As an alternative embodiment, referring to fig. 3, before the step of receiving the target image acquired on the SMD chip production line, the method further includes:
s102, acquiring an original sample set and an enhanced sample set;
in a specific implementation process, the enhanced sample set comprises a plurality of constructed sample images, the constructed sample images are synthesized by a background image and a plurality of character images, the background image and the plurality of character images are obtained from a first image and a second image, the first image is an image of an SMD chip shot at different irradiation angles under the same light source, and the second image is an image of an SMD chip shot at different SMD chip poses under the same light source and the same irradiation angle.
In particular, the acquired original sample image may be segmented into accurate single character images using a vertical projection method. This step includes extracting the single character image using a vertical projection method. The vertical projection is a method for calculating the pixel number of each column of the image to detect the boundary of the adjacent character, and the specific implementation method is not described herein. The original sample image may be a sample image in the original sample set, or may be a sample image in which the contrast between the character and the background is improved by the polarizing plate and the polarizing plate.
The original sample set comprises a plurality of sample images, and the sample images and the constructed sample images comprise character defect information. The sample image may be an image in which the contrast between the character and the background is improved by the polarizing plate and the polarizing plate, or an image in which the contrast between the character and the background is not improved by the polarizing plate and the polarizing plate. The character defect information is defect labeling information.
S104, training the constructed initial character recognition model by utilizing the original sample set and the enhanced sample set to obtain the character recognition model;
in an implementation, the initial character recognition model may be a neural network model, such as a neural network model of OCR. And training through the original sample set and the sample images in the enhanced sample set to obtain a character recognition model.
As an alternative embodiment, before the step of obtaining the original sample set and the enhanced sample set, the method further comprises:
acquiring first images of SMD chips shot at different irradiation angles under the same light source;
acquiring second images of the SMD chips shot by different SMD chip pose under the same light source and the same irradiation angle;
dividing the first image and the second image by using a vertical projection method to obtain a plurality of background images and a plurality of character images;
And fusing the background image and the character image by using a poisson fusion method to obtain the construction sample image.
In a specific implementation process, the first image and the second image are both collected sample images related to the SMD chip, and may be collected based on the same SMD chip or may be collected based on different SMD chips. When image shooting is carried out on an SMD chip production line, all industrial light sources are required to irradiate, the light sources are generally arranged right above the chips, and the distance between the light sources and the SMD chips can be adjusted through a threaded screw rod, so that on the basis of clearly illuminating the SMD chips, a photo with clear characters is expected to be shot. However, because the metal pins of the SMD chip reflect light, and the pose of the SMD chip is different, or the surface shapes of the pins are slightly different, the light reflection effect and direction can be greatly different, and the shot image quality of the SMD chip may be uneven. Therefore, in this embodiment, by combining the images of the SMD chips photographed at different illumination angles under the same light source and the images of the SMD chips photographed at different positions and orientations of the SMD chips under the same light source, different reflective effects that may occur due to the deviation of the light source angle, the positions and orientations of the SMD chips, and the differences of the pins are fully simulated, so that the generalization capability of the model to character recognition of the SMD chips is improved, and the recognition accuracy is improved.
Specifically, the character image is pasted onto the background image using the poisson fusion method to create a construction sample image. The poisson fusion method is a method of synthesizing an image into a background, and can simultaneously preserve the color and brightness characteristics of the image. Therefore, in this embodiment, the character image is pasted onto the background image by using the poisson fusion method, so that the original characteristics can be maintained, and the quality of the constructed sample image is ensured, thereby ensuring the training quality of the model.
As an optional implementation manner, the step of fusing the background image and the character image by using a poisson fusion method to obtain the constructed sample image includes:
randomly screening N character images and a background image from a plurality of character images and a plurality of background images, wherein N is an integer greater than or equal to 2;
and synthesizing the N character images and one background image by using a poisson fusion method to obtain the constructed sample image.
In the implementation process, a plurality of character images comprising different characters and/or different character defects can be pasted in one background image, so that the model training efficiency and the generalization capability of the model can be improved. In addition, under the condition that the types of character images and the types of character defects are large, the generalization capability of the character recognition model can be remarkably improved by adopting a random screening mode.
