WO2022027816A1

WO2022027816A1 - Detection method for image acquisition apparatus, and related apparatus

Info

Publication number: WO2022027816A1
Application number: PCT/CN2020/120043
Authority: WO
Inventors: 叶飞; 何帆
Original assignee: 深圳回收宝科技有限公司
Priority date: 2020-08-03
Filing date: 2020-10-09
Publication date: 2022-02-10
Also published as: CN111935480B; CN111935480A

Abstract

A detection method for an image acquisition apparatus, and a related apparatus. The detection method comprises: acquiring multiple pieces of anomalous image information, wherein each of the multiple pieces of anomalous image information comprises a first preset anomaly (S101); performing anomaly determination on the first preset anomaly in the anomalous image information, and establishing an anomaly model of the first preset anomaly according to a determination result of the anomaly determination (S102); and according to the anomaly model, performing anomaly analysis on a first image acquired by an image acquisition apparatus (S103). By means of the method, the first preset anomaly in the first image acquired by the image acquisition apparatus can be automatically detected and identified.

Description

A detection method and related device for an image acquisition device

technical field

The present application relates to the technical field of abnormality detection of electronic products, and in particular, to a detection method for an image acquisition device and a related device.

Background technique

In the prior art, when abnormal detection is performed on the front camera or rear camera of an electronic product such as a mobile phone, it is usually necessary to use the front camera or rear camera of the electronic product to take pictures, and then the inspectors manually analyze the pictures obtained after taking the pictures. ;

technical problem

In addition, due to the defects in the photos taken by the front and rear cameras of mobile phones and other electronic products, the defects are large, the shape is uncertain, the position is uncertain, the color depth is different, and the background is different. Therefore, using traditional image processing and recognition technology requires manual setting of hundreds of thresholds, and can only deal with a single fixed defect of the scene and defect, and it is difficult to identify defects with huge differences in shape, position, background, number, color, and depth. . Furthermore, when the defect changes or the production line is changed, the traditional image processing and recognition technology needs to reset the threshold, which often requires expert algorithm engineers to spend a lot of time debugging.

technical solutions

The present application provides a detection method for an image acquisition device and a related device to solve the above technical problems.

In order to solve the above technical problems, a technical solution adopted in this application is to provide a detection method for an image acquisition device, the detection method comprising:

acquiring a plurality of abnormal image information, wherein the plurality of abnormal image information includes a first preset abnormality;

Performing an abnormality judgment on the first preset abnormality in the abnormal image information, and establishing an abnormality model of the first preset abnormality according to the judgment result of the abnormality judgment;

An abnormality analysis is performed on the first image acquired by the image acquisition device according to the abnormality model.

Optionally, the step of performing abnormality determination on the first preset abnormality in the abnormal image information includes:

For at least one of the position of the first preset abnormality, the size of the first preset abnormality, the number of the first preset abnormality, and the degree of abnormality of the first preset abnormality in the abnormal image information are marked.

Optionally, the step of marking the abnormality degree of the first preset abnormality includes:

The first predetermined abnormality is scored according to the severity of the first predetermined abnormality.

Optionally, the step of establishing the abnormality model of the first preset abnormality according to the judgment result of the abnormality judgment includes:

A plurality of All the first preset abnormalities in the abnormal image information are divided into a plurality of abnormal levels;

An abnormality model is established, the abnormality model includes the plurality of abnormality levels, and each of the abnormality levels corresponds to at least one image information corresponding to the first predetermined abnormality.

Optionally, the step of performing anomaly analysis on the first image acquired by the image acquisition device according to the anomaly model includes:

The abnormality level of the first image is determined by matching the first image acquired by the image acquisition device with the abnormality image information corresponding to the abnormality model.

Optionally, after the step of performing abnormality judgment on the first preset abnormality in the abnormal image information, and after the abnormality model of the first preset abnormality is established according to the judgment result of the abnormality judgment Before the step of analyzing, the detection method also includes:

Perform enhancement processing on the first preset abnormality in each of the abnormal image information; so as to improve the recognition degree of the first preset abnormality after the enhancement processing.

Optionally, the enhanced processing method includes:

increasing the contrast between an area corresponding to the first preset anomaly and an area other than the first preset anomaly in the abnormal image information; or

Dyeing is performed on the region corresponding to the first preset abnormality in the abnormal image information.

Optionally, the first preset abnormality includes black spot abnormality.

