CN116798025A

CN116798025A - Chip component identification method, model training method and device

Info

Publication number: CN116798025A
Application number: CN202210237975.0A
Authority: CN
Inventors: 赵佳鹏; 佘爽
Original assignee: Guangzhou Shiyuan Electronics Thecnology Co Ltd; Guangzhou Shiyuan Artificial Intelligence Innovation Research Institute Co Ltd
Current assignee: Guangzhou Shiyuan Electronics Thecnology Co Ltd; Guangzhou Shiyuan Artificial Intelligence Innovation Research Institute Co Ltd
Priority date: 2022-03-10
Filing date: 2022-03-10
Publication date: 2023-09-22

Abstract

The embodiment of the disclosure provides a method for identifying chip components, a method for training a model and a device, comprising the following steps: the method comprises the steps of obtaining the polarity attribute of the chip component to be identified, wherein the polarity attribute is polar or nonpolar, calling an identification method for identifying the chip component to be identified according to the polarity attribute, wherein the identification methods corresponding to the polar and the nonpolar are different, and carrying out identification processing on the chip component to be identified according to the called identification method to obtain an identification result, so that the diversity and the flexibility of the identification processing on the chip component are improved, the identification processing on the chip component is targeted, the effectiveness and the reliability of the identification processing are improved, and the technical effect of improving the accuracy of the determined identification result is achieved.

Description

Chip component identification method, model training method and device

Technical Field

The embodiment of the disclosure relates to deep learning in artificial intelligence technology, in particular to a method for identifying chip components, a method for training a model and a device thereof.

Background

A chip (microchip) device is a generic term for semiconductor device products used to form integrated circuits (integrated circuit, ICs).

In the prior art, the method for identifying the chip components mainly comprises the following steps: and acquiring a character string on the chip component, and determining the identification result of the chip component by combining the character string and a preset template.

However, with the above method, there is a technical problem that accuracy of recognition is low.

Disclosure of Invention

The embodiment of the disclosure provides a method for identifying chip components, a method and a device for training a model, which are used for solving the problem of low accuracy of identifying the chip components.

In a first aspect, an embodiment of the present disclosure provides a method for identifying a chip component, including:

acquiring polarity attributes of chip components to be identified, wherein the polarity attributes are polar or nonpolar;

calling an identification method for identifying the chip components to be identified according to the polarity attribute, wherein the identification methods corresponding to the polarity and the non-polarity are different;

and carrying out recognition processing on the chip components to be recognized according to the called recognition method to obtain a recognition result.

In some embodiments, if the polarity attribute is non-polarity, performing recognition processing on the chip component to be recognized according to the invoked recognition method to obtain a recognition result, where the recognition result includes:

Acquiring an image to be identified of the chip component to be identified;

if the image to be identified is an inverted image, the inverted image is identified based on a preset mapping relation, and the identification result is obtained, wherein the mapping relation is used for representing the corresponding relation between the inverted text and the characters.

In some embodiments, if the image to be identified is an inverted image, performing identification processing on the inverted image based on a preset mapping relationship to obtain the identification result, including:

if the image to be identified is an inverted image, acquiring an inverted description text of the chip component to be identified in the inverted image;

and determining characters corresponding to the reversed descriptive text according to the mapping relation, and determining the recognition result according to the determined characters corresponding to the reversed descriptive text.

In some embodiments, determining the recognition result according to the determined character corresponding to the reversed descriptive text includes:

and carrying out reflection processing on the determined characters corresponding to the reversed description text according to the mapping relation to obtain the identification result.

In some embodiments, if the polarity attribute is non-polarity, invoking an identification method for identifying the chip component to be identified according to the polarity attribute, including: invoking a nonpolar recognition model;

and carrying out recognition processing on the chip components to be recognized according to the called recognition method to obtain a recognition result, wherein the recognition result comprises the following steps: performing identification processing on the chip components to be identified based on the nonpolar identification model to obtain the identification result;

the nonpolar recognition model is trained based on a first sample data set, the first sample data set comprises a first sample image, the first sample image is an inverted image, and the first sample image comprises description text of nonpolar chip components.

In some embodiments, if the polarity attribute is polarity, performing recognition processing on the chip component to be recognized according to a called recognition method to obtain a recognition result, where the recognition result includes:

acquiring an image to be identified of the chip component to be identified;

and carrying out image enhancement processing on the image to be identified to obtain an enhanced image, and carrying out identification processing on the enhanced image to obtain the identification result.

In some embodiments, if the polarity attribute is polar, invoking an identification method for identifying the chip component to be identified according to the polarity attribute, including: invoking a polarity recognition model;

and carrying out recognition processing on the chip components to be recognized according to the called recognition method to obtain a recognition result, wherein the recognition result comprises the following steps: performing identification processing on the chip components to be identified based on the polar identification model to obtain the identification result;

the polar recognition model is trained based on a second sample data set, wherein the second sample data set comprises a second sample image, and the second sample image comprises description text of a polar chip component.

In some embodiments, the polar recognition model is obtained by performing image enhancement processing on the second sample image and training based on the second sample image and the enhanced image.

In some embodiments, the polar recognition model is obtained by performing a random masking process on the second sample image, obtaining a masked image, and training based on the second sample image and the masked image.

In some embodiments, the polar recognition model is obtained by performing random masking on the second sample image to obtain a masking region, performing blurring on the masking region to obtain a masked image, and training based on the second sample image and the masked image, wherein the blurring comprises gaussian blurring and/or median blurring.

In a second aspect, an embodiment of the present disclosure provides a model training method, including:

acquiring a first sample data set, wherein the first sample data set comprises a first sample image, and the first sample image is an inverted image;

training according to the first sample data set to obtain a nonpolar identification model, wherein the nonpolar identification model is used for carrying out identification processing on the chip components to be identified if the polarity attribute of the chip components to be identified is nonpolar, so as to obtain an identification result.