As an optional implementation manner, after the step of outputting the character recognition result, the method further includes:
and if the character recognition result shows that the character defect exists, eliminating the target SMD chip.
In the specific implementation process, when the real-time identification is carried out on the SMD chip production line, if the character defects exist, the target SMD chip is removed by the vertical horse, so that the yield of products on the production line can be ensured.
Based on the same inventive concept, referring to fig. 4, an embodiment of the present application further provides a model training method, including:
s200, acquiring an original sample set and an enhanced sample set, wherein the enhanced sample set comprises a plurality of constructed sample images, the constructed sample images are synthesized by a background image and a plurality of character images, the background image and the plurality of character images are obtained from a first image and a second image, the first image is an image of an SMD chip shot at different irradiation angles under the same light source, and the second image is an image of an SMD chip shot at different SMD chip poses under the same light source and the same irradiation angle; the original sample set comprises a plurality of sample images, and the sample images and the constructed sample images comprise character defect information;
S400, training the constructed initial character recognition model by utilizing the original sample set and the enhanced sample set to obtain the character recognition model.
As an alternative embodiment, before the step of obtaining the original sample set and the enhanced sample set, the method further includes:
acquiring first images of SMD chips shot at different irradiation angles under the same light source;
acquiring second images of the SMD chips shot by different SMD chip pose under the same light source and the same irradiation angle;
dividing the first image and the second image by using a vertical projection method to obtain a plurality of background images and a plurality of character images;
and fusing the background image and the character image by using a poisson fusion method to obtain the construction sample image.
As an alternative embodiment, the step of fusing the background image and the character image by using a poisson fusion method to obtain the constructed sample image includes:
randomly screening N character images and a background image from a plurality of character images and a plurality of background images, wherein N is an integer greater than or equal to 2;
and synthesizing the N character images and one background image by using a poisson fusion method to obtain the constructed sample image.
It can be understood that the essential content of the model training method in this embodiment is consistent with the content of the description about model training in the foregoing embodiment, and the specific implementation manner and the achieved technical effect of the model training method in this embodiment may be the same as those in the foregoing embodiment, and will not be repeated here.
Referring to fig. 5, based on the same inventive concept, an embodiment of the present application further provides a character recognition device of a chip, including:
the receiving module is used for receiving a target image acquired on the SMD chip production line, wherein the target image comprises characters of a target SMD chip;
the recognition module is used for inputting the target image into a trained character recognition model so as to obtain a character recognition result of the target SMD chip; wherein the character recognition result comprises target character information and target character defect information; the training image of the character recognition model comprises an image synthesized by a background image and a plurality of character images, wherein the background image and the plurality of character images are obtained from a first image and a second image, the first image is an image of an SMD chip shot at different irradiation angles under the same light source, and the second image is an image of an SMD chip shot at different SMD chip poses under the same light source and the same irradiation angle; the character recognition model comprises character defect information;
And the output module is used for outputting the character recognition result.
It should be noted that, each module in the character recognition device of the chip in this embodiment corresponds to each step in the character recognition method of the chip in the foregoing embodiment one by one, so the specific implementation manner and the achieved technical effect of this embodiment may refer to the implementation manner of the character recognition method of the foregoing chip, and will not be described herein again.
Referring to fig. 6, based on the same inventive concept, an embodiment of the present application further provides a model training apparatus, including:
the acquisition module is used for acquiring an original sample set and an enhanced sample set, wherein the enhanced sample set comprises a plurality of constructed sample images, the constructed sample images are synthesized by background images and a plurality of character images, the background images and the plurality of character images are obtained from a first image and a second image, the first image is an image of an SMD chip shot at different irradiation angles under the same light source, and the second image is an image of an SMD chip shot at different SMD chip pose under the same light source and the same irradiation angle; the original sample set comprises a plurality of sample images, and the sample images and the constructed sample images comprise character defect information;
And the training module is used for training the constructed initial character recognition model by utilizing the original sample set and the enhanced sample set so as to obtain the character recognition model.
It should be noted that, each module in the model training apparatus in this embodiment corresponds to each step in the model training method in the foregoing embodiment one by one, so specific implementation manner and achieved technical effects of this embodiment may refer to implementation manner of the foregoing model training method, and will not be described herein in detail.