In order to solve the above technical problem, a technical solution adopted in the present application is to provide a detection device for an image acquisition device, characterized in that the detection device includes a processor and a memory; the memory is used for storing the processing The computer program executed by the processor and the intermediate data generated when the computer program is executed; when the processor executes the computer program, it is used to implement the detection method described above.

In order to solve the above technical problems, a technical solution adopted in the present application is to provide a computer-readable storage medium, characterized in that the computer-readable storage medium stores program data, and the program data can be executed to realize the The detection method described above.

beneficial effect

The beneficial effects of the present application are as follows: the present application provides a detection method for an image acquisition device and a related device. By using a plurality of first preset abnormalities in the plurality of abnormal image information to perform abnormality determination, the abnormality model of the first abnormality can be preset, and then the first image obtained by the image acquisition device can be compared with the first preset abnormality model in the abnormality model. Anomalies are matched, so that the first image and all the first preset abnormal areas in the first image can be quickly and automatically identified, and the location, size and abnormal score can be marked, thereby improving the accuracy of the first image acquired by the image acquisition device. Anomaly detection efficiency, and robust detection effect for the first preset anomalies with different shapes, sizes, colors, backgrounds, and numbers; at the same time, the use of deep learning technology can further train the anomaly models of the first preset anomalies , so that the adaptability to the first preset abnormality detection of different sizes, shapes, colors, backgrounds, numbers and degrees of abnormality can be improved; further, by using the detection method provided by the present application, the first preset abnormality detection method obtained by the image acquisition device can be used in the follow-up. When an image is detected, the first preset anomalies with different anomaly levels in the model can be used to match the first preset anomalies in the first image, so that each of the first preset anomalies in the first image can be automatically detected. Perform location labeling, area calculations, and anomaly scoring, reducing reliance on technicians.

Description of drawings

In order to illustrate the technical solutions in the embodiments of the present application more clearly, the following briefly introduces the drawings that are used in the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, under the premise of no creative work, other drawings can also be obtained from these drawings, wherein:

FIG. 1 is a schematic flowchart of an embodiment of a detection method for an image acquisition device provided by the present application;

FIG. 2 is a schematic structural diagram of an embodiment of a detection device for an image acquisition device provided by the present application;

FIG. 3 is a schematic structural diagram of an embodiment of a computer-readable storage medium provided by the present application.

Embodiments of the present invention

The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, but not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the protection scope of the present application.

Please refer to FIG. 1 . FIG. 1 is a schematic flowchart of an embodiment of a detection method for an image acquisition device provided by the present application.

Wherein, the detection method for the image acquisition device may specifically include the following steps:

S101: Acquire a plurality of abnormal image information, where the plurality of abnormal image information includes a first preset abnormality.

In this embodiment, the detection method of the image acquisition device may be used to detect a first preset abnormality in an image acquired by the image acquisition device, where the first preset abnormality may be a black spot abnormality in abnormal image information such as photos , or it can also be abnormal problems such as white spots or colored spots, which are not limited here.

Wherein, a plurality of abnormal image information having the first preset abnormality may be acquired first. The multiple pieces of abnormal image information may be images obtained respectively by using multiple image obtaining devices.

In this embodiment, the image acquisition apparatus may include a front or rear camera of a mobile terminal such as a mobile phone.

When the detection method is used to detect the front or rear camera of a mobile terminal such as a mobile phone, a plurality of different mobile terminals such as mobile phones can be obtained, and the front or rear camera of each mobile terminal such as a mobile phone can be used to shoot multiple A photograph is taken, and the photograph is the abnormal image information described above. If the abnormal image information has a first preset abnormality, the abnormal image information can be the abnormal image information described above.

Wherein, each abnormal image information may be a picture obtained by photographing a solid-color background by an image obtaining device. The solid-color background may be a white background, a gray background, or a solid-color background of other colors.

S102: Perform an abnormality determination on the first preset abnormality in the abnormal image information, and establish an abnormality model of the first preset abnormality according to the determination result of the abnormality determination.

When step S101 is completed, the steps of step S102 can be continued, and the details are as follows:

After acquiring a plurality of abnormal image information with the first predetermined abnormality, it is necessary to perform abnormality judgment on a predetermined abnormality among the plurality of abnormal image information, and establish the abnormality model of the first predetermined abnormality according to the judgment result of the abnormality judgment .

The abnormality determination for the first preset abnormality in the plurality of abnormal image information may be to perform abnormality determination on the first preset abnormality of each abnormal image information by means of manual determination. Thereby, the position, size, number and degree of abnormality of the region having the first preset abnormality in each abnormal image information can be determined.

Wherein, each first preset abnormality in the abnormal image information may be marked by manual determination. For example, each first preset abnormality in each abnormal image information can be circled by a label box, wherein the label box corresponding to each first preset abnormality can be the same as the corresponding first preset abnormality. The outer contours overlap.

In this step, the labeling frame of each abnormal image information can be formed by the personnel using a labeling pen to draw along the outer contour of the first preset abnormality; Preset abnormal outline drawing is formed.

After marking the first preset abnormality in each abnormal image information, the number of the first preset abnormality in each abnormal image information may be determined according to the number of marked boxes. At the same time, the area or size of the first preset abnormality corresponding to the marked frame can be confirmed according to the percentage of each marked frame in the entire graphic area of the abnormal image information.

In this step, after marking the first preset abnormality in each abnormal image information, the degree of abnormality of each first preset abnormality in each abnormal image information can also be determined.

When the first preset abnormality is a black spot abnormality, the degree of abnormality of the first predetermined abnormality may be determined according to the obvious degree of the black spot. Wherein, each black spot abnormality can be scored according to the obvious degree of black spot. For example, when the black spot abnormality does not exist, the abnormal score corresponding to the abnormal degree can be set to 0 points, and when the black spot abnormality corresponds to When the picture is purely black, the abnormal score of the black spot can be set to 100 points. Therefore, different degrees of black spot anomalies can be scored according to the gray scale of the black spot.

Alternatively, the degree of abnormality of each first preset abnormality in each abnormal image information can also be scored by an experienced judging person according to their judgment experience.

After the abnormality determination of the first preset abnormality in the abnormal image information is completed by using the above method, an abnormality model of the first predetermined abnormality may also be established by the determination result of the abnormality determination.

In this embodiment, a first preset abnormal abnormality model is established according to the judgment result of the abnormality judgment, wherein the abnormality model can be used to detect whether the image acquisition device is qualified, and the specific steps include:

1. At least one of the position of the first preset abnormality, the size of the first preset abnormality, the number of the first preset abnormality, and the degree of abnormality of the first preset abnormality obtained according to the judgment will have the first preset abnormality The abnormal image information is divided into multiple abnormal levels.

For example, a plurality of abnormal image information may be classified into two abnormal levels. That is, the abnormal image information is classified as pass or fail. Wherein, when the overall area of the abnormal image information is not larger than a certain preset value; and when the abnormality scores of all the first preset abnormality degrees in the abnormal image information are lower than a certain preset score, it can indicate that If the abnormal image information is qualified, it means that the image acquisition device corresponding to the abnormal image information can be qualified; otherwise, the abnormal image information is unqualified, that is, the image acquisition device corresponding to the abnormal image information is qualified.

Alternatively, the abnormal image information having the first preset abnormality may also be divided into at least three abnormality levels, for example, the first preset abnormality of the abnormal image information may be divided into minor, moderate, and severe. Similarly, the abnormal image information corresponding to one or more abnormal levels may be set as acceptable; and the abnormal image information corresponding to the remaining abnormal levels may be set as unqualified.

Further, the abnormality score of each first preset abnormality area in the abnormality image information may be weighted, so as to obtain the overall abnormality score corresponding to the abnormality image information.

Specifically, the overall abnormality score of the abnormal image information may be equal to the sum of the result value of the area of each first preset abnormality multiplied by its abnormality score.

Then, a plurality of abnormal image information are classified into different abnormal levels by the overall abnormal score of each abnormal image information. The specific method for dividing the abnormal level of the abnormal image information can be referred to above, which will not be repeated here.

Or, in other embodiments, the first preset abnormality with the most serious first preset abnormality in each abnormal image information may also be used to classify the abnormality level, wherein the most serious first preset abnormality refers to the first predetermined abnormality. Let the abnormal area multiplied by its abnormal score result value is the largest in the abnormal image information.

2. After the abnormal image information is divided into a plurality of abnormal levels, the abnormal model of the first preset abnormality can be established. Wherein, the abnormality model may include at least two abnormality levels as divided above, and each abnormality level may include at least one abnormality image information corresponding to the abnormality level.

In this step, the FPN target detection technology may be used to perform model parameter training on the above-mentioned multiple abnormal image information, so as to establish the above-mentioned abnormal model.

S103: Perform an anomaly analysis on the first image acquired by the image acquisition device according to the anomaly model.

