Lens detection method and device, computer equipment and storage medium

Abstract

The invention belongs to the technical field of optical lens detection, and relates to a lens detection method and device, computer equipment and a storage medium, and the method comprises the steps: detecting a lens based on a plurality of detection points of the lens, and obtaining a plurality of spectral images; acquiring position information of each detection point, acquiring a standard image according to the position information, matching the standard image with the spectral image, and determining defect information of the lens; predicting the performance of the lens according to the spectral image and the defect information, and determining performance prediction data of the lens; and generating a detection report of the lens according to the performance prediction data. The production efficiency and the product yield are improved, and the detection efficiency of the detection process is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of optical lens detection, and in particular to a lens detection method and device, a computer device and a storage medium. BACKGROUND

[0002] With the rapid development of the mobile intelligent terminal industry, the application of optical lenses such as composite plate lenses and glass lenses is becoming more and more widespread, and the quality of the lenses directly affects the user experience of the terminal product, and the industry has an increasingly urgent demand for efficient and accurate lens quality control throughout the entire process.

[0003] At present, the quality control of optical lenses in the industry mainly adopts a separate detection mode for each module, and special equipment is configured for different detection needs and implemented step by step. For example, visual detection equipment is used to check surface scratches, edge collapse and other appearance defects, composition analysis instruments are used to verify the material composition of the substrate and the coating, and optical performance testers are used to detect key indicators such as transmittance and hardness, and each detection link is relatively independent.

[0004] However, the detection dimensions of the prior art are mutually fragmented, and different equipment needs to be adjusted separately and the process needs to be connected manually, which not only increases the equipment investment and labor costs, but also leads to a long detection cycle and data gaps in the process connection, making it impossible to achieve continuous detection throughout the entire process, ultimately resulting in low detection efficiency and difficulty in adapting to the efficient detection needs of lens mass production.

[0005] Therefore, the present application is proposed. SUMMARY

[0006] The purpose of the embodiments of the present application is to provide a lens detection method, device, computer device and storage medium to solve the technical problem of low detection efficiency in the lens detection process.

[0007] To solve the above technical problems, the embodiments of the present application provide a lens detection method, which adopts the following technical solution, comprising: detecting a lens based on a plurality of detection points of the lens to obtain a plurality of spectral images; obtaining position information of each detection point, obtaining a standard image according to the position information, matching the standard image with the spectral images, and determining defect information of the lens; predicting the performance of the lens according to the spectral images and the defect information, and determining performance prediction data of the lens; generating a detection report of the lens according to the performance prediction data.

[0008] Further, the obtaining of the standard image according to the position information comprises: obtaining a process label of the lens, and screening a defect type to be matched from a plurality of preset defect types according to the process label; determining the standard image according to the defect type to be matched and the position information.

[0009] Further, the defect type to be matched has a plurality of types, and the matching of the standard image with the spectral image to determine the defect information of the lens comprises: obtaining a plurality of weights corresponding to the defect type to be matched; determining a matching order of the standard image according to the weights; matching the standard image with the spectral image according to the matching order to determine the defect information.

[0010] Further, the matching of the standard image with the spectral image according to the matching order to determine the defect information comprises: determining a similarity between the standard image and the spectral image according to the matching order; determining the defect information existing in the spectral image according to the similarity.

[0011] Further, the performance prediction of the lens according to the spectral image and the defect information comprises: extracting a component spectral feature from the spectral image; screening material information corresponding to the component spectral feature from a preset material feature library; fusing the material information, the defect information and the spectral image to obtain fused feature data; predicting the performance of the lens according to the fused feature data to obtain performance prediction data.

[0012] Further, the performance prediction data comprises optical performance data and mechanical performance data, and the prediction of the performance of the lens according to the fused feature data to obtain the performance prediction data comprises: obtaining a first mapping relationship between the fused feature data and an optical performance index, and mapping the fused feature data according to the first mapping relationship to obtain the optical performance data; obtaining a second mapping relationship between the fused feature data and a mechanical performance index and a process parameter of the lens, and mapping the fused feature data according to the second mapping relationship and the process parameter to obtain the mechanical performance data.

[0013] Further, after the obtaining the basic information of the lens, and generating the detection report of the lens according to the basic information and the performance prediction data, the method further comprises: extracting detection data of unqualified items in the detection report; generating process adjustment parameters according to the detection data; adjusting a production process corresponding to the unqualified items according to the process adjustment parameters.

[0014] To solve the above technical problems, the embodiment of the application further provides a lens detection device, which adopts the technical scheme as follows: A lens detection device comprises: a detection module configured to detect a lens based on a plurality of detection points of the lens to obtain a plurality of spectral images; a matching module configured to obtain position information of each detection point, obtain a standard image according to the position information, match the standard image with the spectral images, and determine defect information of the lens; a prediction module configured to predict a performance of the lens according to the spectral images and the defect information, and determine performance prediction data of the lens; a generation module configured to generate a detection report of the lens according to the performance prediction data.

[0015] To solve the above technical problems, the embodiment of the application further provides a computer device, which adopts the technical scheme as follows: A computer device comprises a memory and a processor, the memory stores computer readable instructions, and the processor executes the computer readable instructions to realize the steps of the lens detection method as described above.

[0016] To solve the above technical problems, the embodiment of the application further provides a computer readable storage medium, which adopts the technical scheme as follows: A computer readable storage medium stores computer readable instructions, and the computer readable instructions are executed by a processor to realize the steps of the lens detection method as described above.

[0017] Compared with the prior art, the application has the following beneficial effects: The lens detection method disclosed in the application predicts the performance of the lens according to the spectral image and the defect information. The performance prediction data of the lens is based on the acquisition of the multi-spectral image of the lens at multiple detection points. Different parts of the lens, such as the front and side, can be fully covered, and defects caused by a single detection point can be avoided, laying a complete data foundation for subsequent accurate detection. The corresponding standard image is matched in combination with the detection point position information, which can distinguish the defect characteristics of different positions and greatly improve the accuracy and recognition of defect identification, ensuring accurate defect information determination. The multi-spectral data extracted from the spectral image and the defect information are used to jointly predict the performance of the lens, which can predict whether the lens meets the standard in advance, effectively reducing the waste of resources caused by the flow of unqualified products into subsequent processes. The performance prediction data of the lens is integrated to generate a detection report, which makes the detection results complete and practical. It is convenient to intuitively view the product state, and provides clear data support for production traceability and process optimization, helping to improve production efficiency and product yield, and at the same time, improving the detection efficiency of the detection process. BRIEF DESCRIPTION OF DRAWINGS

[0018] In order to more clearly illustrate the schemes in the application, the following will briefly introduce the drawings needed to be used in the embodiments of the application. Obviously, the drawings in the following description are some embodiments of the application, and other drawings can be obtained by those skilled in the art without creating laborious work.

[0019] Figure 1 is an exemplary system architecture diagram to which the application can be applied; Figure 2 is a flowchart of one embodiment of the lens detection method according to the application; Figure 3 is a structural schematic diagram of one embodiment of the lens detection device according to the application; Figure 4 is a structural schematic diagram of one embodiment of the computer device according to the application. DETAILED DESCRIPTION

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the application belongs; the terms used herein in the specification of the application are only for the purpose of describing specific embodiments and are not intended to limit the application; the specification and claims of the application and the above description of drawings The terms "include" and "have" and any variations thereof in the specification and claims of the above drawings are intended to cover non-exclusive inclusion. The terms "first", "second" and the like in the specification and claims of the application or the above description of drawings are used to distinguish different objects, not to describe a specific order.

[0021] Reference to“an embodiment” herein means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the application. The appearances of the phrase“in an embodiment” in various places in the specification are not necessarily referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. As will be apparent to those of ordinary skill in the art, embodiments described herein can be combined with other embodiments.

[0022] In order to make the person skilled in the art better understand the scheme of the present application, the technical solutions in the embodiments of the present application will be described clearly and completely in conjunction with the drawings below.

[0023] As shown in Figure 1 The system architecture 100 can include a first terminal device 101, a second terminal device 102, a third terminal device 103, a network 104, and a server 105. The network 104 is a medium for providing a communication link between the first terminal device 101, the second terminal device 102, the third terminal device 103, and the server 105. The network 104 can include various connection types, such as wired, wireless communication links, or optical fiber cables, etc.

[0024] A user can use the first terminal device 101, the second terminal device 102, and the third terminal device 103 to interact with the server 105 through the network 104 to receive or send messages, etc. Various communication client applications can be installed on the first terminal device 101, the second terminal device 102, and the third terminal device 103, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social platform software, etc.

[0025] The first terminal device 101, the second terminal device 102, and the third terminal device 103 can be various electronic devices with display screens and supporting web browsing, including but not limited to smartphones, tablet computers, e-book readers, MP3 (Moving Picture Experts Group Audio Layer III) players, MP4 (Moving Picture Experts Group Audio Layer IV) players, laptop computers, and desktop computers, etc.

[0026] The server 105 can be a server that provides various services, such as a background server that provides support for pages displayed on the first terminal device 101, the second terminal device 102, and the third terminal device 103.

[0027] It should be noted that the lens detection method provided in the embodiments of the present application is generally executed by a terminal device, and correspondingly, the lens detection apparatus is generally arranged in the terminal device.

[0028] It should be understood that Figure 1 The number of terminal devices, networks and servers in

[0029] With reference to Figure 2 , a flow chart of one embodiment of the lens detection method according to the present application is shown. The lens detection method includes the following steps: In step S201, the lens is detected based on a plurality of detection points of the lens to obtain a plurality of spectral images.

[0030] In the present embodiment, the electronic device (for example, the terminal device shown in Figure 1 The electronic device on which the lens detection method runs (for example, the terminal device shown in

[0031] In the embodiment, the detection points are a plurality of positions preset on the lens, covering key areas of the lens, such as the lens center, the four-corner area, the lens side, and the like; the lens center and the four corners can cover the distribution area of the sunburst (CD line), the silk screen edge, or the non-conductive vacuum metallization (NCVM) coating area; the side can cover the interlayer bonding surface and the side coating area. The spectral image is a lens image carrying spectral data, which can be obtained by a 400-2500 nm waveband multispectral camera. Specifically, for each detection point, the multispectral camera can be used to collect the full-waveband spectral image, and the push-broom or snapshot imaging mode can be used to complete the collection. The push-broom refers to that the imaging lens scans along the lens surface row by row, and the full-waveband spectral data of all pixel points in each row is recorded, and finally the one-dimensional spectral data is superimposed and spliced with the spatial coordinates of the lens surface to obtain a complete image, that is, the spectral image. The snapshot refers to that the spectral information of the entire lens area is captured at one time through an optical element, and the collection efficiency is higher, which is suitable for batch detection scenarios. During the spectral image collection process, each spatial pixel point synchronously records the reflectivity, absorbance and other optical parameters at the corresponding wavelength. The original collected data is processed, such as noise reduction (eliminating dark current and environmental light interference), calibration (calibrating the reflectivity baseline with a standard white plate), and the like; then the spatial coordinates and one-dimensional spectral coordinates are integrated and reconstructed into a three-dimensional data cube, that is, a complete lens spectral image; finally, the pseudo-color visual image can be converted according to the requirements, which is convenient for intuitive observation of defects and material differences. For example, based on the full-waveband collected data, a spectral intensity matrix image in the form of a cube, that is, a spectral image, is generated, which contains three-dimensional data of spatial position and spectral characteristics. The spectral image is a collection of spectral data, and each spatial pixel point of the spectral image corresponds to a complete set of spectral data, and the spectral image in the defect identification waveband (400-800 nm) can capture the appearance defect characteristics such as surface scratches and silk screen deviation.

[0032] In step S202, the position information of each detection point is obtained, a standard image is obtained according to the position information, the standard image is matched with the spectral image, and the defect information of the lens is determined.

[0033] In the embodiment, the position information refers to the coordinate identification and area label of each detection point. The standard image refers to the standard image of all types of defects of each standard position of the lens, for example, the spectral image of the local part of the lens carrying defect information, which covers common defects such as scratches, CD line breaks, coating unevenness, printing deviation, edge collapse, etc. According to the above position information, a complete set of defect standard images corresponding to each detection point is obtained, for example, for the front four corners 1 (X: 30mm, Y: 30mm), the scratch standard image, the CD line break standard image, the coating unevenness standard image and other defect standard images corresponding to the position are called. The Scale-Invariant Feature Transform (SIFT) algorithm, the Speeded Up Robust Features (SURF) algorithm or the Binary Robust Invariant Scalable Keypoints (BRISK) algorithm can be used to compare the spectral image of each detection point with the standard image of the corresponding position one by one, calculate the feature point matching similarity, and thus determine the defect information of the lens, wherein the defect information includes defect type, defect position, defect quantization parameter (used to describe defect size), defect severity level and defect cause analysis, etc.

[0034] Step S203, according to the spectral image and the defect information, the performance of the lens is predicted, and the performance prediction data of the lens is determined.

[0035] In the embodiment, multispectral data is extracted from the spectral image. For example, near-infrared spectral curves of 800-1500 nm are extracted from the spectral image to detect the substrate composition and the interlayer state; far-infrared spectral features of 1500-2500 nm are extracted to correlate the transmittance and other performances; and near-infrared core spectral data of 800-1100 nm are extracted to predict the infrared cutoff depth. The multispectral data extracted from the spectral image has eliminated system errors, wavelet transform denoising, and standardization to the [0, 1] interval, and has eliminated invalid data generated in the collection process. The performance of the lens is predicted according to the spectral image and the defect information, and performance prediction data of the lens is determined, wherein the performance prediction can be realized by using a gradient boosting tree model, a convolutional neural network and a multi-layer perceptron fusion model, or a random forest model. Specifically, the multispectral data and the defect information are synchronously input into the performance prediction model. The model extracts information related to the optical performance and mechanical performance of the lens through the characteristic signals of each waveband in the multispectral data, combines the defect type, severity level and other key contents in the defect information, analyzes the potential impact of the defect on the performance, and infers the performance parameters of the lens according to the mapping relationship pre-trained in the model. The performance prediction data of the lens includes transmittance, reflectance and special optical indicators suitable for the lens function, such as the cutoff depth OD value of an infrared cutoff filter (IR-Cut) lens; and mechanical parameters such as lens surface hardness and steel wool wear resistance; and the conclusion of whether the lens is qualified is output.

[0036] In step S204, a detection report of the lens is generated according to the performance prediction data.

[0037] In the embodiment, the basic information of the lens and the performance prediction data are combined to form a detection report. The basic information includes the ID of the lens, the lens type, the process label and the detection process. The detection report includes the lens basic information, the material state, the defect condition, the performance prediction result, the qualification determination, the cause analysis and the treatment suggestion. At the same time, the detection report is output in two versions of editable documents and structured data. The editable document version is used for production quality inspection archiving and process personnel viewing, and the structured data version is directly embedded into the manufacturing execution system of the factory to support the linkage and traceability of defect data and performance data with the production process, and to provide complete data support for subsequent process parameter optimization and model iteration. The application can collect multispectral images of multiple detection points of the lens, can comprehensively cover different parts such as the front and side of the lens, and avoid defects missing caused by a single detection point, thereby laying a complete data foundation for subsequent accurate detection; the corresponding standard images are matched according to the position information of the detection points, the defect characteristics of different positions can be distinguished, the accuracy and recognition of defect identification are greatly improved, and the defect information is accurately determined; multispectral data and defect information are extracted from the spectral images to jointly predict the performance of the lens, whether the lens meets the standard can be predicted in advance, and resource waste caused by unqualified products flowing into subsequent processes is effectively reduced; the performance prediction data of the lens are integrated to generate a detection report, so that the detection result has completeness and practicality, the product state can be intuitively observed, clear data support is provided for production traceability and process optimization, the production efficiency and product yield are improved, and the detection efficiency of the detection process is improved.

[0038] In some optional implementations of the embodiment, the step of obtaining the standard image according to the position information comprises: obtaining a process label of the lens, and screening a to-be-matched defect type from a plurality of preset defect types according to the process label; determining the standard image according to the to-be-matched defect type and the position information.

[0039] In the embodiment, the process label is an identifier used to record the substrate type, production process combination and main process of the lens. For example, the process label of the composite board material type includes composite board material transfer printing and coating, such as UV transfer printing and NCVM plating process; the special function type includes IR-Cut 25-layer TiO2 / SiO2 film system plating process; or the basic process type such as simple silk printing process and laser cutting process. The preset defect types include CD line breakage, transfer printing texture deviation, transfer printing pattern defect, coating unevenness, silk printing deviation, interlayer bubble and edge collapse. Specifically, if the process label of the lens is a composite board material double transfer printing, transfer printing and superhard coating process, and in combination with the high occurrence of defects of the process, such as CD line breakage and transfer printing texture deviation in the double transfer printing process, pattern defect in the transfer printing process, and uneven coating in the coating process, the defect types irrelevant to the process are removed, such as silk printing deviation and interlayer bubble, and four to-be-matched defect types, CD line breakage, UV transfer printing texture deviation, transfer printing pattern defect and superhard coating unevenness, are screened. According to the four to-be-matched defect types after screening and the position information of each detection point, the standard images of the to-be-matched defect types in the corresponding positions are called, and one to-be-matched defect type can correspond to multiple standard images in one corresponding position.

[0040] The application filters the matched defect types by process labels, eliminates irrelevant defects, and locates the standard image according to the detection point position information, so that the matching is more targeted. The accuracy of defect recognition is improved, invalid comparison is reduced, detection efficiency is optimized, the detection needs of different processes of lenses are adapted, and the integrated detection process is improved.

[0041] In some optional implementations of the embodiment, the to-be-matched defect types are multiple, and the step of matching the standard image with the spectral image to determine the defect information of the lens includes: obtaining weights corresponding to the to-be-matched defect types; determining a matching order of the standard image according to the weights; matching the standard image with the spectral image according to the matching order to determine the defect information.

[0042] In the embodiment, the weights are set based on defect occurrence probability and influence degree on product performance / appearance. The defect occurrence probability can be obtained based on historical production data. For example, CD line breakage occurs frequently in the UV transfer printing process, and super-hard coating unevenness easily occurs in the coating process. The occurrence probability of the two is significantly higher than that of UV transfer printing texture offset and transfer printing pattern defect. Combined with the influence degree of the defects, such as CD line breakage directly damaging the optical texture integrity and super-hard coating unevenness affecting the light transmittance and wear resistance, the weights of the two are higher. The influence range and severity of UV transfer printing texture offset and transfer printing pattern defect are relatively low. Therefore, the weight distribution can be CD line breakage (weight 0.85), super-hard coating unevenness (weight 0.75), UV transfer printing texture offset (weight 0.6), and transfer printing pattern defect (weight 0.5). According to the high and low of the weights, multiple standard images can be sorted, and the images of CD line breakage and super-hard coating unevenness are matched preferentially. Specifically, the SIFT feature point matching algorithm can be used to compare each detection point spectral image with the corresponding to-be-matched defect standard image in the above order. The similarity threshold can be set to 0.85.

[0043] The application sets the weights based on the defect occurrence probability and influence degree, determines the matching order of the standard image according to the weights, and preferentially matches the defects with high occurrence and high influence. Invalid comparison is reduced, detection efficiency is greatly improved, high-occurrence high-influence defects are avoided, the targeting and practicality of defect recognition are enhanced, and the actual detection needs of lens production processes are adapted.

[0044] In some optional implementations of the embodiment, the step of matching the standard image with the spectral image according to the matching order to determine the defect information includes: determining the similarity of the standard image and the spectral image according to the matching order. According to the similarity, determine defect information existing in the spectral image.

[0045] In the embodiment, the standard images are matched in the spectral image according to the matching order. Specifically, first, Gaussian pyramids (multi-scale blur) and difference pyramids (adjacent scale difference) are constructed for the spatial dimension gray images of the standard images to be matched and the spectral image, respectively, local extreme points are searched in different scale layers, and potential feature points with scale invariance are preliminarily screened out. Secondly, curve fitting is performed on the potential extreme points, unstable points with low contrast and weak edge response are removed, the coordinates of the feature points, the scale to which the feature points belong and the spectral band attributes of the corresponding positions are located, and it is ensured that the feature points can correspond to the defect regions or the standard image template features. Then, the gradient direction and spectral feature distribution histogram in the spatial neighborhood range of each feature point are counted, the peak value corresponding direction is selected as the main direction of the feature point (supporting multiple directions), so that the feature point has rotation invariance and adapts to possible image angle deviation in lens detection. Then, around the main direction of the feature point, the gradient information of the 16x16 pixel block in the adjacent region of the feature point is extracted, divided into 4x4 sub-blocks, and the 8 direction gradient histogram of each sub-block is counted. A fusion feature descriptor of 128-dimensional spatial features and N-dimensional key band spectral features is generated, and after standardization, the influence of light change is eliminated to ensure the consistency of the features under the difference of multi-spectral bands. Then, the similarity between the standard image to be matched and the feature descriptor in the spectral image is calculated by using the Euclidean distance, whether the spectral image exists the defect type corresponding to the standard image is determined according to the similarity, if the defect type corresponding to the standard image exists, the defect position, defect quantitative parameter and defect degree of the defect type are extracted, and the defect cause is analyzed combined with the process label to obtain the defect information.

[0046] The present application matches by order, obtains similarity through multi-scale feature point screening, direction calibration and feature descriptor calculation, and determines defects according to the similarity. It not only avoids the interference brought by angle and multi-spectral band difference, but also can completely extract defect position, parameter and cause information, improves recognition reliability, and provides solid support for subsequent performance prediction and process optimization.

[0047] In some optional implementations of the embodiment, the step of predicting the performance of the lens according to the spectral image and the defect information, and determining the performance prediction data of the lens includes: performing feature extraction on the spectral image to obtain component spectral features; screening material information corresponding to the component spectral features in a preset material feature library; fusing the material information, the defect information and the spectral image to obtain fusion feature data; According to the fusion feature data, performance of the lens is predicted to obtain performance prediction data.

[0048] In the embodiment, targeted feature extraction is performed on the spectral image, 16-dimensional component spectral features are extracted from the near-infrared waveband of 800-1500 nm of the spectral image, including parameters such as feature peak position, peak intensity ratio, half-peak width, and the like, which are directly related to the substrate and film system components, and meanwhile, 4-dimensional correlation features such as the absorbance peak spectral slope of the near-infrared core waveband of 800-1100 nm in the spectral image are supplemented, to form a 20-dimensional component spectral feature vector. A preset material feature library stores standard spectral feature templates of common lens materials such as polycarbonate PC, acrylic PMMA, and super-hard coated materials. Material information corresponding to the component spectral features is screened in the preset material feature library: the extracted 20-dimensional component spectral features are compared with the templates in the library in terms of similarity, matching results with a similarity greater than or equal to 0.95 are screened out, and material information is determined, including substrate type, purity, film system components, and layer thickness. For example, the determined material information is a PC / PMMA composite substrate with a purity of 99.99%, a super-hard coated layer with a thickness of 1.2 μm and uniform components, and no bubbles between layers.

[0049] For example, the material information (substrate type, purity, film system components, layer thickness, and the like), defect information (defect type code, offset normalization value, severity level quantization value, and influence coefficient), and multi-spectral data (absorbance and reflectivity mean values of each key waveband) extracted from the spectral image are fused in a feature splicing manner to generate fusion feature data. The fusion feature data is input into a performance prediction model, wherein the model is based on a gradient boosting tree architecture, the model calls a pre-trained mapping relationship library of fusion features and performance, and comprehensively infers performance prediction data of the lens.

[0050] The present application extracts multi-spectral component features, matches material information, and fuses material, defect, and spectral data to form complete features. This not only makes up for the limitations of single data prediction, enriches feature dimensions, but also improves the comprehensiveness and accuracy of performance prediction, adapts to different process lens requirements, and provides reliable basis for subsequent qualification judgment and process optimization.

[0051] In some optional implementations of the embodiment, the performance prediction data includes optical performance data and mechanical performance data, and the step of predicting performance of the lens according to the fusion feature data to obtain performance prediction data includes: obtaining a first mapping relationship between the fusion feature data and an optical performance index, mapping the fusion feature data according to the first mapping relationship to obtain the optical performance data; Obtain the second mapping relationship between the fusion feature data and the mechanical performance index, as well as the process parameters of the lens. Based on the second mapping relationship and the process parameters, map the fusion feature data to obtain the mechanical performance data.

[0052] In this embodiment, optical performance indicators include basic optical indicators, such as transmittance and reflectance, which affect the image clarity and light utilization of the lens, as well as specialized optical indicators, such as the infrared cutoff depth for IR-Cut lenses, which ensures infrared filtering effectiveness. The first mapping relationship refers to the correlation between fused feature data and optical performance indicators. Mechanical performance indicators are quantitative data used to characterize the mechanical resilience of the lens, directly related to its durability, and their prediction error is adapted to industrial testing needs. These include the surface hardness of the lens, which reflects its resistance to scratches and pressure, and the number of steel wool abrasion resistance tests, reflecting the lens surface's resistance to wear and tear. The second mapping relationship is the correlation between fused feature data and mechanical performance indicators. Both the first and second mapping relationships are generated through training with a large number of samples.

[0053] For example, if the substrate purity is ≥99.99%, there are no bubbles / cracks between layers, the average absorbance in the 1500-2500nm performance-related wavelength band is ≤0.08, and the coating uniformity deviation is ≤0.02, it can be inferred that the light transmittance of the lens can reach 99%. If the substrate purity is <99.95% or there are bubbles between layers (diameter ≤0.05mm) and the average absorbance is >0.10, it can be inferred that the light transmittance of the lens reaches 97%, and the light transmittance of the lens does not meet the standard.

[0054] For example, by integrating the substrate purity ≥99.99%, uniform coating composition, absence of coating unevenness defects (defect characteristics), and a wear resistance correlation value ≥0.9 in the 1500-2500nm range, combined with the coating gas ratio (argon:oxygen = 9:1) and curing temperature of 120-130℃ in the process parameters, the wear resistance count of the lens can be projected to meet the standard. If the substrate purity <99.95%, the coating unevenness area ≥2mm², and the correlation characteristic value <0.8, combined with an incorrect coating gas ratio or abnormal curing temperature in the process, the wear resistance count of the lens will not meet the standard. By sequentially mapping the optical performance indicators and mechanical performance indicators, the optical performance data and mechanical performance data can be obtained, i.e., performance prediction data.

[0055] This application derives performance predictions from optical and mechanical properties using a specific mapping relationship, with mechanical property predictions further inferred from process parameters. This approach makes performance predictions more targeted, avoids biases from single mapping methods, and significantly improves data reliability, providing a valid basis for lens qualification assessment and production process optimization.

[0056] In some optional implementations of this embodiment, after the step of generating the lens test report based on the performance prediction data, the method further includes: Extract the test data for the non-compliant items in the test report; Based on the detection data, process adjustment parameters are generated; The production process corresponding to the non-conforming item is adjusted according to the process adjustment parameters.

[0057] In this embodiment, the test data for non-conforming items are extracted from the test report. For example, the abrasion resistance of steel wool in the mechanical properties of the lens does not meet the standard, while other performance indicators meet industry standards. Therefore, relevant test data related to the non-compliance of the abrasion resistance is extracted, such as material characteristic data indicating that the substrate purity does not reach a preset threshold, the compositional uniformity deviation of the superhard coating layer exceeds the standard range, defect characteristic data indicating slight coating unevenness, and process parameter data for gas ratio and curing temperature in the coating process. Based on the relevant processes of these non-conforming items, process adjustment parameters are generated. For example, to address the problem of insufficient substrate purity, the raw material acceptance process parameters can be adjusted, the substrate raw material purity screening threshold can be increased, and a secondary purity verification process can be added before the raw materials are put into storage. To address the problems of uneven coating composition and abnormal gas ratio, the superhard coating process parameters can be corrected, the mixing ratio of argon and oxygen can be optimized, and the gas pressure in the coating chamber can be adjusted to a suitable range. To address the problem of low curing temperature, the temperature parameters of the curing process after superhard coating can be increased, and the isothermal curing time can be extended to ensure a tight bond between the film layer and the substrate. According to the generated process adjustment parameters, the corresponding processes for non-conforming items are adjusted, and the next batch of lenses produced after the adjustment is tested again according to this plan.

[0058] This application identifies the causes of lens defects, generates corresponding process adjustment parameters, and optimizes the corresponding production processes in a targeted manner, avoiding blind adjustments. This not only quickly eliminates performance shortcomings and improves production yield, but also establishes a sound management model for production optimization through inspection, driving continuous process improvement and upgrading.

[0059] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence (AI) refers to the theories, methods, technologies, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0060] Foundational technologies for artificial intelligence generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies mainly encompass computer vision, robotics, biometrics, speech processing, natural language processing, and machine learning / deep learning.

[0061] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by instructing related hardware through computer-readable instructions. These computer-readable instructions can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above methods. The aforementioned storage medium can be a non-volatile storage medium such as a magnetic disk, optical disk, or read-only memory (ROM), or random access memory (RAM).

[0062] It should be understood that although the steps in the flowcharts of the accompanying figures are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the accompanying figures may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0063] Further reference Figure 3 As a response to the above Figure 2 To implement the method shown, this application provides an embodiment of a lens detection device, which is similar to... Figure 2 Corresponding to the method embodiments shown, this device can be specifically applied to various electronic devices.

[0064] like Figure 3 As shown, the lens detection device 300 described in this embodiment includes: a detection module 301, a matching module 302, a prediction module 303, and a generation module 304. Wherein: The detection module 301 is used to detect the lens based on multiple detection points of the lens to obtain multiple spectral images; The matching module 302 is used to acquire the position information of each detection point, acquire a standard image based on the position information, match the standard image with the spectral image, and determine the defect information of the lens; Prediction module 303 is used to predict the performance of the lens based on the spectral image and the defect information, and determine the performance prediction data of the lens; The generation module 304 is used to generate a test report for the lens based on the performance prediction data.

[0065] The lens inspection device provided in this application predicts the performance of the lens based on the spectral image and the defect information. The performance prediction data is determined by collecting multispectral images from multiple inspection points on the lens, comprehensively covering different parts such as the front and sides of the lens, avoiding defect omissions caused by single inspection points, and laying a complete data foundation for subsequent accurate inspection. By combining the inspection point location information with corresponding standard images, it can specifically distinguish defect features at different locations, significantly improving the accuracy and distinctiveness of defect identification, ensuring accurate defect information judgment. Extracting multispectral data from the spectral image and jointly predicting lens performance with defect information allows for early prediction of whether the lens meets standards, effectively reducing resource waste caused by defective products flowing into subsequent processes. Integrating the lens performance prediction data to generate an inspection report ensures that the inspection results are both complete and practical, facilitating intuitive viewing of product status and providing clear data support for production traceability and process optimization, thus helping to improve production efficiency and product yield, while also improving the inspection efficiency of the inspection process.

[0066] In some optional implementations of this embodiment, the matching module 302 is further configured to: Obtain the manufacturing label of the lens, and based on the manufacturing label, filter out the defect type to be matched from multiple preset defect types; The standard image is determined based on the type of defect to be matched and the location information.

[0067] The lens inspection device provided in this application filters suitable defect types for matching by using process labels, eliminates irrelevant defect interference, and then combines the location information of the inspection points to locate standard images, making the matching more targeted. This not only improves the accuracy of defect identification, reduces invalid comparisons, optimizes inspection efficiency, but also adapts to the inspection needs of lenses with different processes, thus perfecting the integrated inspection process.

[0068] In some optional implementations of this embodiment, the matching module 302 is further configured to: Obtain the weights corresponding to the multiple defect types to be matched; The matching order of the standard images is determined based on the weights. The standard image is matched with the spectral image according to the matching order to determine the defect information.

[0069] The lens inspection device provided in this application sets weights based on the probability of defect occurrence and the degree of impact, and determines the standard image matching order according to the weights, prioritizing the matching of high-incidence and high-impact defects. This reduces invalid comparisons, significantly improves inspection efficiency, avoids missing high-incidence and high-impact defects, enhances the pertinence and practicality of defect identification, and adapts to the actual inspection needs of lens manufacturing processes.

[0070] In some optional implementations of this embodiment, the matching module 302 is further configured to: The similarity between the standard image and the spectral image is determined according to the matching order. Based on the similarity, the defect information present in the spectral image is determined.

[0071] The lens inspection device provided in this application obtains similarity through matching sequence comparison, multi-scale feature point screening, orientation calibration, and feature descriptor calculation, and determines defects based on the similarity. This avoids interference caused by differences in angle and multispectral bands, and can completely extract information such as defect location, parameters, and causes, improving the reliability of identification and providing solid support for subsequent performance prediction and process optimization.

[0072] In some optional implementations of this embodiment, the prediction module 303 is further configured to: Feature extraction is performed on the spectral image to obtain component spectral features; Filter material information corresponding to the composition spectral characteristics from a preset material feature library; The material information, the defect information, and the spectral image are fused to obtain fused feature data; Based on the fusion feature data, the performance of the lens is predicted to obtain performance prediction data.

[0073] The lens inspection device provided in this application extracts multispectral component features, matches material information, and fuses material, defect, and spectral data to form complete features. This not only overcomes the limitations of single-data prediction and enriches feature dimensions, but also improves the comprehensiveness and accuracy of performance prediction, adapting to the needs of lenses with different processes and providing a reliable basis for subsequent qualification judgment and process optimization.

[0074] In some optional implementations of this embodiment, the prediction module 303 is further configured to: Obtain a first mapping relationship between the fused feature data and the optical performance index; and map the fused feature data according to the first mapping relationship to obtain the optical performance data. Obtain the second mapping relationship between the fusion feature data and the mechanical performance index, as well as the process parameters of the lens. Based on the second mapping relationship and the process parameters, map the fusion feature data to obtain the mechanical performance data.

[0075] The lens testing device provided in this application derives its performance through optical and mechanical properties using a specific mapping relationship. Mechanical performance prediction is further inferred by combining process parameters. This makes performance prediction more targeted, avoids biases from single mapping methods, and significantly improves data reliability, providing a valid basis for lens qualification and production process optimization.

[0076] In some optional implementations of this embodiment, the generation module 304 is further configured to: Extract the test data for the non-compliant items in the test report; Based on the detection data, process adjustment parameters are generated; The production process corresponding to the non-conforming item is adjusted according to the process adjustment parameters.

[0077] The lens inspection device provided in this application identifies the causes of lens defects, generates corresponding process adjustment parameters, and optimizes the corresponding production processes in a targeted manner, avoiding blind adjustments. It can quickly eliminate performance shortcomings, improve production yield, and establish a positive management model for production optimization through inspection, driving continuous process improvement and upgrading.

[0078] To address the aforementioned technical problems, embodiments of this application also provide a computer device. Please refer to [link / reference needed]. Figure 4 , Figure 4 This is a basic structural block diagram of the computer device in this embodiment.

[0079] The computer device 4 includes a memory 41, a processor 42, and a network interface 43 that are interconnected via a system bus. It should be noted that only the computer device 4 with components 41, 42, and 43 is shown in the figure; however, it should be understood that it is not required to implement all the components shown, and more or fewer components can be implemented alternatively. Those skilled in the art will understand that the computer device described here is a device capable of automatically performing numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes, but is not limited to, microprocessors, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc.

[0080] The computer device can be a desktop computer, laptop, handheld computer, or cloud server, etc. The computer device can interact with the user via a keyboard, mouse, remote control, touchpad, or voice control.

[0081] The memory 41 includes at least one type of readable storage medium, including flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 41 may be an internal storage unit of the computer device 4, such as the hard disk or memory of the computer device 4. In other embodiments, the memory 41 may also be an external storage device of the computer device 4, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the computer device 4. Of course, the memory 41 may include both the internal storage unit and its external storage device of the computer device 4. In this embodiment, the memory 41 is typically used to store the operating system and various application software installed on the computer device 4, such as computer-readable instructions for lens detection methods. In addition, the memory 41 can also be used to temporarily store various types of data that have been output or will be output.

[0082] In some embodiments, the processor 42 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chip. The processor 42 is typically used to control the overall operation of the computer device 4. In this embodiment, the processor 42 is used to execute computer-readable instructions stored in the memory 41 or to process data, for example, to execute computer-readable instructions for the lens detection method.

[0083] The network interface 43 may include a wireless network interface or a wired network interface, which is typically used to establish communication connections between the computer device 4 and other electronic devices.

[0084] The computer equipment provided in this application predicts the performance of the lens based on the spectral image and the defect information. The predicted performance data is determined by collecting multispectral images from multiple detection points on the lens, comprehensively covering different parts such as the front and sides of the lens, avoiding defect omissions caused by single detection points, and laying a complete data foundation for subsequent accurate testing. By combining the detection point location information with corresponding standard images, it can specifically distinguish defect features at different locations, significantly improving the accuracy and distinctiveness of defect identification, ensuring accurate defect information judgment. Extracting multispectral data from the spectral image and jointly predicting lens performance with defect information allows for advance prediction of whether the lens meets standards, effectively reducing resource waste caused by defective products flowing into subsequent processes. Integrating the lens performance prediction data to generate a test report ensures that the test results are both complete and practical, facilitating intuitive viewing of product status and providing clear data support for production traceability and process optimization, thus helping to improve production efficiency and product yield, while also improving the efficiency of the testing process.

[0085] This application also provides another embodiment, namely, providing a computer-readable storage medium storing computer-readable instructions that can be executed by at least one processor to cause the at least one processor to perform the steps of the lens detection method described above.

[0086] The computer-readable storage medium provided in this application predicts the performance of the lens based on the spectral image and the defect information. The performance prediction data is determined by collecting multispectral images from multiple detection points on the lens, comprehensively covering different parts such as the front and sides of the lens, avoiding defect omissions caused by single detection points, and laying a complete data foundation for subsequent accurate testing. By combining the detection point location information with corresponding standard images, it can specifically distinguish defect features at different locations, significantly improving the accuracy and distinctiveness of defect identification, ensuring accurate defect information judgment. Extracting multispectral data from the spectral image and jointly predicting lens performance with defect information allows for advance prediction of whether the lens meets standards, effectively reducing resource waste caused by defective products flowing into subsequent processes. Integrating the lens performance prediction data to generate a test report ensures that the test results are both complete and practical, facilitating intuitive viewing of product status and providing clear data support for production traceability and process optimization, thus helping to improve production efficiency and product yield, while also improving the efficiency of the testing process.

[0087] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0088] Obviously, the embodiments described above are only some embodiments of this application, not all embodiments. The accompanying drawings show preferred embodiments of this application, but do not limit the patent scope of this application. This application can be implemented in many different forms; rather, these embodiments are provided to provide a more thorough and comprehensive understanding of the disclosure of this application. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing specific embodiments, or make equivalent substitutions for some of the technical features. Any equivalent structures made using the content of this application's specification and drawings, directly or indirectly applied to other related technical fields, are similarly within the scope of patent protection of this application.

Claims

1. A method for testing lenses, characterized in that, Includes the following steps: The lens is tested at multiple detection points to obtain multiple spectral images; The location information of each detection point is obtained, a standard image is obtained based on the location information, and the standard image is matched with the spectral image to determine the defect information of the lens; The performance of the lens is predicted based on the spectral image and the defect information, and the performance prediction data of the lens is determined. Based on the performance prediction data, a test report for the lens is generated.

2. The lens testing method according to claim 1, characterized in that, The step of obtaining a standard image based on the location information includes: Obtain the manufacturing label of the lens, and based on the manufacturing label, filter out the defect type to be matched from multiple preset defect types; The standard image is determined based on the type of defect to be matched and the location information.

3. The lens testing method according to claim 2, characterized in that, There are multiple defect types to be matched. Matching the standard image with the spectral image to determine the defect information of the lens includes: Obtain the weights corresponding to the multiple defect types to be matched; The matching order of the standard images is determined based on the weights. The standard image is matched with the spectral image according to the matching order to determine the defect information.

4. The lens testing method according to claim 3, characterized in that, The step of matching the standard image with the spectral image according to the matching order to determine the defect information includes: The similarity between the standard image and the spectral image is determined according to the matching order. Based on the similarity, the defect information present in the spectral image is determined.

5. The lens testing method according to claim 1, characterized in that, The step of predicting the performance of the lens based on the spectral image and the defect information, and determining the performance prediction data of the lens, includes: Feature extraction is performed on the spectral image to obtain component spectral features; Filter material information corresponding to the composition spectral characteristics from a preset material feature library; The material information, the defect information, and the spectral image are fused to obtain fused feature data; Based on the fusion feature data, the performance of the lens is predicted to obtain performance prediction data.

6. The lens testing method according to claim 5, characterized in that, The performance prediction data includes optical performance data and mechanical performance data. The process of predicting the performance of the lens based on the fusion feature data to obtain performance prediction data includes: Obtain a first mapping relationship between the fused feature data and the optical performance index; and map the fused feature data according to the first mapping relationship to obtain the optical performance data. Obtain the second mapping relationship between the fusion feature data and the mechanical performance index, as well as the process parameters of the lens. Based on the second mapping relationship and the process parameters, map the fusion feature data to obtain the mechanical performance data.

7. The lens testing method according to any one of claims 1 to 6, characterized in that, After generating the lens test report based on the performance prediction data, the method further includes: Extract the test data for the non-compliant items in the test report; Based on the detection data, process adjustment parameters are generated; The production process corresponding to the non-conforming item is adjusted according to the process adjustment parameters.

8. A lens testing device, characterized in that, include: The detection module is used to detect the lens based on multiple detection points to obtain multiple spectral images; The matching module is used to obtain the location information of each detection point, obtain a standard image based on the location information, match the standard image with the spectral image, and determine the defect information of the lens; The prediction module is used to predict the performance of the lens based on the spectral image and the defect information, and to determine the performance prediction data of the lens; The generation module is used to generate a test report for the lens based on the performance prediction data.

9. A computer device, characterized in that, The device includes a memory and a processor, wherein the memory stores computer-readable instructions, and the processor executes the computer-readable instructions to implement the steps of the lens detection method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the steps of the lens detection method as described in any one of claims 1 to 7.

Images