Furthermore, in an embodiment, the present application also provides a computer storage medium, on which a computer program is stored, which computer program, when being executed by a processor, implements the steps of the method in the previous embodiment.
In some embodiments, the computer readable storage medium may be FRAM, ROM, PROM, EPROM, EEPROM, flash memory, magnetic surface memory, optical disk, or CD-ROM; but may be a variety of devices including one or any combination of the above memories. The computer may be a variety of computing devices including smart terminals and servers.
In some embodiments, the executable instructions may be in the form of programs, software modules, scripts, or code, written in any form of programming language (including compiled or interpreted languages, or declarative or procedural languages), and they may be deployed in any form, including as stand-alone programs or as modules, components, subroutines, or other units suitable for use in a computing environment.
As an example, the executable instructions may, but need not, correspond to files in a file system, may be stored as part of a file that holds other programs or data, for example, in one or more scripts in a hypertext markup language (HTML, hyper Text Markup Language) document, in a single file dedicated to the program in question, or in multiple coordinated files (e.g., files that store one or more modules, sub-programs, or portions of code).
As an example, executable instructions may be deployed to be executed on one computing device or on multiple computing devices located at one site or, alternatively, distributed across multiple sites and interconnected by a communication network.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or system. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or system that comprises the element.
The foregoing embodiment numbers of the present application are merely for the purpose of description, and do not represent the advantages or disadvantages of the embodiments.
From the above description of embodiments, it will be clear to a person skilled in the art that the above embodiment method may be implemented by means of software plus a necessary general hardware platform, but may of course also be implemented by means of hardware, but in many cases the former is a preferred embodiment. Based on this understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art in the form of a software product stored in a storage medium (e.g. read-only memory/random-access memory, magnetic disk, optical disk) comprising several instructions for causing a multimedia terminal device (which may be a mobile phone, a computer, a television receiver, or a network device, etc.) to perform the method according to the embodiments of the present application.
The foregoing disclosure is merely illustrative of some embodiments of the present application and it is not to be construed as limiting the scope of the application, as a person of ordinary skill in the art will appreciate that all or part of the above-described embodiments may be practiced with equivalent variations which fall within the scope of the application as defined in the appended claims.
Claims (12)
1. The character recognition method of the chip is characterized by comprising computer equipment used for an SMD chip production line, wherein a light source used for irradiating the SMD chip is arranged on the SMD chip production line, a vibration starting sheet is arranged at the position of the light source, and a polarizing sheet is arranged in front of an industrial camera lens used for collecting images; the method comprises the following steps:
controlling the light source provided with the vibration starting sheet to irradiate a target SMD chip on the SMD chip production line;
controlling the industrial camera lens provided with the polaroid to shoot the target SMD chip irradiated by a light source so as to obtain the target image;
receiving a target image acquired on an SMD chip production line, wherein the target image comprises characters of a target SMD chip;
inputting the target image into a trained character recognition model to obtain a character recognition result of the target SMD chip; wherein the character recognition result comprises target character information and target character defect information; the training image of the character recognition model comprises an image synthesized by a background image and a plurality of character images, wherein the background image and the plurality of character images are obtained from a first image and a second image, the first image is an image of an SMD chip shot at different irradiation angles under the same light source, and the second image is an image of an SMD chip shot at different SMD chip poses under the same light source and the same irradiation angle; the character recognition model comprises character defect information, the character defect information comprises character dirt, missing printing and pinholes, and the lighting mode of the first image and the second image during shooting is a front bright field;
And outputting the character recognition result.
2. The method of claim 1, wherein prior to the step of receiving the target image acquired on the SMD chip production line, further comprising:
acquiring an original sample set and an enhanced sample set; wherein the enhanced sample set comprises a plurality of constructed sample images synthesized from the background image and a plurality of the character images; the original sample set comprises a plurality of sample images, and the sample images and the constructed sample images comprise character defect information;
training the constructed initial character recognition model by utilizing the original sample set and the enhanced sample set to obtain the character recognition model.
3. The method of claim 2, wherein prior to the steps of obtaining the original sample set and enhancing the sample set, further comprising:
acquiring first images of SMD chips shot at different irradiation angles under the same light source;
acquiring second images of the SMD chips shot by different SMD chip pose under the same light source and the same irradiation angle;
dividing the first image and the second image by using a vertical projection method to obtain a plurality of background images and a plurality of character images;
And fusing the background image and the character image by using a poisson fusion method to obtain the construction sample image.