After completing step S102 to establish the abnormality model of the first preset abnormality, the steps of step S103 are continued, and the details are as follows:

An abnormality analysis is performed on the first image acquired by the image acquisition device according to the abnormality model to determine whether the image acquisition device is qualified.

In this step, an abnormality analysis can be performed on each image acquisition device through the abnormal model established above to determine whether the image acquisition device is qualified.

Specifically, the solid-color background as described above can be photographed by an image acquisition device, thereby acquiring the first image. Then, by matching the first image with the abnormal image information corresponding to the abnormal model described above, the abnormal level of the first image is determined. Specifically, when the first preset abnormality in the first image can be matched with abnormal image information corresponding to a certain abnormality level, it is determined that the abnormality level of the first image is the abnormality level corresponding to the abnormal image information. Wherein, if the abnormality level corresponds to qualified, it means that the image acquisition device that captures the first image is qualified; otherwise, it is unqualified.

Further, in this embodiment, a deep learning technology may also be used to further train the abnormal model. For example, when the first image cannot be matched with the abnormal image information corresponding to the abnormal model described above, the position, size, number and degree of normality of the abnormality of the first preset abnormality in the first image can be further carried out. judgment, thereby establishing a new abnormal level corresponding to the first image in the abnormal model model.

Further, in other implementation manners, position labeling, area calculation, and anomaly scoring may also be performed for each first preset anomaly in the first image according to the anomaly model.

In this embodiment, after the step S102 of performing abnormality determination on the first predetermined abnormality in the abnormal image information, and before the step of establishing an abnormality model analysis of the first predetermined abnormality according to the determination result of the abnormality determination, the detection The method further includes: performing enhancement processing on the first predetermined abnormality in each abnormal image information, so as to improve the recognition degree of the first predetermined abnormality after the enhancement processing. The enhancement processing method includes: increasing the contrast between an area corresponding to the first preset abnormality and an area other than the first preset abnormality in the abnormal image information; or increasing the contrast between the area corresponding to the first preset abnormality in the abnormal image information Do dyeing. Therefore, the recognition degree of the first preset abnormality in each abnormal image information can be improved.

At this time, after using the first image acquired by the image acquisition device, the same enhancement processing method as above can also be used to perform enhancement processing on the first preset abnormality in the first image, thereby improving the detection of the first abnormality in the first image. A preset abnormal recognition ability.

Based on the same inventive concept, the present application also proposes a detection device for an image acquisition device.

Please refer to FIG. 2 , which is a schematic structural diagram of an embodiment of a detection device for an image acquisition device provided by the present application.

The detection device 20 includes a processor 210 and a memory 220; the memory 220 is used to store the computer program executed by the processor 210 and the intermediate data generated when the computer program is executed; when the processor 210 executes the computer program, it is used to receive image acquisition The image information of the first image acquired by the device, so as to implement the detection method as described above, so as to perform graphic detection on the first image acquired by the image acquisition device.

Based on the same inventive concept, the present application also proposes a computer-readable storage medium, please refer to FIG. 3 , which is a schematic structural diagram of an embodiment of the computer-readable storage medium provided by the present application. The computer-readable storage medium 30 stores program data 31, and the program data 31 may be a program or an instruction, and the program data can be executed to implement the above coordinate calibration method.

In one embodiment, the computer-readable storage device 30 may be a storage chip in the terminal, a hard disk, or other readable and writable storage tools such as a mobile hard disk, a USB flash drive, an optical disk, or the like, or a server or the like.

In the several embodiments provided by the present invention, it should be understood that the disclosed method and apparatus may be implemented in other manners. For example, the device implementations described above are only illustrative. For example, the division of processors or memories is only a logical function division. In actual implementation, there may be other divisions, such as multiple processors and memories. The functions may be combined or may be integrated into another system, or some features may be omitted, or not implemented. On the other hand, the shown or discussed mutual coupling or direct coupling or connection may be through some interfaces, indirect coupling or connection of devices or units, and may be in electrical, mechanical or other forms.

Units described as separate components may or may not be physically separated, and components shown as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of this implementation manner.

In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically alone, or two or more units may be integrated into one unit. The above-mentioned integrated units may be implemented in the form of hardware, or may be implemented in the form of software functional units.

The integrated unit, if implemented as a software functional unit and sold or used as a stand-alone product, may be stored in a computer-readable storage medium. Based on such understanding, the technical solution of the present invention essentially or the part that contributes to the prior art or the whole or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a computer-readable The storage medium includes several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) or a processor (processor) to execute all or part of the steps of the methods in various embodiments of the present invention. The aforementioned computer-readable storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory). Various media that can store program codes, such as Memory), magnetic disks, or optical disks.

In conclusion, the present application provides a detection method for an image acquisition device and a related device. By using a plurality of first preset abnormalities in the plurality of abnormal image information to perform abnormality determination, the abnormality model of the first abnormality can be preset, and then the first image obtained by the image acquisition device can be compared with the first preset abnormality model in the abnormality model. Anomalies are matched, so that the first image and all the first preset abnormal areas in the first image can be quickly and automatically identified and the location, size and abnormal score can be marked, so as to improve the accuracy of the first image acquired by the image acquisition device. Anomaly detection efficiency, and robust detection effect for the first preset anomalies with different shapes, sizes, colors, backgrounds, and numbers; at the same time, the use of deep learning technology can further train the anomaly models of the first preset anomalies , so that the adaptability to the first preset abnormality detection of different sizes, shapes, colors, backgrounds, numbers and degrees of abnormality can be improved; further, by using the detection method provided by the present application, the first preset abnormality detection method obtained by the image acquisition device can be used in the follow-up. When an image is detected, the first preset anomalies with different anomaly levels in the model can be used to match the first preset anomalies in the first image, so that each of the first preset anomalies in the first image can be automatically detected. Perform location labeling, area calculations, and anomaly scoring, reducing reliance on technicians.

The above descriptions are only the embodiments of the present application, and are not intended to limit the scope of the patent of the present application. Any equivalent structure or equivalent process transformation made by using the contents of the description and drawings of the present application, or directly or indirectly applied to other related technologies Fields are similarly included within the scope of patent protection of this application.

Claims

A detection method for an image acquisition device, characterized in that the detection method comprises:

acquiring a plurality of abnormal image information, wherein the plurality of abnormal image information includes a first preset abnormality;

Performing an abnormality judgment on the first preset abnormality in the abnormal image information, and establishing an abnormality model of the first preset abnormality according to the judgment result of the abnormality judgment;

An abnormality analysis is performed on the first image acquired by the image acquisition device according to the abnormality model.
The detection method according to claim 1, wherein the step of performing abnormality determination on the first preset abnormality in the abnormal image information comprises:

For at least one of the position of the first preset abnormality, the size of the first preset abnormality, the number of the first preset abnormality, and the degree of abnormality of the first preset abnormality in the abnormal image information are marked.
The detection method according to claim 2, wherein the step of marking the abnormality degree of the first preset abnormality comprises:

The first predetermined abnormality is scored according to the severity of the first predetermined abnormality.
The detection method according to claim 2, wherein the step of establishing the abnormality model of the first preset abnormality according to the judgment result of the abnormality judgment comprises:

A plurality of All the first preset abnormalities in the abnormal image information are divided into a plurality of abnormal levels;

An abnormality model is established, the abnormality model includes the plurality of abnormality levels, and each of the abnormality levels corresponds to at least one image information corresponding to the first predetermined abnormality.
The detection method according to claim 4, wherein the step of performing anomaly analysis on the first image acquired by the image acquisition device according to the anomaly model comprises:

The abnormality level of the first image is determined by matching the first image acquired by the image acquisition device with the abnormality image information corresponding to the abnormality model.
The detection method according to any one of claims 1-5, characterized in that after the step of performing abnormality determination on the first preset abnormality in the abnormal image information, and after the abnormality is determined according to the abnormality Before the step of establishing the abnormal model analysis of the first preset abnormality based on the judgment result of the judgment, the detection method further includes:

Perform enhancement processing on the first preset abnormality in each of the abnormal image information; so as to improve the recognition degree of the first preset abnormality after the enhancement processing.
The detection method according to claim 6, wherein the enhancement processing method comprises:

increasing the contrast between an area corresponding to the first preset anomaly and an area other than the first preset anomaly in the abnormal image information; or

Dyeing is performed on the region corresponding to the first preset abnormality in the abnormal image information.
The detection method according to claim 6, wherein the first preset abnormality comprises a black spot abnormality.
A detection device for an image acquisition device, characterized in that the detection device includes a processor and a memory; the memory is used to store a computer program executed by the processor and a computer program generated when the computer program is executed. Intermediate data; when the processor executes the computer program, it is used to implement the detection method according to any one of claims 1-8.
A computer-readable storage medium, characterized in that, the computer-readable storage medium stores program data, and the program data can be executed to implement the detection method according to any one of claims 1-8.