In some embodiments, the first sample image includes descriptive text for a chip component that is nonpolar; training according to the first sample data set to obtain a nonpolar recognition model, wherein the nonpolar recognition model comprises the following steps:

constructing a mapping relation according to the descriptive text in the first sample image, wherein the mapping relation is used for representing the corresponding relation between the reversed text and the characters;

And training to obtain the nonpolar recognition model according to the first sample image and the mapping relation.

In some embodiments, training to obtain the nonpolar recognition model according to the first sample image and the mapping relation includes:

predicting predicted characters of the descriptive text in the first sample image according to a preset basic network model;

and training the basic network model according to the predicted character and the mapping relation to obtain the nonpolar recognition model.

In some embodiments, training the basic network model according to the predicted character and the mapping relation to obtain the nonpolar recognition model includes:

determining real characters of the descriptive text in the first sample image according to the mapping relation;

and calculating the difference information between the predicted character and the real character, and adjusting the model parameters of the basic network model according to the difference information to obtain the nonpolar recognition model.

In some embodiments, obtaining the first sample data set includes:

acquiring an image of a nonpolar chip component;

and acquiring an inverted image from the image of the nonpolar chip component, and constructing the first sample data set according to the inverted image.

In a third aspect, an embodiment of the present disclosure provides an identification apparatus for a chip component, including:

the first acquisition unit is used for acquiring the polarity attribute of the chip component to be identified, wherein the polarity attribute is polar or nonpolar;

the calling unit is used for calling the identification method for identifying the chip components to be identified according to the polarity attribute, wherein the identification methods corresponding to the polarity and the non-polarity are different;

and the identification unit is used for carrying out identification processing on the chip components to be identified according to the called identification method to obtain an identification result.

In some embodiments, if the polarity attribute is non-polar, the identifying unit includes:

the first acquisition subunit is used for acquiring the image to be identified of the chip component to be identified;

and the first recognition subunit is used for recognizing the inverted image based on a preset mapping relation to obtain the recognition result if the image to be recognized is the inverted image, wherein the mapping relation is used for representing the corresponding relation between the inverted text and the character.

In some embodiments, the first identification subunit comprises:

The acquisition module is used for acquiring the reversed description text of the chip components to be identified in the reversed image if the image to be identified is the reversed image;

and the determining module is used for determining the characters corresponding to the reversed descriptive text according to the mapping relation and determining the recognition result according to the determined characters corresponding to the reversed descriptive text.

In some embodiments, the determining module is configured to perform, according to the mapping relationship, a de-mapping process on the determined character corresponding to the inverted description text, to obtain the recognition result.

In some embodiments, the invoking unit is configured to invoke the non-polarity recognition model;

the identification unit is used for carrying out identification processing on the chip components to be identified based on the nonpolar identification model to obtain the identification result;

In some embodiments, if the polarity attribute is polar, the identifying unit includes:

the second acquisition subunit is used for acquiring the image to be identified of the chip component to be identified;

the processing subunit is used for carrying out image enhancement processing on the image to be identified to obtain an enhanced image;

and the second recognition subunit is used for carrying out recognition processing on the enhanced image to obtain the recognition result.

In some embodiments, if the polarity attribute is polarity, the calling unit is configured to call a polarity identification model;

the identification unit is used for carrying out identification processing on the chip components to be identified based on the polar identification model to obtain the identification result;

In a fourth aspect, an embodiment of the present disclosure provides a model training apparatus, including:

a second acquisition unit configured to acquire a first sample data set, wherein the first sample data set includes a first sample image, and the first sample image is an inverted image;

the training unit is used for training according to the first sample data set to obtain a nonpolar identification model, wherein the nonpolar identification model is used for carrying out identification processing on the chip components to be identified if the polarity attribute of the chip components to be identified is nonpolar, so as to obtain an identification result.

In some embodiments, the first sample image includes descriptive text for a chip component that is nonpolar; the training unit comprises:

a construction subunit, configured to construct a mapping relationship according to the descriptive text in the first sample image, where the mapping relationship is used to characterize a correspondence between the inverted text and the character;

and the training subunit is used for training to obtain the nonpolar recognition model according to the first sample image and the mapping relation.

In some embodiments, the training subunit comprises:

the prediction module is used for predicting the predicted characters of the descriptive text in the first sample image according to a preset basic network model;

and the training module is used for training the basic network model according to the predicted character and the mapping relation to obtain the nonpolar recognition model.

In some embodiments, the training module comprises:

a determining submodule, configured to determine, according to the mapping relationship, a real character of a descriptive text in the first sample image;

a calculation sub-module for calculating difference information between the predicted character and the real character;

and the adjustment sub-module is used for adjusting the model parameters of the basic network model according to the difference information to obtain the nonpolar identification model.

In some embodiments, the second acquisition unit includes:

a third obtaining subunit, configured to obtain an image of a nonpolar chip component, and obtain an inverted image from the image of the nonpolar chip component;

a construction subunit for constructing the first sample data set from the inverted image.

In a fifth aspect, embodiments of the present disclosure provide an electronic device, including: a memory, a processor;

a memory; a memory for storing the processor-executable instructions;

wherein the processor is configured to perform the method of the first or second aspect.

In a sixth aspect, embodiments of the present disclosure provide a computer-readable storage medium having stored therein computer-executable instructions for implementing the method according to the first or second aspect when executed by a processor.

In a seventh aspect, the presently disclosed embodiments provide a computer program product comprising a computer program which, when executed by a processor, implements the method according to the first or second aspect.

According to the identification method, the model training method and the device for the chip components, the polarity attribute of the chip components to be identified is acquired, the identification method corresponding to the polarity attribute is called to conduct identification processing on the chip components to be identified, the diversity and the flexibility of identification processing on the chip components are improved, the identification processing on the chip components is targeted, the effectiveness and the reliability of the identification processing are improved, and the technical effect of improving the accuracy of the determined identification result is achieved.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the disclosure and together with the description, serve to explain the principles of the disclosure.

FIG. 1 is a schematic diagram of a chip component according to one embodiment;

FIG. 2 is a schematic diagram of a chip component according to another embodiment;

FIG. 3 is a schematic diagram of a method of identifying chip components according to one embodiment of the disclosure;

FIG. 4 is a schematic diagram of a method of identifying chip components according to another embodiment of the disclosure;

FIG. 5 is a schematic diagram of a method of identifying chip components according to another embodiment of the disclosure;

FIG. 6 is a schematic diagram of a chip component according to another embodiment;

FIG. 7 is a schematic diagram of a model training method according to one embodiment of the present disclosure;

FIG. 8 is a schematic diagram of an identification device for chip components according to one embodiment of the disclosure;

FIG. 9 is a schematic diagram of a model training apparatus according to one embodiment of the present disclosure;

fig. 10 is a block diagram of an electronic device of a method of identifying chip components, a method of model training, according to an embodiment of the present disclosure.

Specific embodiments of the present disclosure have been shown by way of the above drawings and will be described in more detail below. These drawings and the written description are not intended to limit the scope of the disclosed concepts in any way, but rather to illustrate the disclosed concepts to those skilled in the art by reference to specific embodiments.

Detailed Description

Reference will now be made in detail to exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, the same numbers in different drawings refer to the same or similar elements, unless otherwise indicated. The implementations described in the following exemplary examples are not representative of all implementations consistent with the present disclosure. Rather, they are merely examples of apparatus and methods consistent with some aspects of the present disclosure as detailed in the accompanying claims.

The chip components are used to form an integrated circuit, and the chip components may be classified into a chip component having a polarity and a chip component having no polarity according to whether the chip component has a polarity.

In the related art, the same identification method is generally used to identify the chip components having polarity and the chip components having no polarity. That is, when the chip component is identified, it is not necessary to distinguish whether the chip component is a polar chip component or a nonpolar chip component.

For example, the following method may be used to identify the chip components:

in one example, a character string of the chip component may be obtained, the obtained character string is segmented to obtain a single character, and the single character is matched with a preset character template to obtain a recognition result of the chip component, i.e., to obtain text content of the chip component.

However, since the background of the chip component is complex, and the image of the chip component may be relatively low, as shown in fig. 1, it is difficult to segment a single character by using the above method, so that the technical problem of low recognition accuracy is caused.

In another example, the chip components may be identified based on a deep learning algorithm, such as a time series class classification (Connectionist Temporal Classification, CTC) algorithm through an encoding-decoding network and a neural network, to obtain an identification result of the chip components.

However, since the chip components may be inverted, as shown in fig. 2, accurate recognition cannot be performed by the above method.

In some embodiments, the direction recognition model may be trained based on the above examples, so as to perform recognition processing on the chip component in combination with the direction recognition model.

However, on the one hand, training the direction recognition model requires time and cost, and in recognition processing, since the direction recognition model is added, the time of recognition processing is increased, that is, recognition efficiency is relatively reduced, and the risk of recognition errors is relatively added, so that the accuracy of recognition is reduced.

In order to avoid at least one of the above technical problems, the inventors of the present disclosure have made creative efforts to obtain the inventive concept of the present disclosure: the polarity attribute of the chip component to be identified is obtained, different identification methods are called according to different polarity attributes, and the chip component to be identified is identified based on the called identification methods.

The following describes the technical solutions of the present disclosure and how the technical solutions of the present disclosure solve the above technical problems in detail with specific embodiments. The following embodiments may be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments. Embodiments of the present disclosure will be described below with reference to the accompanying drawings.

Referring to fig. 3, fig. 3 is a schematic diagram of a method for identifying a chip component according to an embodiment of the disclosure, including:

s301: and acquiring the polarity attribute of the chip component to be identified.

Wherein the polarity attribute is polar or non-polar.

The execution body of the embodiment is an identification device (hereinafter simply referred to as an identification device) of a chip component, and the identification device may be a server (such as a cloud server, a local server, or a server cluster), may also be a terminal device, may also be a computer, may also be a processor, may also be a chip, or the like, and the embodiment is not limited.

In combination with the above analysis, the chip component has a polarity attribute, such as polarity or non-polarity, so that the chip component to be identified also has a polarity attribute, for example, for characterizing that the chip component to be identified is a chip component with polarity, or for characterizing that the chip component to be identified is a chip component of an inorganic type.

The manner of acquiring the polarity attribute is not limited in this embodiment. For example, when the chip components to be identified are deployed, if the chip components to be identified need to be deployed in a certain insertion mode, the polarity attribute of the chip components to be identified is described as having polarity; otherwise, if the chip components to be identified can be deployed in a random insertion mode, the polarity attribute of the chip components to be identified is described as non-polarity.

S302: and calling an identification method for identifying the chip components to be identified according to the polarity attribute.

Wherein, the identification methods corresponding to the polarity and the nonpolarity are different.

Exemplary methods for identifying chip components include two methods, one for identifying chip components having a polarity and one for identifying chip components having no polarity.

Correspondingly, if the chip component to be identified is a chip component with polarity, invoking a method for identifying the chip component with polarity so as to identify the chip component to be identified based on the identification method; and if the chip component to be identified is a nonpolar chip component, invoking a method for identifying the nonpolar chip component so as to identify the chip component to be identified based on the identification method.

The method for identifying the polar chip components can be that the polar chip components are analyzed to obtain common characteristics of the polar chip components and the common characteristics are generated based on the common characteristics.

Correspondingly, the method for identifying the nonpolar chip components can be that the nonpolar chip components are analyzed to obtain the common characteristics of the nonpolar chip components and the common characteristics are generated based on the common characteristics.

S303: and carrying out recognition processing on the chip components to be recognized according to the called recognition method to obtain a recognition result.

Based on the above analysis, the embodiment of the disclosure provides a method for identifying a chip component, which includes: the method comprises the steps of obtaining the polarity attribute of the chip component to be identified, wherein the polarity attribute is polar or non-polar, calling an identification method for identifying the chip component to be identified according to the polarity attribute, wherein the identification methods corresponding to the polar and non-polar are different, and carrying out identification processing on the chip component to be identified according to the called identification method to obtain an identification result.

Referring to fig. 4, fig. 4 is a schematic diagram of a method for identifying a chip component according to another embodiment of the disclosure, including:

s401: and acquiring the polarity attribute of the chip component to be identified.

Wherein the polarity attribute is polar or non-polar.

It should be understood that, in order to avoid a cumbersome statement, the present embodiment is not limited with respect to the same technical features as the above-described embodiment.

S402: and if the polarity attribute is nonpolar, acquiring an image to be identified of the chip component to be identified.

In other embodiments, the image to be identified of the chip component to be identified may also be obtained when the polarity attribute of the chip component to be identified is obtained.

Alternatively, the image to be recognized of the chip component to be recognized may be obtained first, and then the polarity attribute of the chip component to be recognized may be obtained.

That is, the present embodiment does not limit the order of acquiring the polarity attribute and the images to be identified. In order to facilitate the determination of which recognition method is used to recognize the chip components to be recognized, it is possible to determine the polarity attribute first and then acquire the image to be recognized.

The manner of acquiring the image to be identified is not limited in this embodiment, for example:

In one example, the recognition device may be coupled to the image capture device and receive the image to be recognized transmitted by the image capture device.

In another example, the identification device may provide an image-loading tool by which a user may transmit an image to be identified to the identification device.

The image loading tool can be an interface used for being connected with external equipment, such as an interface used for being connected with other storage equipment, and the image to be identified transmitted by the external equipment is obtained through the interface; the image loading tool may also be a display device, for example, the recognition device may input an interface for loading the image function on the display device, through which the user may import the image to be recognized into the recognition device, and the recognition device obtains the imported image to be recognized.

S403: if the image to be identified is the reverse image, the reverse image is identified based on a preset mapping relation, and an identification result is obtained.

Wherein the mapping relationship is used to characterize the correspondence between the inverted text and the character.

For example, after the chip component to be identified is determined to be a nonpolar chip component and the image to be identified is obtained, whether the image to be identified is an inverted image may be determined, and if the image to be identified is the inverted image as shown in fig. 2, the inverted image is identified according to the mapping relationship.

In this embodiment, a mapping relationship may be pre-constructed to represent a correspondence between the inverted text and the characters by the mapping relationship, so as to identify the inverted image based on the mapping relationship, thereby obtaining the characters corresponding to the inverted image, and further obtaining the identification result.

In combination with the above analysis, in some embodiments, the reverse image may be identified by training the direction identification model, however, the method may reduce the identification efficiency, may reduce the identification accuracy, and requires time and resources to train.

In some embodiments, S403 may include the steps of:

a first step of: and if the image to be identified is the reverse image, acquiring the reverse description text of the chip components to be identified in the reverse image.

The description text refers to a text describing the chip components, such as a description of the type of the chip components as diodes shown in fig. 2.

In this embodiment, since the chip component to be identified is an upside-down chip component, and the image to be identified is an upside-down image, the text obtained for describing the chip component to be identified is an upside-down text (i.e., upside-down description text).

And a second step of: and determining the characters corresponding to the reversed descriptive text according to the mapping relation, and determining the recognition result according to the determined characters corresponding to the reversed descriptive text.

In combination with the above analysis, the mapping relationship is characterized as a correspondence relationship between the inverted text and the characters, and then, after the inverted descriptive text is acquired, the characters having the correspondence relationship with the inverted descriptive text can be determined from the mapping relationship.

Illustratively, in the mapping relationship, "inverted A" is associated with the character "≡! "having correspondence, if the description text of the inversion includes" inverted A ", the character" correspondingto "inverted A" -! ".

It should be understood that the above-described mapping relationships are for exemplary purposes only and are not to be construed as limiting the mapping relationships. In some embodiments, the mapping relationship may be constructed based on a requirement, a history, and a test, which is not limited in this embodiment.

In this embodiment, the reverse description text is determined, so that the recognition result is determined by combining the mapping relation and the reverse description text, so that the defects of resource consumption, efficiency reduction and the like caused by performing recognition processing on the chip components to be recognized based on the direction recognition model, which are obtained by training in the above embodiment, are avoided, and the technical effects of efficiency and reliability of performing recognition processing on the chip components to be recognized are improved.

In some embodiments, determining the recognition result based on the determined character corresponding to the inverted descriptive text includes: and carrying out reflection processing on the determined characters corresponding to the reversed descriptive text according to the mapping relation to obtain a recognition result.

Illustratively, in connection with the above analysis, if the character corresponding to the reversed descriptive text is "+|! And carrying out reflection processing according to the mapping relation, so that a non-inverted description text including 'A' can be obtained, and the identification result is determined.

Referring to fig. 2, the reversed descriptive text includes a plurality of characters (such as letters and numbers shown in fig. 2), then the characters corresponding to each character can be determined according to the mapping relationship, and the reflection processing is performed on each character, so as to obtain non-reversed characters corresponding to each character, and the splicing processing is performed on each non-reversed character, so as to obtain the recognition result.

In other embodiments, in addition to determining the identification result of the chip component to be identified in the above manner, the identification result of the chip component to be identified may also be determined in a model manner.

For example, an apolar recognition model for recognizing a chip component with an apolar attribute may be constructed in advance, so that when it is determined that the polarity attribute of the chip component to be recognized is apolar, the apolar recognition model is called to determine a recognition result of the chip component to be recognized based on the apolar recognition model.

The nonpolar recognition model is trained based on a first sample data set, wherein the first sample data set comprises a first sample image, the first sample image is an inverted image, and the first sample image comprises description text of nonpolar chip components.

It should be understood that "first" in a first sample data set is used to distinguish from other sample data sets, such as from a second sample data set hereafter, and is not to be construed as limiting the first sample data set.

Similarly, a "first" of the first sample images is used to distinguish from other sample images, such as from a second sample image hereafter, and is not to be construed as limiting the first sample image.

The number of the first sample images may be determined based on the requirements, the history, the test, and the like, which is not limited in this embodiment.

For example, the number of first sample images may be relatively large for relatively high accuracy scenes, and the number of first sample images may be relatively small for relatively low accuracy scenes.

That is, a plurality of reversed images of the non-polar chip components may be acquired, each reversed image including a description text for describing the non-polar chip components, the description text being reversed text, and the first sample dataset being constructed based on the reversed image of each non-polar chip component to be trained to obtain the non-polar identification model based on the first sample dataset.

The present embodiment is not limited as to the manner in which the nonpolar recognition model is obtained based on the training of the first sample data set. For example, the base network model may be trained based on the first sample data set to obtain a non-polar recognition model, and the framework and structure of the base network model is not limited in this embodiment.

It should be noted that, in this embodiment, the nonpolar recognition model is obtained by combining the reverse image training of the description text including the nonpolar chip components, so that when the polarity attribute of the chip components to be recognized is determined to be nonpolar, the nonpolar recognition model is called to perform recognition processing on the chip components to be recognized, so as to obtain a recognition result, and pertinence of the recognition processing on the nonpolar chip components can be realized, thereby improving the accuracy and reliability of the recognition processing.

In combination with the above analysis, the determination result of determining whether the image to be identified is an inverted image may be that the image to be identified is an inverted image as in the above embodiment, or that the image to be identified is a non-inverted image.

For example, if the image to be recognized is a non-inverted image, that is, an image that is normally displayed, the image to be recognized may be recognized by using an optical character recognition (Optical Character Recognition, OCR) technique, for example, so as to obtain a recognition result.

Referring to fig. 5, fig. 5 is a schematic diagram of a method for identifying a chip component according to another embodiment of the disclosure, including:

s501: and acquiring the polarity attribute of the chip component to be identified.

Wherein the polarity attribute is polar or non-polar.

Similarly, in order to avoid a cumbersome statement, the present embodiment is not limited to the same technical features as the above-described embodiment.

S502: and if the polarity attribute is polar, acquiring an image to be identified of the chip component to be identified.

S503: and performing image enhancement processing on the image to be identified to obtain an enhanced image, and performing recognition processing on the enhanced image to obtain a recognition result.

In this embodiment, if the polarity attribute of the chip component to be identified is polar, the image to be identified is a non-inverted image, that is, an image displayed normally, then image enhancement processing may be performed on the image to be identified, so as to avoid problems such as low contrast and unclear contrast of the image to be identified, that is, the enhanced image is clearer than the image to be identified, so that reliability and accuracy of the identification processing may be improved by performing identification processing on the enhanced image, and further, the identification result has a technical effect of higher accuracy and reliability.

In other embodiments, the image to be identified may be identified by using a model mode, so as to obtain an identification result.

For example, a polar recognition model for recognizing a chip component with a polar attribute may be constructed in advance, so that when it is determined that the polar attribute of the chip component to be recognized is polar, the polar recognition model is called to determine a recognition result of the chip component to be recognized based on the polar recognition model.

The polar identification model is trained based on a second sample data set, wherein the second sample data set comprises a second sample image, and the second sample image comprises description text of the polar chip components.

Similarly, the number of second sample images may be determined based on the requirements, the history, and the test, which is not limited in this embodiment.

That is, a plurality of images of the chip components with polarities, that is, a plurality of second sample images, may be acquired, and each of the second sample images includes a description text for describing the chip components with polarities, and a second sample data set is constructed based on the second sample image of each chip component with polarities, so as to obtain a polarity recognition model based on the training of the second sample data set.

The present embodiment is not limited as to the manner in which the polar recognition model is obtained based on the training of the second sample data set. For example, the base network model may be trained based on the second sample data set to obtain a polar recognition model, and the framework and structure of the base network model is not limited in this embodiment.

Similarly, in this embodiment, the polar recognition model is obtained by combining the second sample image including the description text of the polar chip component, so that when the polar attribute of the chip component to be recognized is determined to be polar, the polar recognition model is called to perform recognition processing on the chip component to be recognized, so as to obtain a recognition result, and pertinence of the recognition processing on the polar chip component can be realized, thereby improving the accuracy and reliability of the recognition processing.

In some embodiments, training the polar recognition model may include the steps of:

a first step of: and acquiring a second sample image, and performing image enhancement processing on the second sample image to obtain an enhanced image.

The image enhancement processing manner is not limited in this embodiment, for example, the image enhancement processing may be performed on the second sample image by using a geometric transformation manner, so as to obtain an enhanced image; for another example, the second sample image may be enhanced by using a color transform method, to obtain an enhanced image, and so on, which are not listed here.

And the second sample image is subjected to image enhancement processing in a geometric transformation mode, and when the enhanced image is obtained, the second sample image can be subjected to image enhancement processing in a turnover, rotation, cutting, deformation, scaling and other modes, so that the enhanced image is obtained.

And a second step of: and training according to the second sample image and the image after the enhancement processing to obtain a polar identification model.

In combination with the above analysis, the second sample image is an image before the image enhancement processing, that is, in this embodiment, the image before and after the image enhancement processing is taken as the second sample data set, so that the polar recognition model is obtained by training based on the second sample data set, so that the polar recognition model has a relatively strong discrimination capability, and even if the text in the image to be recognized of the chip component to be recognized is blurred, accurate recognition can be performed, thereby improving the technical effects of determining the accuracy and reliability of the recognition result.

In other embodiments, training the polar recognition model may include the steps of:

a first step of: and acquiring a second sample image, and carrying out random masking processing on the second sample image to obtain a masked image.

And a second step of: and training based on the second sample image and the image subjected to the mask processing to obtain a polar identification model.

The random mask processing means that part of the area in the second sample image is randomly covered, so that the polar recognition model has strong recognition capability, and even when the part of the area in the image to be recognized of the chip component to be recognized is blurred, the recognition result of the chip component to be recognized, namely the text content of the chip component to be recognized, can be accurately determined.

In some embodiments, performing a random masking process on the second sample image, resulting in a masked image may include: and carrying out random masking processing on the second sample image to obtain a masking region, and carrying out blurring processing on the masking region to obtain a masked image.

Wherein the blurring process includes a gaussian blurring process and/or a median blurring process.

For example, if the second sample image is AS shown in fig. 6, text content (or referred to AS a tag) of the second sample image, that is, "AS431A", may be acquired, and the number of words subjected to the random masking may be predetermined (may be a fixed value), or the number of words subjected to the random masking may be determined according to the length of the text content.

The width of each word, such as the width of "a", may be predicted, the width of the word may be determined as the width of the mask region, and the width of the mask region may be determined by dividing the length of the second sample image by the total number of words in the text content.

The different mask areas are obtained by randomly initializing the starting positions of the mask areas. For example, the mask area may be an area where "a" is located, or may be an area where "S" is located, which is not listed here.

And carrying out Gaussian blur processing on the mask area, or carrying out median blur processing on the mask area, or carrying out Gaussian blur processing and median blur processing on the mask area, thereby obtaining an image after mask processing.

The Gaussian Blur (Gaussian Blur) process may also be referred to as Gaussian smoothing process, for reducing image noise and reducing the level of detail, as in the present embodiment, for reducing image noise of a mask region and reducing the level of detail.

The median blurring process may also be referred to as median filtering process, which is a nonlinear smoothing technique that sets the gray value of each pixel to be the median of the gray values of all pixels in a certain neighborhood window of the pixel, as in the present embodiment, the gray value of the pixel in the mask region is set to be the median of the gray values of regions adjacent to the mask region.

Of course, in other embodiments, the mask area may be blurred in other manners, such as mean filtering, bilateral filtering, etc., which are not listed here.

In this embodiment, the mask-processed image is determined through the random mask processing and the blurring processing, and the polarity recognition model is constructed by combining the mask-processed image, so that the recognition performance of the polarity recognition model can be improved, and the accuracy and the reliability of the recognition processing can be improved when the recognition processing is performed on the to-be-recognized image of the chip component to be recognized based on the polarity recognition model.

In combination with the above analysis, in some embodiments, a polar recognition model for recognizing a polar chip component may be constructed in advance, and an apolar recognition model for recognizing an apolar chip component may be constructed, so that when the chip component to be recognized needs to be recognized, the polarity attribute of the chip component to be recognized is determined first, that is, whether the chip component to be recognized is a polar chip component or an apolar chip component is determined. If the chip component to be identified is a polar chip component, invoking a polar identification model to identify the chip component to be identified, and if the chip component to be identified is a nonpolar chip component, invoking a nonpolar identification model to identify the chip component to be identified, so as to improve the flexibility and diversity of the chip component identification.

In the above embodiments, the construction of the polarity recognition model is described in detail, and a method of constructing the non-polarity recognition model that may be employed is described with reference to fig. 7.

Referring to fig. 7, fig. 7 is a schematic diagram of a model training method according to an embodiment of the present disclosure.

As shown in fig. 7, the method includes:

s701: a first sample data set is acquired.

Wherein the first sample data set includes a first sample image, the first sample image being an inverted image.

In some embodiments, S701 may include the steps of:

a first step of: an image of a non-polar chip component is acquired.

And a second step of: and acquiring an inverted image from the image of the nonpolar chip component, and constructing a first sample data set according to the inverted image.

S702: training according to the first sample data set to obtain a nonpolar recognition model.

And if the polarity attribute of the chip component to be identified is nonpolar, carrying out identification processing on the chip component to be identified to obtain an identification result.

In some embodiments, the first sample image includes descriptive text for a chip component that is nonpolar; s702 may include the steps of:

a first step of: and constructing a mapping relation according to the descriptive text in the first sample image.

And a second step of: training to obtain a nonpolar recognition model according to the first sample image and the mapping relation.

Illustratively, in connection with the above analysis, the first sample image is an inverted image, and the text in the first sample image for describing the nonpolar chip component is an inverted text, and thus, in this embodiment, the mapping relationship may be constructed based on the description text (i.e., the inverted text) in the first sample image.

In some embodiments, the second step may comprise the sub-steps of:

a first substep: and predicting the predicted characters of the descriptive text in the first sample image according to a preset basic network model.

The type and structure of the basic network model are not limited in this embodiment. For example, the base network model may be a convolutional neural network (CRNN), and in order to reduce the time consuming training of the model, a lightweight network (MobileNet V3) of the neural network model (backbone) may be used, while decoding may be performed using CTC algorithms.

A second substep: and training the basic network model according to the predicted character and the mapping relation to obtain the nonpolar recognition model.

For example, the real character of the descriptive text in the first sample image may be determined according to the mapping relationship, and the difference information between the predicted character and the real character may be calculated, so that the model parameters of the basic network model may be adjusted according to the difference information, thereby obtaining the nonpolar recognition model.

Referring to fig. 8, fig. 8 is a schematic diagram of an identification device of a chip component according to an embodiment of the disclosure.

As shown in fig. 8, the chip component recognition apparatus 800 includes:

the first obtaining unit 801 is configured to obtain a polarity attribute of a chip component to be identified, where the polarity attribute is polar or non-polar.

And a calling unit 802, configured to call, according to the polarity attribute, a recognition method for recognizing the chip component to be recognized, where the recognition methods corresponding to the polarity and the non-polarity are different.

And the identification unit 803 is used for carrying out identification processing on the chip components to be identified according to the called identification method to obtain an identification result.

In some embodiments, if the polarity attribute is non-polar, in conjunction with fig. 8, the identifying unit 803 includes:

The first obtaining subunit 8031 is configured to obtain an image to be identified of the chip component to be identified.

The first recognition subunit 8032 is configured to, if the image to be recognized is an inverted image, perform recognition processing on the inverted image based on a preset mapping relationship, so as to obtain the recognition result, where the mapping relationship is used to characterize a correspondence between inverted text and characters.

In some embodiments, the first recognition subunit 8032 includes:

and the acquisition module is used for acquiring the description text of the inversion of the chip component to be identified in the inversion image if the image to be identified is the inversion image.

In some embodiments, the invoking unit 802 is configured to invoke the non-polarity recognition model.

And the identifying unit 803 is configured to identify the chip component to be identified based on the nonpolar identifying model, so as to obtain the identifying result.

In some embodiments, if the polarity attribute is polar, in conjunction with fig. 8, the identifying unit 803 includes:

a second obtaining subunit 8033, configured to obtain an image to be identified of the chip component to be identified;

a processing subunit 8034, configured to perform image enhancement processing on the image to be identified, so as to obtain an enhanced image;

and a second recognition subunit 8035, configured to perform recognition processing on the enhanced image, so as to obtain the recognition result.

In some embodiments, if the polarity attribute is polarity, the calling unit 802 is configured to call a polarity identification model.

And the identifying unit 803 is configured to identify the chip component to be identified based on the polar identification model, so as to obtain the identification result.

Referring to fig. 9, fig. 9 is a schematic diagram of a model training apparatus according to an embodiment of the present disclosure.

As shown in fig. 9, the model training apparatus 900 includes:

a second obtaining unit 901, configured to obtain a first sample data set, where the first sample data set includes a first sample image, and the first sample image is an inverted image.

In some embodiments, as shown in connection with fig. 9, the second obtaining unit 901 includes:

a third acquiring subunit 9011, configured to acquire an image of a nonpolar chip component, and acquire an inverted image from the image of the nonpolar chip component;

a construction subunit 9012 is configured to construct the first sample data set from the inverted image.

The training unit 902 is configured to train according to the first sample data set to obtain a nonpolar identification model, where the nonpolar identification model is configured to identify a chip component to be identified if a polarity attribute of the chip component to be identified is nonpolar, so as to obtain an identification result.

In some embodiments, the first sample image includes descriptive text for a chip component that is nonpolar; as shown in conjunction with fig. 9, the training unit 902 includes:

a construction subunit 9021, configured to construct a mapping relationship according to the descriptive text in the first sample image, where the mapping relationship is used to characterize a correspondence between the inverted text and the character;

And a training subunit 9022, configured to train to obtain the nonpolar recognition model according to the first sample image and the mapping relationship.

In some embodiments, training subunit 9022 comprises:

and the prediction module is used for predicting the predicted characters of the descriptive text in the first sample image according to a preset basic network model.

In some embodiments, the training module comprises:

According to embodiments of the present disclosure, the present disclosure also provides an electronic device and a readable storage medium.

According to an embodiment of the present disclosure, the present disclosure also provides a computer program product comprising: a computer program stored in a readable storage medium, from which at least one processor of an electronic device can read, the at least one processor executing the computer program causing the electronic device to perform the solution provided by any one of the embodiments described above.

As shown in fig. 10, a block diagram of an electronic device of a method for identifying chip components and a method for training a model according to an embodiment of the present disclosure is shown. Electronic devices are intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular telephones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be exemplary only, and are not meant to limit implementations of the disclosure described and/or claimed herein.

As shown in fig. 10, the electronic device includes: one or more processors 1001, memory 1002, and interfaces for connecting the components, including a high-speed interface and a low-speed interface. The various components are interconnected using different buses and may be mounted on a common motherboard or in other manners as desired. The processor may process instructions executing within the electronic device, including instructions stored in or on memory to display graphical information of the GUI on an external input/output device, such as a display device coupled to the interface. In other embodiments, multiple processors and/or multiple buses may be used, if desired, along with multiple memories and multiple memories. Also, multiple electronic devices may be connected, each providing a portion of the necessary operations (e.g., as a server array, a set of blade servers, or a multiprocessor system). One processor 1001 is illustrated in fig. 10.

Memory 1002 is a non-transitory computer-readable storage medium provided by the present disclosure. The memory stores instructions executable by the at least one processor to enable the at least one processor to execute the method for identifying the chip components and the method for training the model provided by the present disclosure. The non-transitory computer-readable storage medium of the present disclosure stores computer instructions for causing a computer to execute the method of identifying a chip component and the method of model training provided by the present disclosure.

The memory 1002 is used as a non-transitory computer readable storage medium, and may be used to store non-transitory software programs, non-transitory computer executable programs, and modules, such as program instructions/modules corresponding to the method for identifying chip components and the method for training a model in the embodiments of the present disclosure. The processor 1001 executes a non-transitory software program, instructions, and modules stored in the memory 1002 to perform various functional applications and data processing of the server, that is, to implement the method for identifying chip components and the method for training a model in the above-described method embodiment.

Memory 1002 may include a storage program area that may store an operating system, at least one application program required for functionality, and a storage data area; the storage data area may store data created according to the identification method of the chip components, the use of the electronic device of the model training method, and the like. In addition, the memory 1002 may include high-speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory 1002 may optionally include memory remotely located with respect to processor 1001, which may be connected to the electronic device of the identification method, model training method of the chip components via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.

The electronic device of the chip component identification method and the model training method may further include: an input device 1003 and an output device 1004. The processor 1001, memory 1002, input device 1003, and output device 1004 may be connected by a bus or other means, for example by a bus connection in fig. 10.

The input device 1003 may receive input numeric or character information and generate key signal inputs related to the recognition method of the chip components, user settings of the electronic device of the model training method, and function control, such as a touch screen, a keypad, a mouse, a track pad, a touch pad, a pointer stick, one or more mouse buttons, a track ball, a joystick, and the like. The output means 1004 may include a display device, auxiliary lighting means (e.g., LEDs), tactile feedback means (e.g., vibration motors), and the like. The display device may include, but is not limited to, a Liquid Crystal Display (LCD), a Light Emitting Diode (LED) display, and a plasma display. In some implementations, the display device may be a touch screen.

Various implementations of the systems and techniques described here can be realized in digital electronic circuitry, integrated circuitry, application specific ASIC (application specific integrated circuit), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs, the one or more computer programs may be executed and/or interpreted on a programmable system including at least one programmable processor, which may be a special purpose or general-purpose programmable processor, that may receive data and instructions from, and transmit data and instructions to, a storage system, at least one input device, and at least one output device.

These computing programs (also referred to as programs, software applications, or code) include machine instructions for a programmable processor, and may be implemented in a high-level procedural and/or object-oriented programming language, and/or in assembly/machine language. As used herein, the terms "machine-readable medium" and "computer-readable medium" refer to any computer program product, apparatus, and/or device (e.g., magnetic discs, optical disks, memory, programmable Logic Devices (PLDs)) used to provide machine instructions and/or data to a programmable processor, including a machine-readable medium that receives machine instructions as a machine-readable signal. The term "machine-readable signal" refers to any signal used to provide machine instructions and/or data to a programmable processor.

To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and pointing device (e.g., a mouse or trackball) by which a user can provide input to the computer. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic input, speech input, or tactile input.

The systems and techniques described here can be implemented in a computing system that includes a background component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such background, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), wide Area Networks (WANs), and the internet.

The computer system may include a client and a server. The client and server are typically remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other.

Other embodiments of the disclosure will be apparent to those skilled in the art from consideration of the specification and practice of the disclosure disclosed herein. This disclosure is intended to cover any adaptations, uses, or adaptations of the disclosure following the general principles of the disclosure and including such departures from the present disclosure as come within known or customary practice within the art to which the disclosure pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the disclosure being indicated by the following claims.

It is to be understood that the present disclosure is not limited to the precise arrangements and instrumentalities shown in the drawings, and that various modifications and changes may be effected without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A method for identifying chip components includes:

2. The method of claim 1, wherein if the polarity attribute is non-polarity, performing recognition processing on the chip component to be recognized according to the invoked recognition method to obtain a recognition result, including:

acquiring an image to be identified of the chip component to be identified;

3. The method according to claim 2, wherein if the image to be identified is an inverted image, performing identification processing on the inverted image based on a preset mapping relationship to obtain the identification result, including:

4. A method according to claim 3, wherein determining the recognition result from the determined character corresponding to the reversed descriptive text comprises:

5. The method according to any one of claims 1-4, wherein if the polarity attribute is non-polar, invoking an identification method for identifying the chip component to be identified according to the polarity attribute, including: invoking a nonpolar recognition model;

6. The method of claim 1, wherein if the polarity attribute is polarity, performing recognition processing on the chip component to be recognized according to the invoked recognition method to obtain a recognition result, including:

acquiring an image to be identified of the chip component to be identified;

7. The method of claim 1, wherein if the polarity attribute is polar, invoking an identification method for identifying the chip component to be identified according to the polarity attribute, including: invoking a polarity recognition model;

8. The method of claim 7, wherein the polar recognition model is image-enhanced on the second sample image and is trained based on the second sample image and the enhanced image.

9. The method of claim 7, wherein the polarity recognition model is obtained by performing a random masking process on the second sample image, obtaining a masked image, and training based on the second sample image and the masked image.

10. The method of claim 9, wherein the polarity recognition model is obtained by performing a random masking process on the second sample image to obtain a masked region, performing a blurring process on the masked region to obtain a masked image, and training based on the second sample image and the masked image, wherein the blurring process includes a gaussian blurring process and/or a median blurring process.

11. A model training method, comprising:

12. The method of claim 11, wherein the first sample image comprises descriptive text for a chip component that is nonpolar; training according to the first sample data set to obtain a nonpolar recognition model, wherein the nonpolar recognition model comprises the following steps:

13. The method of claim 12, wherein training the nonpolar recognition model based on the first sample image and the mapping relationship comprises:

14. The method of claim 13, wherein training the base network model to obtain the non-polar recognition model based on the predicted character and the mapping relationship comprises:

15. The method of any of claims 11-14, wherein obtaining a first sample dataset comprises:

acquiring an image of a nonpolar chip component;

16. An identification device for chip components, comprising:

17. A model training apparatus comprising:

18. An electronic device, comprising:

at least one processor; and

a memory communicatively coupled to the at least one processor; wherein,

The memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-10; or to enable the at least one processor to perform the method of any one of claims 11-15.

19. A non-transitory computer readable storage medium storing computer instructions for causing the computer to perform the method of any one of claims 1-10; alternatively, the computer instructions are for causing the computer to perform the method according to any one of claims 11-15.

20. A computer program product comprising a computer program which, when executed by a processor, implements the steps of the method of any of claims 1-10; alternatively, the computer program, when executed by a processor, implements the steps of the method of any of claims 11-15.