4. A method according to claim 3, wherein the step of fusing the background image and the character image using a poisson fusion method to obtain the constructed sample image comprises:
randomly screening N character images and a background image from a plurality of character images and a plurality of background images, wherein N is an integer greater than or equal to 2;
and synthesizing the N character images and one background image by using a poisson fusion method to obtain the constructed sample image.
5. The method of claim 1, wherein after the step of outputting the character recognition result, further comprising:
and if the character recognition result shows that the character defect exists, eliminating the target SMD chip.
6. A method of model training, comprising:
acquiring an original sample set and an enhanced sample set, wherein the enhanced sample set comprises a plurality of constructed sample images, the constructed sample images are synthesized by a background image and a plurality of character images, the background image and the plurality of character images are obtained from a first image and a second image, the first image is an image of an SMD chip shot at different irradiation angles under the same light source, and the second image is an image of an SMD chip shot at different SMD chip poses under the same light source and the same irradiation angle; the original sample set comprises a plurality of sample images, the sample images and the constructed sample images comprise character defect information, the character defect information comprises character dirt, printing omission and pinholes, and the lighting mode of the first image and the second image during shooting is a front bright field;
Training the constructed initial character recognition model by utilizing the original sample set and the enhanced sample set to obtain a character recognition model.
7. The method of claim 6, wherein prior to the steps of obtaining the original sample set and enhancing the sample set, further comprising:
acquiring first images of SMD chips shot at different irradiation angles under the same light source;
acquiring second images of the SMD chips shot by different SMD chip pose under the same light source and the same irradiation angle;
dividing the first image and the second image by using a vertical projection method to obtain a plurality of background images and a plurality of character images;
and fusing the background image and the character image by using a poisson fusion method to obtain the construction sample image.
8. The method of claim 7, wherein the step of fusing the background image and the character image using a poisson fusion method to obtain the constructed sample image comprises:
randomly screening N character images and a background image from a plurality of character images and a plurality of background images, wherein N is an integer greater than or equal to 2;
And synthesizing the N character images and one background image by using a poisson fusion method to obtain the constructed sample image.
9. The character recognition device of the chip is characterized by comprising computer equipment used for an SMD chip production line, wherein a light source used for irradiating the SMD chip is arranged on the SMD chip production line, a vibration starting sheet is arranged at the position of the light source, and a polarizing sheet is arranged in front of an industrial camera lens used for collecting images; comprising the following steps:
the receiving module is used for controlling the light source provided with the vibration starting sheet to irradiate a target SMD chip on the SMD chip production line; controlling the industrial camera lens provided with the polaroid to shoot the target SMD chip irradiated by a light source so as to obtain the target image; receiving a target image acquired on an SMD chip production line, wherein the target image comprises characters of a target SMD chip;
the recognition module is used for inputting the target image into a trained character recognition model so as to obtain a character recognition result of the target SMD chip; wherein the character recognition result comprises target character information and target character defect information; the training image of the character recognition model comprises an image synthesized by a background image and a plurality of character images, wherein the background image and the plurality of character images are obtained from a first image and a second image, the first image is an image of an SMD chip shot at different irradiation angles under the same light source, and the second image is an image of an SMD chip shot at different SMD chip poses under the same light source and the same irradiation angle; the character recognition model comprises character defect information, the character defect information comprises character dirt, missing printing and pinholes, and the lighting mode of the first image and the second image during shooting is a front bright field;
And the output module is used for outputting the character recognition result.
10. A model training device, comprising:
the acquisition module is used for acquiring an original sample set and an enhanced sample set, wherein the enhanced sample set comprises a plurality of constructed sample images, the constructed sample images are synthesized by background images and a plurality of character images, the background images and the plurality of character images are obtained from a first image and a second image, the first image is an image of an SMD chip shot at different irradiation angles under the same light source, and the second image is an image of an SMD chip shot at different SMD chip pose under the same light source and the same irradiation angle; the original sample set comprises a plurality of sample images, the sample images and the constructed sample images comprise character defect information, the character defect information comprises character dirt, printing omission and pinholes, and the lighting mode of the first image and the second image during shooting is a front bright field;
and the training module is used for training the constructed initial character recognition model by utilizing the original sample set and the enhanced sample set so as to obtain a character recognition model.
11. A computer device, characterized in that it comprises a memory in which a computer program is stored and a processor which executes the computer program, implementing the method according to any of claims 1-8.
12. A computer readable storage medium, having stored thereon a computer program, the computer program being executable by a processor to implement the method of any of claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310633652.8A CN116343213B (en) | 2023-05-31 | 2023-05-31 | Model training and chip character recognition method, device, equipment and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310633652.8A CN116343213B (en) | 2023-05-31 | 2023-05-31 | Model training and chip character recognition method, device, equipment and medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116343213A CN116343213A (en) | 2023-06-27 |
| CN116343213B true CN116343213B (en) | 2023-08-25 |
Family
ID=86880838
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310633652.8A Active CN116343213B (en) | 2023-05-31 | 2023-05-31 | Model training and chip character recognition method, device, equipment and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116343213B (en) |
Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2017116556A (en) * | 2017-02-22 | 2017-06-29 | 株式会社東芝 | Defect detection device and defect detection method |
| CN108982508A (en) * | 2018-05-23 | 2018-12-11 | 江苏农林职业技术学院 | A kind of plastic-sealed body IC chip defect inspection method based on feature templates matching and deep learning |
| CN110763685A (en) * | 2019-10-22 | 2020-02-07 | 陕西源杰半导体技术有限公司 | Artificial intelligent detection method and device for DFB semiconductor laser chip surface defects |
| CN110929773A (en) * | 2019-11-18 | 2020-03-27 | 上海集成电路研发中心有限公司 | Chip surface defect classification device and method based on deep learning |
| CN111145164A (en) * | 2019-12-30 | 2020-05-12 | 上海感图网络科技有限公司 | IC chip defect detection method based on artificial intelligence |
| CN112082999A (en) * | 2020-07-27 | 2020-12-15 | 深检数据科技(江苏)有限公司 | Industrial product defect detection method and industrial intelligent camera |
| CN112730460A (en) * | 2020-12-08 | 2021-04-30 | 北京航天云路有限公司 | Welding defect and intensive rosin joint detection technology for communication IC chip |
| CN112816493A (en) * | 2020-05-15 | 2021-05-18 | 奕目(上海)科技有限公司 | Chip routing defect detection method and device |
| CN113221889A (en) * | 2021-05-25 | 2021-08-06 | 中科芯集成电路有限公司 | Anti-interference recognition method and device for chip characters |
| CN113256598A (en) * | 2021-06-09 | 2021-08-13 | 合肥中科星翰科技有限公司 | Visual detection system for chip production |
| CN114998192A (en) * | 2022-04-19 | 2022-09-02 | 深圳格芯集成电路装备有限公司 | Defect detection method, device and equipment based on deep learning and storage medium |
| CN115308222A (en) * | 2022-07-11 | 2022-11-08 | 江苏汤谷智能科技有限公司 | System and method for identifying bad chip appearance based on machine vision |
| CN115830599A (en) * | 2023-02-08 | 2023-03-21 | 成都数联云算科技有限公司 | Industrial character recognition method, model training method, device, equipment and medium |
| CN115909359A (en) * | 2022-10-25 | 2023-04-04 | 扬州工业职业技术学院 | Capacitance character detection method based on intelligent vision |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7525659B2 (en) * | 2003-01-15 | 2009-04-28 | Negevtech Ltd. | System for detection of water defects |
| AU2003241977A1 (en) * | 2003-05-30 | 2005-01-21 | Advantest Corporation | Electronic component test instrument |
-
2023
- 2023-05-31 CN CN202310633652.8A patent/CN116343213B/en active Active
Patent Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2017116556A (en) * | 2017-02-22 | 2017-06-29 | 株式会社東芝 | Defect detection device and defect detection method |
| CN108982508A (en) * | 2018-05-23 | 2018-12-11 | 江苏农林职业技术学院 | A kind of plastic-sealed body IC chip defect inspection method based on feature templates matching and deep learning |
| CN110763685A (en) * | 2019-10-22 | 2020-02-07 | 陕西源杰半导体技术有限公司 | Artificial intelligent detection method and device for DFB semiconductor laser chip surface defects |
| CN110929773A (en) * | 2019-11-18 | 2020-03-27 | 上海集成电路研发中心有限公司 | Chip surface defect classification device and method based on deep learning |
| CN111145164A (en) * | 2019-12-30 | 2020-05-12 | 上海感图网络科技有限公司 | IC chip defect detection method based on artificial intelligence |
| CN112816493A (en) * | 2020-05-15 | 2021-05-18 | 奕目(上海)科技有限公司 | Chip routing defect detection method and device |
| CN112082999A (en) * | 2020-07-27 | 2020-12-15 | 深检数据科技(江苏)有限公司 | Industrial product defect detection method and industrial intelligent camera |
| CN112730460A (en) * | 2020-12-08 | 2021-04-30 | 北京航天云路有限公司 | Welding defect and intensive rosin joint detection technology for communication IC chip |
| CN113221889A (en) * | 2021-05-25 | 2021-08-06 | 中科芯集成电路有限公司 | Anti-interference recognition method and device for chip characters |
| CN113256598A (en) * | 2021-06-09 | 2021-08-13 | 合肥中科星翰科技有限公司 | Visual detection system for chip production |
| CN114998192A (en) * | 2022-04-19 | 2022-09-02 | 深圳格芯集成电路装备有限公司 | Defect detection method, device and equipment based on deep learning and storage medium |
| CN115308222A (en) * | 2022-07-11 | 2022-11-08 | 江苏汤谷智能科技有限公司 | System and method for identifying bad chip appearance based on machine vision |
| CN115909359A (en) * | 2022-10-25 | 2023-04-04 | 扬州工业职业技术学院 | Capacitance character detection method based on intelligent vision |
| CN115830599A (en) * | 2023-02-08 | 2023-03-21 | 成都数联云算科技有限公司 | Industrial character recognition method, model training method, device, equipment and medium |
Non-Patent Citations (1)
| Title |
|---|
| 基于改进Hu矩和矩形拟合的芯片识别定位算法;寸毛毛等;《包装工程》(第03期);第161-166页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116343213A (en) | 2023-06-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10943107B2 (en) | Simulating image capture | |
| JP2019100917A (en) | Inspection program generation system, generation method of inspection program and generation program of inspection program | |
| CN110110714A (en) | Method and system are corrected automatically on a kind of line of papery operation | |
| CN114266773A (en) | Display panel defect positioning method, device, equipment and storage medium | |
| CN116109631B (en) | PCBA (printed circuit board assembly) cold joint defect detection method, device, equipment and medium | |
| CN110147787A (en) | Bank's card number automatic identifying method and system based on deep learning | |
| CN109379534A (en) | Handle method, apparatus, terminal and the storage medium of image | |
| CN111507325B (en) | Industrial visual OCR recognition system and method based on deep learning | |
| CN116343213B (en) | Model training and chip character recognition method, device, equipment and medium | |
| CN116091500B (en) | Diffusion plate defect detection method, model training method, device, equipment and medium | |
| CN113298069A (en) | Steel seal character recognition method and device, computer equipment and storage medium | |
| CN115830599B (en) | Industrial character recognition method, model training method, device, equipment and medium | |
| CN116579363A (en) | Intelligent recognition and error correction system and method for cigarette bar codes based on artificial intelligence | |
| JP2021022028A (en) | Image processing system, image processing method and image processing program | |
| JP3230218U (en) | Defect identification system for paintings | |
| CN112183526A (en) | Visible light positioning method, device, equipment and medium | |
| CN115689994B (en) | Nameplate and bar code defect detection method, equipment and storage medium | |
| CN111556251A (en) | Electronic book generation method, device and medium | |
| CN115439720B (en) | CAM image reconstruction method, training method, device, equipment and medium | |
| Avgunštincic et al. | Optical Character Recognition for Expiration Dates Validation | |
| CN117119113B (en) | Camera self-calibration method and device of electronic equipment and electronic equipment | |
| US20220188995A1 (en) | Imaging processing device and imaging processing system | |
| CN116188258A (en) | Data acquisition method, model training method, image stitching method and related devices | |
| CN116753496A (en) | Stage lamp maintenance method and system | |
| CN117908742A (en) | Image processing method and device, electronic equipment and 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |