Method and device for correcting machining error of optical device

Abstract

The invention provides a processing error correction method and device for an optical device, and the method comprises the steps: carrying out the LED irradiation and detection of a to-be-corrected optical device in a plurality of directions and a plurality of wave bands, so as to obtain initial detection data containing reflection distribution information, transmission distribution information and an interference fringe mode; generating an error distribution diagram indicating coordinate positions and corresponding error types according to the data, and performing grinding treatment, polishing treatment or local film layer repairing treatment on each region to obtain updated data formed by the finished local morphology, the film layer characteristics and the coordinate information of the finished region; lED irradiation and detection are carried out again in the whole surface range, and a comprehensive correction result is obtained. The method realizes timely correction of local deviation while ensuring complete detection of a multilayer film system and a high-precision surface, and is suitable for batch production of various high-end optical devices.

Description

Technical Field

[0001] The present invention relates to the field of error correction, and in particular to a method and device for correcting processing errors of optical devices. Background Art

[0002] Optical devices play an important role in modern industry and scientific research. For example, they are used in imaging lenses, laser systems, spectral analysis equipment and other fields. Generally speaking, the manufacturing process of optical devices includes blank preparation, mechanical grinding, precision polishing and multi-layer thin film coating. Since optical materials usually have high precision requirements on surface shape and film distribution, any small geometric error or film unevenness will have a great impact on the optical performance of the device. For high-end lenses or reflectors, errors at the micron or even nanometer level may cause imaging distortion, increased light loss or wavefront distortion. In order to meet such strict optical performance indicators, it is necessary to detect the surface morphology and film thickness of optical devices in real time during the processing and perform corresponding trimming operations based on the detection results.

[0003] In the traditional processing methods of optical devices, single-wavelength light sources or mechanical contact measurement methods are usually used for detection to identify the surface flatness and film thickness distribution of the device. However, since single-wavelength detection cannot fully reflect the differences in transmission and reflection of multi-layer film systems in different bands, and mechanical contact measurement may cause scratches or microcracks on the optical surface, the existing technology is prone to insufficient detection information and inaccurate error positioning when it comes to multi-band spectral requirements and high-precision film control. Summary of the invention

[0004] The main purpose of the present invention is to solve the technical problems that the existing single wavelength light source or mechanical contact measurement means easily lead to insufficient detection information and inaccurate error positioning in the processing of optical devices; A first aspect of the present invention provides a method for correcting processing errors of an optical device, the method comprising: The optical device to be corrected is subjected to LED irradiation processing in multiple orientations and multiple wavebands to obtain initial detection data containing reflection distribution information, transmission distribution information and interference fringe patterns; According to the initial detection data, the reference parameters of the optical device are divided into regions and compared with each other to obtain an error distribution map indicating the coordinate position and the corresponding error type; According to the error distribution map, grinding, polishing or local film repairing is performed on the corresponding area of ​​the optical device to obtain updated data including the local morphology after repair, film characteristics and coordinate information of the repaired area; According to the updated data, the optical device is again subjected to LED irradiation and detection processing over the entire surface to obtain a comprehensive correction result that marks the overall reflection change, transmission change and fringe change.

[0005] Optionally, in a first implementation of the first aspect of the present invention, the optical device to be corrected is subjected to LED irradiation processing in multiple orientations and multiple wavebands, and the initial detection data containing reflection distribution information, transmission distribution information and interference fringe pattern is obtained, including: According to the appearance characteristics of the optical device to be corrected, multiple LED sub-light sources are arranged in a ring or matrix to obtain a multi-channel illumination array with sub-bands and sub-angles; According to the preset timing control information, the LED sub-light sources of different wavelength bands or different incident angles in the multi-channel illumination array are sequentially illuminated to obtain the reflection image, transmission image and fringe image of the optical device surface in each wavelength band or orientation; The reflection image, transmission image and fringe image are respectively subjected to data separation processing to obtain initial detection data containing reflection distribution information, transmission distribution information and interference fringe pattern.

[0006] Optionally, in a second implementation of the first aspect of the present invention, performing region division and difference comparison processing on reference parameters of the optical device according to the initial detection data to obtain an error distribution diagram indicating coordinate positions and corresponding error types includes: Performing difference calculation processing on the reflection distribution information in the initial detection data and the reference reflection parameters of the optical device to obtain a reflection difference matrix that associates the corresponding optical coordinates with the reflection deviation values; Performing difference calculation processing on the transmission distribution information in the initial detection data and the reference transmission parameters of the optical device to obtain a transmission difference matrix that associates the corresponding optical coordinates with the transmission deviation values; Performing phase resolution processing on the interference fringe pattern in the initial detection data, and performing comparison operation processing in combination with reference fringe parameters to obtain a wavefront offset matrix; Performing multi-dimensional data fusion processing on the reflection difference matrix, the transmission difference matrix and the wavefront deviation matrix to obtain a comprehensive data set including optical coordinates, reflection deviation, transmission deviation and phase change information; The comprehensive data set is processed by partitioning and labeling using a coordinate indexing algorithm to obtain an error distribution map indicating coordinate positions and corresponding error types.

[0007] Optionally, in a third implementation of the first aspect of the present invention, performing phase resolution processing on the interference fringe pattern in the initial detection data and performing comparison operation processing in combination with reference fringe parameters to obtain a wavefront shift matrix includes: Performing pixel-level extraction processing on the fringe peak-valley distribution in the interference fringe pattern to obtain preliminary fringe information including fringe position coordinates and intensity changes; According to the preliminary fringe information, a phase unwrapping process is performed on the fringe image to obtain a phase distribution matrix based on pixel coordinates; Performing difference comparison processing on the phase distribution matrix and the reference fringe parameters to obtain phase difference data with phase offset values ​​associated with fringe pixel coordinates; Performing interpolation fitting processing on the phase difference data to obtain a phase distortion distribution that characterizes a wavefront distortion trend; The phase distortion distribution is combined with the coordinate mapping relationship of the interference image to perform matrix conversion processing to obtain a wavefront shift matrix marked with optical coordinates and phase shift amounts.

[0008] Optionally, in a fourth implementation of the first aspect of the present invention, the partition labeling process is performed on the comprehensive data set using a coordinate index algorithm to obtain an error distribution map indicating coordinate positions and corresponding error types, including: Performing coordinate registration processing on the reflection deviation, transmission deviation and phase change information contained in the comprehensive data set to obtain a multi-dimensional error point cloud indexed by a unified optical coordinate system; According to the multi-dimensional error point cloud, each pixel or sampling point is divided into different threshold values ​​to obtain a plurality of error units showing similar deviation characteristics; Performing coordinate clustering processing on the error unit to obtain a cluster center coordinate and a corresponding deviation information index table; Perform label assignment processing on each error unit according to the deviation information index table to obtain a region label list including optical coordinates and error type labels; Visual mapping is performed on the area mark list to obtain an error distribution map indicating coordinate positions and corresponding error types.

[0009] Optionally, in a fifth implementation of the first aspect of the present invention, the grinding, polishing or local film layer repairing of the corresponding area of ​​the optical device according to the error distribution map to obtain updated data including the local morphology after repair, film layer characteristics and coordinate information of the repaired area includes: Determine the area marked as surface geometric error and the area marked as film thickness anomaly in the error distribution map; Performing a cutting depth calculation process on the area marked as the surface geometric error to obtain a cutting depth parameter for the grinding work, and performing a grinding process or a polishing process on the area of ​​the surface geometric error to obtain temporary topography data after the surface flatness is improved; Performing film layer increase / decrease estimation processing on the area marked as abnormal film layer thickness to obtain film layer control instructions for local deposition or removal, and performing thin film increase / decrease operation processing on the area marked as abnormal film layer to obtain local update data containing compensated film layer thickness information; The temporary morphology data and the local update data are integrated to obtain temporary data including the local morphology after trimming and the film layer characteristics, and the coordinate index information is added to the temporary data and a unified marking process is performed to obtain updated data including the local morphology after trimming, the film layer characteristics and the coordinate information of the trimming area.

[0010] Optionally, in a sixth implementation of the first aspect of the present invention, the optical device is subjected to LED irradiation and detection processing again over the entire surface according to the updated data to obtain a comprehensive correction result that marks the overall reflection change, transmission change, and fringe change, including: According to the trimming area coordinate information in the update data, the entire surface range of the optical device is divided into detection areas to obtain a detection distribution map covering all trimming areas and untrimmed areas; According to the detection distribution diagram, performing sub-band triggering processing on multiple LED sub-light sources to obtain a multi-angle illumination scheme adapted to each coordinate area; Performing full-surface detection processing on the optical device under the multi-angle illumination scheme to obtain secondary sampling data including reflection field, transmission field and interference fringe field information; The secondary sampling data is compared with the trimmed local morphology and film layer features marked in the update data to obtain a comprehensive correction result of the overall reflection change, transmission change and fringe change.

[0011] A second aspect of the present invention provides a processing error correction device for an optical device, the processing error correction device for an optical device comprising: A multi-source illumination module is used to perform LED illumination processing on the optical device to be corrected in multiple orientations and multiple wavebands to obtain initial detection data containing reflection distribution information, transmission distribution information and interference fringe patterns; A difference comparison module, used to perform area division and difference comparison processing on reference parameters of the optical device according to the initial detection data, and obtain an error distribution map indicating coordinate positions and corresponding error types; A local trimming module, used to perform grinding, polishing or local film repairing on the corresponding area of ​​the optical device according to the error distribution map, and obtain updated data including the local morphology after trimming, film characteristics and coordinate information of the trimming area; The global correction module is used to perform LED irradiation and detection processing on the entire surface of the optical device again according to the updated data, so as to obtain a comprehensive correction result that marks the overall reflection change, transmission change and fringe change.

[0012] A third aspect of the present invention provides a processing error correction device for an optical device, comprising: a memory and at least one processor, wherein instructions are stored in the memory, and the memory and the at least one processor are interconnected via lines; the at least one processor calls the instructions in the memory to enable the processing error correction device for the optical device to perform the steps of the above-mentioned processing error correction method for the optical device.

[0013] A fourth aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the steps of the above-mentioned method for correcting processing errors of optical devices.

[0014] The processing error correction method and device of the above optical device adopts LED irradiation and detection of the optical device to be corrected in multiple orientations and multiple bands to obtain initial detection data containing reflection distribution information, transmission distribution information and interference fringe patterns; based on the data, an error distribution map indicating the coordinate position and the corresponding error type is generated, and each area is ground, polished or partially film-repaired to obtain updated data consisting of the local morphology after repair, film characteristics and coordinate information of the repaired area; then LED irradiation and detection are performed again over the entire surface to obtain a comprehensive correction result. This method ensures complete detection of multi-layer film systems and high-precision surfaces while achieving timely correction of local deviations, and is suitable for mass production of various high-end optical devices.

[0015] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention are realized and obtained by the structures particularly pointed out in the description, claims and drawings.

[0016] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a schematic diagram of a first embodiment of a method for correcting machining errors of an optical device according to an embodiment of the present invention; Figure 2 Schematic diagram of an embodiment of a processing error correction device for an optical device in an embodiment of the present invention; Figure 3It is a schematic diagram of an embodiment of a processing error correction device for an optical device in an embodiment of the present invention. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0019] The terms "including" and "having" and any variations thereof mentioned in the embodiments of the present invention are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device end including a series of steps or units is not limited to the listed steps or units, but may optionally include other steps or units that are not listed, or may optionally include other steps or units that are inherent to these processes, methods, products or device ends.

[0020] To facilitate understanding of this embodiment, a method for correcting processing errors of an optical device disclosed in an embodiment of the present invention is first described in detail. Figure 1 As shown, the method comprises the following steps: 101. Perform LED irradiation processing on the optical device to be calibrated in multiple directions and multiple bands to obtain initial detection data containing reflection distribution information, transmission distribution information and interference fringe pattern; In one embodiment of the present invention, the optical device to be corrected is subjected to LED irradiation processing in multiple orientations and multiple bands to obtain initial detection data containing reflection distribution information, transmission distribution information and interference fringe patterns, including: according to the appearance characteristics of the optical device to be corrected, multiple LED sub-light sources are arranged in a ring or matrix to obtain a multi-channel illumination array with different bands and angles; according to preset timing control information, LED sub-light sources of different bands or different incident angles in the multi-channel illumination array are sequentially illuminated to obtain reflection images, transmission images and fringe images of the optical device surface in each band or orientation; and the reflection image, transmission image and fringe image are respectively subjected to data separation processing to obtain initial detection data containing reflection distribution information, transmission distribution information and interference fringe patterns.

[0021] Specifically, when constructing the lighting environment, it is necessary to determine the specific placement form of the LED sub-light source according to the shape of the optical device to be corrected, and to arrange it in a ring or matrix to obtain sufficient optical response signals in multiple directions and bands. If the optical device is circular or has an outer periphery close to a spherical surface, the LED sub-light source can be laid out in a concentric ring along the periphery of the device to obtain multi-angle illumination in the circumferential direction; if the device shape is close to a rectangle or an irregular polygon, the LED sub-light source can be arranged in a matrix consisting of several rows and columns, and auxiliary light sources can be added in the corners or edge areas in combination with the device edge information to improve the illumination coverage of the corner area. In order to make the band characteristics of each LED sub-light source output consistent with the design requirements, it is necessary to allocate appropriate driving current and pulse width control at the circuit level, and use LEDs with different chips or different packaging types to meet the multi-band combination. The LED sub-light source corresponding to each band can be classified according to the central wavelength of the spectrum, so that there is a slight overlap between adjacent bands, thereby forming a complete spectrum coverage in physical space. When the LED sub-light sources are arranged in a ring shape, the radial distance between the central area and the peripheral area will affect the way the light beams converge. Therefore, the size of the optical device and the number of light sources need to be considered comprehensively when arranging the number of rings. For a matrix layout, equidistant or non-equidistant distributions can be set in the X-axis and Y-axis directions respectively to adapt to various surface features of the optical device, and LEDs of different bands or different powers can be configured in each grid unit. If the LEDs in a row are mainly responsible for the visible light range, the adjacent rows can consider arranging infrared or ultraviolet bands, so as to have flexible band distribution in both the horizontal and vertical directions. After the layout is completed, the coordinate index of each LED sub-light source is marked in turn to facilitate accurate mapping during subsequent timing control and image recording. Finally, a multi-channel illumination array with sub-bands and sub-angles is formed through a ring or matrix arrangement.

[0022] Specifically, after the multi-channel illumination array is constructed, it is necessary to sequentially light up the LED sub-light sources of different bands or different incident angles according to the preset timing control information, so as to divide the multi-band illumination process into multiple time segments, so as to distinguish the respective optical effects in the image acquisition stage. The timing control information is usually saved by a digital controller or an industrial computer as a tabular instruction set, including the light source number, lighting sequence, on-time, and standby time. The controller synchronously schedules the LED sub-light source and the image acquisition device according to this information. During the period when the LED sub-light source of a certain band is powered on, the other bands remain off, and switch to the next band or the next incident angle after the change, so as to separate the illumination effects under different light source conditions. The reason for doing this is that if multiple wavelengths are started at the same time, the reflected light and the transmitted light will be mixed into the same picture, and it will be difficult to distinguish the intensity distribution corresponding to each band in the image post-processing. The image acquisition device can be a high-resolution camera or a detection system based on an array photoelectric sensor. Whenever the controller sends a trigger signal, the image sensor records the surface response of the optical device under the current band and angle conditions within the specified exposure time. If the optical device has a multilayer film or a curvature change, when the LED irradiation light interferes with the device surface, alternating light and dark stripes will appear in the image, and such stripes can be called part of the interference image. By lighting up each band in turn, a large number of independent data frames can be obtained in one round of detection, including reflection images, transmission images, and stripe images. The reflection image highlights the reflection intensity distribution of the light in this band on the surface of the device, the transmission image shows the transmittance of the optical device inside this band, and the stripe image focuses on the interference pattern formed by the optical path difference. If the incident angle distribution of the device is diversified, multi-angle information can be obtained by irradiating in different directions within the same band, presenting a richer spatial resolution.

[0023] Specifically, after the sequential lighting process is completed and a series of images are collected, these images need to be processed for data separation to obtain initial detection data containing reflection distribution information, transmission distribution information and interference fringe patterns. The image processing software will first read the light source number and time stamp in the timing record, and then extract the corresponding frame or frame group under the number from the image sequence, and use this to identify which frames belong to reflection images, which frames record transmission images, and which frames contain significant interference fringes. In order to extract the reflection distribution information, the grayscale value or brightness value of the pixel on the surface of the optical device can be corrected during the image processing process, and converted into a light intensity ratio relative to the reference background or dark field image, so as to obtain a reflectivity distribution map on a two-dimensional coordinate. Similarly, the acquisition of transmission distribution information can be based on the calculation of the ratio of transmitted light intensity to incident light intensity, and the noise signal of the non-transmission area is eliminated in the image, so as to obtain an intuitive display of the internal transmission performance of the device. The fringe image can use the interference fringe recognition algorithm to highlight the alternating light and dark stripes, and use the phase extraction method to preliminarily encode the optical path difference. The fringe center and fringe spacing can indicate the local film thickness or the convexity and concavity of the curved surface. If further phase calculation or quantitative analysis of fringes is required, in-depth solution can be performed in the subsequent links. Here, the fringe image is only marked as a storage method for the interference fringe pattern. After the data separation process is completed, all the extracted reflection distribution information, transmission distribution information and interference fringe patterns will be integrated into a unified structure and mapped with the physical coordinates or pixel coordinates of the optical device, so that the optical properties of a certain point or area can be quickly located during the subsequent difference comparison. The initial detection data thus formed covers three levels: spatial domain, wavelength domain and interference domain, and each information can be called in the form of a matrix or image after digitization, which is conducive to the subsequent error identification, area division and processing correction operations.

[0024] 102. Based on the initial detection data, perform region division and difference comparison processing on the reference parameters of the optical device to obtain an error distribution map indicating the coordinate position and the corresponding error type; In one embodiment of the present invention, the reference parameters of the optical device are divided into regions and subjected to difference comparison processing according to the initial detection data to obtain an error distribution diagram indicating the coordinate position and the corresponding error type, including: performing difference operation processing on the reflection distribution information in the initial detection data and the reference reflection parameters of the optical device to obtain a reflection difference matrix that associates the corresponding optical coordinates with the reflection deviation value; performing difference operation processing on the transmission distribution information in the initial detection data and the reference transmission parameters of the optical device to obtain a transmission difference matrix that associates the corresponding optical coordinates with the transmission deviation value; performing phase resolution processing on the interference fringe pattern in the initial detection data, and performing comparison operation processing in combination with the reference fringe parameters to obtain a wavefront offset matrix; performing multi-dimensional data fusion processing on the reflection difference matrix, the transmission difference matrix and the wavefront offset matrix to obtain a comprehensive data set containing optical coordinates, reflection deviation, transmission deviation and phase change information; and performing partition labeling processing on the comprehensive data set using a coordinate indexing algorithm to obtain an error distribution diagram indicating the coordinate position and the corresponding error type.

[0025] Specifically, when the reflection distribution information in the initial detection data and the reference reflection parameters of the optical device are subjected to difference calculation processing to obtain the reflection difference matrix associated with the corresponding optical coordinates and the reflection deviation value, it is necessary to first combine the reflection distribution information with the device surface coordinates in a pixel-level or sampling point-level manner, so as to ensure that the measured value of each coordinate point can be aligned to the correct position during the calculation process. If the initial detection data is stored in the form of a two-dimensional image, it is necessary to establish a mapping between the row and column indexes of the image and the physical coordinates of the device so that each pixel corresponds to a specific spatial position. If the device has a non-planar or non-spherical morphology, it is also necessary to introduce a geometric correction function to transform the image coordinates into optical coordinates. Next, the corresponding reference reflection parameters can be stored in a lookup table or function expression to characterize the ideal reflectivity or reflection intensity distribution at different positions. If some areas are designed as high-reflection film layers and other areas are designed as anti-reflection structures, the correct reference should be located according to the pre-arranged reference values ​​during the comparison. The difference operation obtains ΔR. When ΔR is greater than a certain set threshold, it means that there is a significant reflection error here. If ΔR is lower than the threshold, it can be regarded as relatively normal. In order to continuously observe these deviations in the spatial domain, each ΔR value can be stored in a two-dimensional matrix or a three-dimensional array according to the coordinate mapping, so that its index position is consistent with the physical coordinate, thereby obtaining a reflection difference matrix. This matrix not only provides the size of the deviation value, but also reflects the distribution form of the deviation between neighborhoods, which has a good indication for detecting large-scale consistency or local anomalies. In the specific implementation, it is necessary to import the measured reflection distribution information into the image processing software, and then perform pixel-to-pixel or point-to-point subtraction operations with the reference reflection parameters. If necessary, the reflection value can also be normalized to eliminate interference caused by fluctuations in the brightness of the LED light source or differences in the camera gain setting. After the operation is completed, the system will write the results into a storage structure and package them synchronously with the coordinate field to form a reflection difference matrix that contains both the coordinate index and the reflection deviation value. In the subsequent process, this matrix will be fused with the transmission and interference results to provide more refined information support for identifying multiple error types.

[0026] Specifically, when the transmission distribution information in the initial detection data and the reference transmission parameters of the optical device are subjected to difference calculation processing to obtain the transmission difference matrix associated with the corresponding optical coordinates and the transmission deviation value, it is necessary to perform special data correction and comparison according to the collection characteristics of the transmission data. Since the transmittance or the transmitted light intensity is often affected by the device material, the film thickness and the incident angle, it is necessary to first select the working conditions or corresponding wavelengths consistent with the reference transmission parameters according to the band identification and angle identification in the initial detection data. If the design of the optical device requires extremely high transmittance in some areas, and optical filtering or reflection enhancement structures may be used in other areas, it is necessary to assign corresponding reference transmission values ​​or transmission functions to different functional areas. After obtaining the accurate corresponding relationship, the calculation process is similar to the generation of the reflection difference matrix. It is necessary to first normalize the light intensity in the image or point cloud, and divide the transmitted light intensity by the incident light intensity to eliminate the influence caused by different LED power outputs. At the same time, it should be checked whether there are multiple interferences or absorption peak shifts according to the matching of the device thickness and the band, so as to find the correct parameter comparison in the reference data. After the difference operation is completed, the ΔT at each position is output to the corresponding two-dimensional matrix unit and bundled with the coordinate index to form a transmission difference matrix. If the same device area presents a large positive ΔT value, it means that the actual transmittance is significantly higher than the design target. If it is a negative value with a large amplitude, it means that the area is insufficiently transmitted or the film layer blocks too much light.

[0027] Specifically, when the interference fringe pattern in the initial detection data is subjected to phase solution processing and compared with the reference fringe parameters to obtain the wavefront offset matrix, a series of image processing and mathematical operations need to be performed on the fringe image or fringe field data. This matrix has a similar data structure to the aforementioned reflection and transmission difference matrix, but the information it carries focuses on the wavefront deformation of the device or the optical path difference effect between multilayer films. The denser the fringe edges or the larger the phase gradient, the more attention it needs to pay to the subsequent trimming process.

[0028] Specifically, when performing multi-dimensional data fusion processing on the reflection difference matrix, transmission difference matrix and wavefront deviation matrix to obtain a comprehensive data set containing optical coordinates, reflection deviation, transmission deviation and phase change information, it is necessary to match and combine the corresponding elements in the three matrices with the same coordinate system as the core. After the fusion is completed, the storage structure of the comprehensive data set can be in the form of a multi-dimensional array or table, with each unit corresponding to a coordinate point and three main deviation attributes and optional band information.

[0029] Specifically, when the coordinate index algorithm is used to partition and label the comprehensive data set, and an error distribution diagram indicating the coordinate position and the corresponding error type is obtained, it is necessary to classify and define the error category of each area based on the various deviation information stored in the above comprehensive data set. The implementation method generally first uses a scanning algorithm to traverse all the coordinate points of the comprehensive data set. During the scanning process, if the deviation amount of some positions meets the conditions of "high reflection difference and low transmission difference", it can be marked as a speculative partition such as surface deformation or coating thickness abnormality; if its wavefront offset value is large and the transmission deviation is also obvious, it can be attributed to the optical path mismatch caused by film stress. The partitioning strategy can be automatically partitioned by distance measurement or clustering method, such as using a density-based algorithm or a connected domain-based search method. Once a homogeneous neighborhood is detected, the neighborhood is collected as an error area, and different types of error labels are further distinguished internally. Thereafter, one or more labels are assigned to each error area, such as "too thick film layer", "local depression", "refractive index drift", etc., and these labels and the vertex coordinates of the area boundary are saved in the annotation list. In order to display the partition annotation results on the graphical interface, the error distribution map can be drawn as a two-dimensional thermal or pseudo-color map, with each area filled with different colors or wireframes to convey different types of deviations. Since the comprehensive data set includes reflection, transmission and wavefront information, the annotation process can more accurately present the multiple causes of errors. If certain areas meet multiple difference criteria, two labels can also be superimposed to remind subsequent processes to pay attention to the complex situation of this area. When all coordinate points are processed, an error distribution map indicating the coordinate position and the corresponding error type will be obtained.

[0030] Furthermore, the phase solution processing of the interference fringe pattern in the initial detection data and the comparison operation processing in combination with the reference fringe parameters to obtain the wavefront offset matrix include: pixel-level extraction processing of the fringe peak and valley distribution in the interference fringe pattern to obtain preliminary fringe information including fringe position coordinates and intensity changes; based on the preliminary fringe information, phase unfolding processing is performed on the fringe image to obtain a phase distribution matrix based on pixel coordinates; difference comparison processing is performed on the phase distribution matrix and the reference fringe parameters to obtain phase difference data with phase offset values ​​associated with fringe pixel coordinates; interpolation fitting processing is performed on the phase difference data to obtain a phase distortion distribution that characterizes the wavefront distortion trend; matrix conversion processing is performed on the phase distortion distribution in combination with the coordinate mapping relationship of the interference image to obtain a wavefront offset matrix marked with optical coordinates and phase offsets.

[0031] Specifically, before performing phase solution on the interference fringe pattern in the initial detection data, it is necessary to perform pixel-level extraction processing on the fringe peak and valley distribution to obtain preliminary fringe information including fringe position coordinates and intensity changes. When implementing this process, first load the interference image or interference fringe data into the image analysis module, and use the filtering method to reduce random noise so that the grayscale contrast between the fringe area and the background area is clearer. Subsequently, by scanning pixel by pixel and performing gradient analysis on the neighborhood grayscale, the local maximum point can be determined as the peak position and the local minimum point can be determined as the valley position. If the fringe has a large curvature or slight distortion, it is necessary to combine threshold comparison and jump detection to ensure the accuracy of the peak and valley position while performing extreme value judgment. When the pixel coordinates of the fringe peaks and valleys are recorded in one step, a rotational association can be established between adjacent peaks and adjacent valleys in the fringe chain reconstruction step. This approach makes the overall direction of the fringe appear as a traceable path at the data level, rather than just discrete scattered points. If the contrast of the interference fringe appears spatially uneven due to the light source distribution or surface reflectivity, the brightness range can be normalized: ; Through this normalization process, the peak-to-valley differences at different positions can be made easier to identify gradients. After the extraction is completed, each peak-to-valley pixel index and its intensity value are uniformly encoded to form a preliminary stripe data set containing a variety of information such as stripe center coordinates, grayscale values, stripe segment affiliation relationships, etc. This data set can provide a basis for judging the stripe period in subsequent steps and provide peak-to-valley reference points for the phase unwrapping process. If the image resolution is high, a sub-pixel interpolation strategy can also be introduced in the peak-to-valley detection process to enable the positioning accuracy of the peak-to-valley coordinates to exceed one pixel unit, thereby maintaining a finer spatial resolution in subsequent phase calculations. Through pixel-level extraction processing, the preliminary stripe information has correlations in geometric position and grayscale features. This information can be output in a point cloud format or a two-dimensional matrix according to needs, so that the subsequent phase unwrapping step can directly retrieve and interpolate the phase value for each stripe period. This pixel-level extraction helps capture subtle distortions of the wavefront.

[0032] Specifically, according to the preliminary fringe information, when performing phase unwrapping processing on the fringe image, it is necessary to map the light and dark distribution between the peaks and valleys to the continuous phase space so that each pixel has a corresponding phase value. When implementing this process, the pixel coordinates of the adjacent peaks and valleys in the fringe chain are first read, and the corresponding grayscale value range is determined, and then the phase is allocated between the peaks and valleys based on the interpolation model. If linear interpolation is introduced, the grayscale from valley to peak can correspond to or interval; if there is a strong bend in the stripes or nonlinear brightness changes, the grayscale gradient distribution of each stripe segment can be mapped to the phase gradient curve using a piecewise spline function or polynomial fitting method. To overcome the periodic break caused by phase shift jumps, when performing phase stitching on the stripe area, the phase trend of the adjacent stripe segments can be referenced to keep the phase smooth at the connection. If a phase flip occurs in the image area beyond If there is a phenomenon, the position is corrected during the expansion process and unified to or If the interference pattern is collected from a multi-band LED light source, it is also necessary to perform phase unwrapping on the fringes in different bands separately, and generate a phase distribution matrix for each band to reflect the fringes phase at different wavelengths. A composite interpolation function can be defined ; in is the initial phase reference, and Represent the grayscale values ​​of the valley and peak positions of the stripes respectively. This function allocates the complete phase range within the peak-to-valley interval. If more precise local adaptation is required, the function can be segmented or regularized to suppress interpolation oscillation. After the calculation of each pixel is completed, it is stored in the phase distribution matrix, and the row and column indexes are consistent with the image coordinates. In this way, a phase distribution matrix based on pixel coordinates is obtained, which can clearly indicate the degree of phase abnormality of each pixel point when compared with the reference stripe parameters in subsequent steps.

[0033] Specifically, when performing difference comparison processing on the phase distribution matrix and the reference fringe parameters, it is necessary to make the ideal fringe reference also have a phase distribution form that can be directly aligned. If the reference fringe parameters are derived from optical simulation, an ideal corresponding image for Dalville at the same resolution and image can be called If the reference data comes from experimental calibration or standard device images, the same phase unwrapping operation as in step 2 is performed on that image. , get the phase difference data, each pixel It can be negative or positive. The larger the value, the more the measured phase advances at the corresponding coordinate. Interval folding, performing phase normalization before or after the difference operation, Mapping back or If the stripes have a topological structure surrounding the center, it is also necessary to detect the phase discontinuity problem at the closed edge of the stripes and perform phase stitching for these surrounding areas separately. When visualizing the matrix, if you see large areas of high or low brightness, this means that the device has significant phase distortion relative to the reference in this area. To measure the degree of distortion at a spatial statistical level, you can calculate the RMS value or maximum deviation value and set a threshold for the phase difference based on the image coordinates to pick out the coordinate groups with more abnormalities. After the values ​​are collected, they are sent to the subsequent interpolation fitting or clustering module for further analysis. Through difference comparison, the stripe pixel coordinates and phase offset values ​​are encapsulated into a phase difference matrix, and each pixel is connected to its specific coordinates and The values ​​form a high-resolution data that can correlate subtle local differences, providing a more accurate basis for surface or film correction strategies for optical devices.

[0034] Specifically, when the phase difference data is interpolated and fitted to obtain the phase distortion distribution that characterizes the wavefront distortion trend, it is necessary to convert the discrete The values ​​are smoothed into a phase continuous surface. The specific implementation can introduce algorithms such as two-dimensional spline interpolation, polynomial fitting or radial basis function expansion to convert pixel coordinates As an independent variable, As the target value, establish a surface , so that the surface is based on the minimum residual error at all pixel points. If the device surface has obvious aspheric effects, Zernike polynomials or high-order polynomials can be used to pass the local curvature changes sharply in order to capture the depth deformity. The fitting process can define an objective function ; Among themReg is a regularization term used to prevent excessive oscillation or unreasonable oscillation peaks on the surface. To adjust the weight of this item in the total error. By iterating or solving a system of linear equations, a smooth distribution can be obtained. , which is at each pixel and If the phase difference information in different bands needs to be processed uniformly, then Multiple channels are merged into high-dimensional input, allowing the interpolation process to fit each band at the same time and introduce the band index as an independent variable. After the fitting is completed, Saved in a two-dimensional array, it becomes a more holistic phase distortion distribution, which can be considered as the overall darkening of the ideal wavefront produced by the weighing optical device. This distribution will not jitter excessively when there is random noise or local deviation in a single pixel, providing a smooth and differentiable phase function description for subsequent coordinate mapping or optical analysis, and can intuitively reflect which area has more significant distortion in subsequent matrix conversion operations.

[0035] Specifically, when performing matrix transformation processing on the phase distortion distribution combined with the coordinate mapping relationship of the interference image to obtain a wavefront shift matrix annotated with the optical coordinates and the phase shift, the image coordinates need to be Projection to the device's optical coordinate system or polar coordinates If the device has an aspherical or free-form surface, the calibration step can be completed before detection, and the mapping relationship between the image coordinates and the three-dimensional coordinates can be saved as a geometric transformation matrix or polynomial function. , query its coordinates on the device surface and from Retrieve the phase distortion value from to write into the wavefront shift matrix If the device is a spherical or rotationally symmetrical body, it can also be simplified to polar coordinate mapping, through Determine the position of each pixel under the axisymmetric reference and assign the corresponding phase value. If some areas lose data during shooting, invalid values ​​can be marked after mapping or neighborhood interpolation can be performed to complete the filling. After the mapping is completed, a wavefront offset matrix that can be adapted to the optical coordinates is formed, giving the phase offset of the device surface or film layer at each coordinate point. If ray tracing or diffraction simulation is required later, the matrix can be called as the input of the surface error or film layer error, and combined with information such as the refractive index and film thickness function to perform a more in-depth simulation of the optical performance. If you want to perform CNC trimming, you can refer to The position where the wavefront distortion is too large is used to locate the grinding or film repair operation area. The matrix conversion process elevates the interpolated phase distortion field from the image pixel domain to the device real coordinate domain, which is more conducive to executing the correction command based on the spatial coordinates in the processing control, and is also convenient for generating a three-dimensional wavefront topology map in the visualization software, thereby presenting the distortion distribution of the optical device at the actual spatial scale.

[0036] Furthermore, the use of a coordinate indexing algorithm to perform partition labeling processing on the comprehensive data set to obtain an error distribution map indicating coordinate positions and corresponding error types includes: performing coordinate registration processing on the reflection deviation, transmission deviation and phase change information contained in the comprehensive data set to obtain a multidimensional error point cloud indexed by a unified optical coordinate system; performing difference threshold division processing on each pixel or sampling point based on the multidimensional error point cloud to obtain a number of error units showing similar deviation characteristics; performing coordinate clustering processing on the error units to obtain a cluster center coordinate and a corresponding deviation information index table; performing label assignment processing on each error unit based on the deviation information index table to obtain a region mark list containing optical coordinates and error type labels; performing visual mapping processing on the region mark list to obtain an error distribution map indicating coordinate positions and corresponding error types.

[0037] Specifically, when the reflection deviation, transmission deviation and phase change information contained in the comprehensive data set are processed for coordinate registration, the error values ​​in different dimensions need to be mapped to the optical coordinates in a consistent manner. If the data comes from the image pixel domain, each pixel or sampling point needs to be projected in a two-dimensional or three-dimensional space based on a geometric calibration function or a surface equation obtained by prior measurement, so that it has corresponding coordinates within the optical coordinate system (u, v) or (r, θ). When the device has a spherical or aspherical shape, the surface equation z=f(x, y)z = f(x, y)z=f(x, y) can be used during mapping to transform the pixel index into surface coordinates, and then store the ΔR, ΔT and Δϕ values ​​of the point. After the projection is completed, the three deviation values ​​can be retrieved at any coordinate point (u, v)(u, v)(u, v) to form a multidimensional error vector [ΔR, ΔT, Δϕ]. All points can be collected to obtain a multidimensional error point cloud, and sampling points whose coordinates exceed the device outline can be removed or marked with invalid values. If different band indices λ are generated in multi-band measurements, λ can be added as the fourth dimension to achieve error tracking for spectral and spatial coupling. This point cloud data structure can provide more complete information for subsequent difference analysis of single points or local areas. It can be stored in the form of coordinates and deviation vectors in a matrix or table, or converted into a point cloud visualization format to check the correlation characteristics between various errors when scanning the coordinate domain.

[0038] Specifically, according to the multidimensional error point cloud, each pixel or sampling point is divided by a difference threshold, and a group of points that are spatially adjacent to each other and have similar values ​​can be classified as an error unit. A comprehensive metric function can be defined, for example, the vector modulus of ΔR, ΔT, Δϕ of the same sampling point is calculated to obtain the vector modulus D, and DDD is compared with a preset threshold. If DDD does not exceed a certain limit, it means that the total three-dimensional deviation of the point is small, and it can be merged with adjacent points with the same deviation range. If it exceeds the limit, it is regarded as another error level when dividing. At this time, if a very low deviation is expected for a part of the device, a stricter threshold can be used for the area. If a certain deviation is allowed to float in other areas, the threshold can be appropriately relaxed. Through this type of hierarchical division processing, points with similar error information in the multidimensional point cloud can be divided into multiple areas, called error units. Each unit can represent a local deviation block or an error cluster of the same cause, which is convenient for accurately finding blocks with more consistent error types during subsequent clustering.

[0039] Specifically, coordinate clustering processing is performed on the above error units to extract representative coordinate centers and corresponding main deviation attributes from local clusters of multidimensional point clouds. If there are many points in a unit and multiple extreme value centers are distributed, clustering algorithms can be introduced again within the same unit, such as density-based DBSCAN or distance-based k-means, to further divide these data points on spatial coordinates (u,v)(u,v)(u,v) and deviation vectors. If some subclusters have similar ΔR, ΔT, Δϕ distributions, they can be classified as the same type. After clustering, the position of the centroid or weighted centroid in the cluster can be calculated to obtain the coordinates of the cluster center, and then the average or median of the reflection deviation, transmission deviation and phase change are calculated from the data points of the cluster to form a deviation information index table. The index table records the cluster number, center coordinates and main error indicators of the cluster, and can also mark the number of nodes or the maximum deviation value to provide a basis for determining different cluster types in subsequent label assignment. For cluster centers, if multi-band joint analysis is required, sub-band weighting or overall weighting can be performed on [ΔRλ, ΔTλ, Δϕλ] of multiple bands to compare their deviations from the design values, so as to determine whether the cluster is closer to geometric defects or uneven film interference.

[0040] Specifically, when assigning labels to each error unit according to the deviation information index table, it is necessary to establish a corresponding relationship between the cluster number or the main deviation feature of the unit and the specific error type with the help of a preset error type dictionary or judgment rule. If the cluster is significantly positive in reflection deviation and the phase difference distribution is a small gradient, it may belong to the film over-reflection unit; if the transmission deviation and phase change are both high, it can be judged as a multilayer film thickness deviation unit; if ΔR and ΔT are both small values ​​but the Δϕ gradient is large, it is mostly a surface deformation unit. Several labels can be defined for each type, such as "film layer is too thick", "film layer refractive index offset", "surface concave", etc. If multiple deviation modes appear in the same cluster, they can be split into further labels or classified as "mixed" identification. After completing the label judgment, the cluster center coordinates and label content of the unit are stored in the regional mark list, and the coordinate range of all sampling points in the cluster is recorded. The list is a [unit ID, center coordinates, error type label, coordinate boundary] structure, which can provide a basis for subsequent visualization or processing decisions. If some units behave differently under multi-band detection, you can also add a band index to the tag field to distinguish anomalies under specific bands.

[0041] Specifically, when visual mapping is performed on the area marking list to obtain an error distribution diagram indicating the coordinate position and the corresponding error type, the error units and their labels defined in the previous stage need to be drawn in two-dimensional or three-dimensional coordinates. If the shape of the device can be approximated by a plane, different colors or textures can be applied to the coordinate area of ​​each error unit on the two-dimensional plane, and the corresponding error label can be displayed at the coordinate of the unit center; if the device is a sphere or a free-form surface, a surface mesh can be generated in three-dimensional space, and the mesh facets belonging to the same error unit can be rendered in the same color, and corresponding text or symbol labels can be added to the center or boundary of the mesh, so that the graphical interface can intuitively present the location of the high deviation area or the abnormal film layer area. If the data generated by the system for multiple bands needs to be displayed in parallel, the distribution differences in different bands can be displayed in layers according to the band index or by using animation switching.

[0042] 103. According to the error distribution map, grinding, polishing or local film repairing is performed on the corresponding area of ​​the optical device to obtain updated data including the local morphology after repair, film characteristics and coordinate information of the repaired area; In one embodiment of the present invention, the grinding, polishing or local film repairing of the corresponding area of ​​the optical device according to the error distribution map to obtain updated data including the local morphology after trimming, film characteristics and coordinate information of the trimmed area includes: determining the area marked as surface geometric error and the area marked as abnormal film thickness in the error distribution map; performing a removal depth calculation process on the area marked as surface geometric error to obtain a cutting depth parameter for the grinding work, and performing a grinding or polishing process on the area of ​​the surface geometric error to obtain temporary morphology data with improved surface flatness; performing a film increase or decrease amount estimation process on the area marked as abnormal film thickness to obtain a film control instruction for local deposition or removal, and performing a thin film increase or decrease operation process on the area of ​​the abnormal film layer to obtain local updated data including compensated film thickness information; integrating the temporary morphology data with the local updated data to obtain temporary stored data including the local morphology after trimming and film characteristics, and adding the coordinate index information to the temporary stored data and performing a unified marking process to obtain updated data including the local morphology after trimming, film characteristics and coordinate information of the trimmed area.

[0043] Specifically, before grinding, polishing or local film repairing the corresponding area of ​​the optical device according to the error distribution map, it is necessary to first determine which parts of the error distribution map are marked as surface geometric error areas and which parts are marked as abnormal film thickness areas. This process will parse the coordinate information and error type labels obtained after the previous detection and difference comparison, and classify and sort out the labels of each pixel or sampling point in the two-dimensional or three-dimensional coordinate domain. If there is no obvious change in reflection deviation and transmission deviation in a certain area on the surface of the device, but the phase change amplitude exceeds the threshold, it is often judged as a geometric error area, such as local scratches, depressions or curvature mismatch; if it is found in multi-band detection that the transmission distribution of this area is significantly low, the reflection is enhanced, or the interference fringe period is offset inconsistent with the film calculation, it will be recorded as an abnormal film thickness area. In order to make the judgment process highly targeted, it is necessary to use the error stratification strategy at the algorithm level, and make different threshold judgments for the phase distortion caused by surface deformation and the spectral response deviation caused by uneven film thickness, and make special cases for those complex areas with both morphology and film abnormalities. If there are both large-scale deformation and significant film defects in the same coordinate block, it may be split into two adjacent sub-blocks for grinding or film addition and subtraction operations. In order to keep the structure clear in the program implementation, a regional information table can be defined to quantify and store the error type, coordinate boundary, area size and severity of each coordinate zone, and add several identification fields to it, such as "GeoErr" to mark geometric errors and "ThkErr" to mark film thickness abnormalities. When this recognition process is completed, the surface geometric error area and the film thickness abnormality area have been finely divided in the coordinate space, and each area has a clear positioning index and error cause mark. This information is integrated into the process planning module to guide the next targeted operation.

[0044] Specifically, when calculating the removal depth of the area marked as the surface geometric error to obtain the cutting depth parameters for the grinding work, it is necessary to determine how much material should be removed during grinding or polishing based on the error quantification results of the area and the material characteristics of the device. If the error distribution diagram shows a large phase offset or the surface height deviation exceeds the expected tolerance in this area, the height deviation value Δ can be regarded as the preliminary removal target, and the cutting coefficient κ generated by the grinding head or polishing pad on the material in actual processing is considered to calculate the cutting depth δ=κ⋅Δh. If the device material is relatively brittle, it is necessary to control the grinding speed and grinding path to avoid excessive cutting that causes cracks or micro-notches on the surface. After completing the calculation of the cutting depth parameters, the machine tool or polishing equipment will perform a quantitative removal operation, perform local or layered grinding on the area, and flatten the uneven or scratched areas. In order to prevent a steep transition between the grinding area and the surrounding qualified area, a progressive depth change can be set in the path planning to enable the grinding head to achieve smooth Z-direction feed in this area. After grinding, polishing can be performed to further reduce surface roughness and improve finish. At this time, finer grinding fluid or polishing powder is often used in combination with a polishing pad to evenly distribute local friction and suppress residual microscopic scratches. After the treatment, a temporary measurement or topography scan of the current status of the area is required to obtain temporary topography data after the surface flatness is improved. These data can be stored as a point cloud or a two-dimensional table, and combined with the coordinate information to index to the corresponding geometric error area to form a comparison record of the status before and after trimming. If the device morphology is close to the target surface or tolerance range, the area can be marked as "ground" in the temporary topography data, otherwise you can return to this step for incremental removal until the flatness meets the design requirements.

[0045] Specifically, when estimating the film increase or decrease amount in the area marked as abnormal film thickness to generate the film control instruction for local deposition or removal, it is necessary to refer to the reflection and transmission deviation obtained by multi-band spectral detection and combine the film system design parameters of the area. If the test results show that the film layer in the area is thin, a local film increase instruction can be issued to the control system, and additional deposition can be performed in the form of a partition mask through a sputtering or evaporation device. If the test results show that the film layer is too thick or the refractive index deviates too much, the local film layer can be thinned by ion beam etching, plasma etching or chemical stripping process, and the etching operation is performed if it is negative. In the partition deposition process, it is necessary to set a mask or a movable physical mask for the area according to the coordinate boundary of the error distribution map, so that the coating or etching device is only effective for the specified coordinate range to avoid affecting the surrounding areas that have reached the standard. After performing the film increase or film removal operation, the film thickness should be measured again to confirm whether the difference has dropped to an acceptable range. If there is still a deviation, the deposition or etching of small steps can be repeated multiple times to avoid excessive repair at one time and cause new problems. When integrating the temporary morphology data with the local update data, the surface geometry information obtained after grinding or polishing and the thickness change information obtained by the film layer addition and subtraction operations need to be unified into the same data structure. After obtaining the temporary data, it is also necessary to add the coordinate index information to it and perform unified marking processing to generate update data containing the trimmed local morphology, film layer characteristics and trimmed area coordinate information. This processing requires unified encoding of the integrated area information, and packaging the coordinate boundary, layer index and trimming results of each area together.

[0046] 104. According to the updated data, the optical device is again subjected to LED irradiation and detection processing over the entire surface to obtain a comprehensive correction result that marks the overall reflection change, transmission change and fringe change.

[0047] In one embodiment of the present invention, the optical device is subjected to LED irradiation and detection processing again over the entire surface range according to the updated data to obtain a comprehensive correction result with overall reflection changes, transmission changes and fringe changes marked, including: according to the trimmed area coordinate information in the updated data, the entire surface range of the optical device is divided into detection areas to obtain a detection distribution map covering all trimmed areas and untrimmed areas; according to the detection distribution map, a plurality of LED sub-light sources are subjected to band-based triggering processing to obtain a multi-angle irradiation scheme adapted to each coordinate area; the optical device is subjected to full-surface detection processing under the multi-angle irradiation scheme to obtain secondary sampling data containing reflection field, transmission field and interference fringe field information; the secondary sampling data is compared with the trimmed local morphology and film layer features marked in the updated data to obtain a comprehensive correction result of overall reflection changes, transmission changes and fringe changes.

[0048] Specifically, when dividing the detection area of ​​the entire surface range of the optical device according to the coordinate information of the trimmed area in the update data, it is necessary to rely on a complete coordinate mapping relationship to determine the distribution of all trimmed areas and unprocessed areas. If the surface of the optical device is aspherical or free-form, the three-dimensional coordinate boundaries of each trimmed area and the corresponding relationship with the surface reference surface will be recorded in the update data. The process first reads the coordinate interval and label information of each trimmed area, and then classifies this area as a "trimmed area" in the overall coordinate domain, and the remaining coordinate areas that have not been recorded in the update data are designated as "untrimmed areas". In order to plan the detection task more flexibly, the surface can be divided into several blocks, and the size of different blocks can be determined by the previous detection accuracy and the specific morphology of the device. If there is an obvious curvature mutation in some surface parts, the number of sub-blocks is increased during the division, so that the detection system can implement more intensive irradiation and collection of these positions. By integrating the above division results into a coordinate index map, a detection distribution map covering all trimmed and untrimmed areas can be generated, and the high-error sensitive areas to be inspected are distinguished from the low-error or trimmed areas for routine inspection. The distribution map can be stored in the form of a grid matrix, vector polygon or irregular grid, depending on the complexity of the device shape and the processing accuracy requirements. If it is necessary to automatically schedule the partition inspection order, the priority field can also be set on the distribution map to give priority to LED irradiation scanning in areas with larger areas or complex geometries.

[0049] Specifically, according to the detection distribution map, the process of performing sub-band triggering processing on multiple LED sub-light sources relies on a set of timing control logic and spatial matching strategies to achieve flexible irradiation of different coordinate areas. First, the specific positions of the trimmed area and the untrimmed area in the coordinate domain have been marked in the detection distribution map, and each block may involve different band requirements or incident angle requirements. For example, if a block has been optimized for the short band during the previous coating trimming, a higher priority medium and long band detection can be used at this time, and vice versa. In order to smoothly distinguish the optical effects of different bands and regions during a detection process, it is necessary to load a band-region correspondence table in the controller or host computer, which records the LED sub-light source number, lighting order and incident angle range specified for each block. Subsequently, the controller performs sub-band triggering arrangement according to this correspondence table before the detection starts, and arranges the LED sub-light sources to be lit in time segments to avoid signal superposition caused by turning on multiple bands at the same time. At the same time, in order to adapt to each coordinate area, the detection system can add a motorized platform or an adjustable optical device between the optical device and the LED sub-light source, so that the irradiation angle or irradiation position is accurately aligned with a certain block and switched to the next block when appropriate. If a multi-angle irradiation scheme is set in the corresponding table, the LED of the same band can be used to repeatedly observe the device at different tilt angles to obtain its reflection and transmission distribution at multiple viewing angles.

[0050] Specifically, when the optical device is subjected to full-surface inspection under the multi-angle illumination scheme, it is necessary to let the camera or optical sensor cooperate with the sub-band trigger sequence to collect the reflected light field, the transmitted light field and the interference fringe distribution under the corresponding angle and band conditions frame by frame. According to the time segment and spatial orientation arranged in the previous stage, the LED sub-light source will light up and illuminate the specific partition at the specified power at different times, and the sensor will record the light information returned from the device surface at the corresponding time. If the device has multi-layer film interference phenomenon in certain bands, it will be reflected as a fringe pattern in the camera image or the sensor receiving signal, which can be processed by phase solution or spectrum analysis method. If you do not want to slow down the inspection speed due to multiple maneuvers, you can place the device on a rotatable platform, let the LED array be in a relatively fixed position, and make each partition enter the inspection field of view by rotating the platform. This inspection process traverses all trimmed and untrimmed areas in a timing cycle, so as to form a batch of secondary sampling data containing multi-band and multi-angle dimensions after the end. For example, multiple frames of images can be collected for a certain block, each frame corresponding to a different band or a different tilt angle, and finally merged into a set of matrices representing local optical properties. If different priorities are set for the detection distribution map, more images can be collected in the high-priority blocks to accumulate more abundant data for error diagnosis. In this data stream, the reflection field component is used to quantify the reflection intensity of each pixel point to a specific band of light, the transmission field component shows the transmission ratio of the same area to light, and the interference fringe field presents the phase modulation result caused by the multilayer film or surface curvature. When processing the data, the image grayscale can be converted into the corresponding normalized light intensity, making it easier to identify spectral differences in subsequent analysis. If it is necessary to verify the surface flatness of the local area after trimming, the fringe spacing and distribution morphology of the block can be captured and compared with the previous initial detection stage to determine whether the interference distortion has been alleviated. If most areas have met the standards, the subsequent repeated detection can be reduced, otherwise further correction plans will be instructed in the next cycle. When performing comparison processing on the secondary sampling data and the trimmed local morphology and film layer features marked in the update data, it is necessary to first correspond the spatial index of the secondary sampling data with the coordinate information in the update data. If the updated data indicates that a certain trimming area has been ground to a specific depth or subjected to a specific film-increasing operation, then the corresponding expected value or threshold can be set for the block in the comparison algorithm. When the reflection field or transmission field in the secondary sampling data is not much different from the expected range corresponding to the block, it means that the trimming effect is significant; if the difference is still obvious, it means that the current trimming measures are not enough to eliminate the deviation or the area needs further fine-tuning. The final comprehensive correction result includes global information in the three dimensions of reflection, transmission and stripes, which can indicate the overall optical consistency and deeply display the degree of convergence of each local difference, thereby ensuring that the trimming goal of the optical device is achieved over the entire surface range.

[0051] In this embodiment, the optical device to be corrected is irradiated and tested by LED in multiple orientations and multiple bands to obtain initial test data including reflection distribution information, transmission distribution information and interference fringe pattern; an error distribution map indicating the coordinate position and the corresponding error type is generated based on the data, and each area is ground, polished or partially film-repaired to obtain updated data consisting of the local morphology after repair, film characteristics and coordinate information of the repaired area; then LED irradiation and testing are performed again over the entire surface to obtain a comprehensive correction result. This method ensures complete detection of multi-layer film systems and high-precision surfaces while achieving timely correction of local deviations, and is suitable for mass production of various high-end optical devices.

[0052] The above describes the processing error correction method of the optical device in the embodiment of the present invention. The following describes the processing error correction device of the optical device in the embodiment of the present invention. Figure 2 , an embodiment of the processing error correction device of an optical device in an embodiment of the present invention comprises: A multi-source illumination module is used to perform LED illumination processing on the optical device to be corrected in multiple orientations and multiple wavebands to obtain initial detection data containing reflection distribution information, transmission distribution information and interference fringe patterns; A difference comparison module, used to perform area division and difference comparison processing on reference parameters of the optical device according to the initial detection data, and obtain an error distribution map indicating coordinate positions and corresponding error types; A local trimming module, used to perform grinding, polishing or local film repairing on the corresponding area of ​​the optical device according to the error distribution map, and obtain updated data including the local morphology after trimming, film characteristics and coordinate information of the trimming area; The global correction module is used to perform LED irradiation and detection processing on the entire surface of the optical device again according to the updated data, so as to obtain a comprehensive correction result that marks the overall reflection change, transmission change and fringe change.

[0053] In an embodiment of the present invention, the processing error correction device of the optical device runs the processing error correction method of the optical device, and the processing error correction device of the optical device uses LED irradiation and detection of the optical device to be corrected in multiple orientations and multiple bands to obtain initial detection data containing reflection distribution information, transmission distribution information and interference fringe patterns; based on the data, an error distribution map indicating the coordinate position and the corresponding error type is generated, and each area is ground, polished or partially film-repaired to obtain updated data consisting of the local morphology after repair, film characteristics and coordinate information of the repaired area; then LED irradiation and detection are performed again over the entire surface to obtain a comprehensive correction result. This method ensures that the multi-layer film system and high-precision surface are fully tested, while achieving timely correction of local deviations, and is suitable for mass production of various high-end optical devices.

[0054] above Figure 2 The processing error correction device for optical devices in the embodiment of the present invention is described in detail from the perspective of modular functional entities. The processing error correction device for optical devices in the embodiment of the present invention is described in detail from the perspective of hardware processing.

[0055] Figure 3 3 is a schematic diagram of the structure of a processing error correction device for an optical device provided by an embodiment of the present invention. The processing error correction device 300 for an optical device may have relatively large differences due to different configurations or performances, and may include one or more processors (central processing units, CPU) 310 (for example, one or more processors) and a memory 320, and one or more storage media 330 (for example, one or more mass storage device terminals) storing application programs 333 or data 332. Among them, the memory 320 and the storage medium 330 may be temporary storage or permanent storage. The program stored in the storage medium 330 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations in the processing error correction device 300 for optical devices. Furthermore, the processor 310 may be configured to communicate with the storage medium 330, and execute a series of instruction operations in the storage medium 330 on the processing error correction device 300 for optical devices to implement the steps of the processing error correction method for optical devices described above.

[0056] The optical device processing error correction device 300 may also include one or more power supplies 340, one or more wired or wireless network interfaces 350, one or more input and output interfaces 360, and / or one or more operating systems 331, such as Windows Serve, Mac OS X, Unix, Linux, FreeBSD, etc. It can be understood by those skilled in the art that Figure 3 The structure of the processing error correction device for an optical device shown does not constitute a limitation on the processing error correction device for an optical device provided by the present invention, and may include more or less components than shown in the figure, or a combination of certain components, or a different arrangement of components.

[0057] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions are executed on a computer, the computer executes the steps of the method for correcting processing errors of an optical device.

[0058] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described system, device, or unit can refer to the corresponding process in the aforementioned method embodiment and will not be repeated here.

[0059] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is 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. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program code.

[0060] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for correcting processing errors of optical devices, characterized in that: The processing error correction method of the optical device comprises: The optical device to be corrected is subjected to LED irradiation processing in multiple orientations and multiple wavebands to obtain initial detection data containing reflection distribution information, transmission distribution information and interference fringe patterns; According to the initial detection data, the reference parameters of the optical device are divided into regions and compared with each other to obtain an error distribution map indicating the coordinate position and the corresponding error type; According to the error distribution map, grinding, polishing or local film repairing is performed on the corresponding area of ​​the optical device to obtain updated data including the local morphology after repair, film characteristics and coordinate information of the repaired area; According to the updated data, the optical device is again subjected to LED irradiation and detection processing over the entire surface to obtain a comprehensive correction result that marks the overall reflection change, transmission change and fringe change.

2. The method for correcting processing errors of optical devices according to claim 1, characterized in that: The optical device to be corrected is subjected to LED irradiation processing in multiple orientations and multiple wavebands to obtain initial detection data containing reflection distribution information, transmission distribution information and interference fringe patterns, including: According to the appearance characteristics of the optical device to be corrected, multiple LED sub-light sources are arranged in a ring or matrix to obtain a multi-channel illumination array with sub-bands and sub-angles; According to the preset timing control information, the LED sub-light sources of different wavelength bands or different incident angles in the multi-channel illumination array are sequentially illuminated to obtain the reflection image, transmission image and fringe image of the optical device surface in each wavelength band or orientation; The reflection image, transmission image and fringe image are respectively subjected to data separation processing to obtain initial detection data containing reflection distribution information, transmission distribution information and interference fringe pattern.

3. The method for correcting processing errors of optical devices according to claim 1, characterized in that: The step of performing area division and difference comparison processing on the reference parameters of the optical device according to the initial detection data to obtain an error distribution diagram indicating the coordinate position and the corresponding error type includes: Performing difference calculation processing on the reflection distribution information in the initial detection data and the reference reflection parameters of the optical device to obtain a reflection difference matrix that associates the corresponding optical coordinates with the reflection deviation values; Performing difference calculation processing on the transmission distribution information in the initial detection data and the reference transmission parameters of the optical device to obtain a transmission difference matrix that associates the corresponding optical coordinates with the transmission deviation values; Performing phase resolution processing on the interference fringe pattern in the initial detection data, and performing comparison operation processing in combination with reference fringe parameters to obtain a wavefront offset matrix; Performing multi-dimensional data fusion processing on the reflection difference matrix, the transmission difference matrix and the wavefront deviation matrix to obtain a comprehensive data set including optical coordinates, reflection deviation, transmission deviation and phase change information; The comprehensive data set is processed by partitioning and labeling using a coordinate indexing algorithm to obtain an error distribution map indicating coordinate positions and corresponding error types.

4. The method for correcting processing errors of optical devices according to claim 3, characterized in that: The phase resolution processing is performed on the interference fringe pattern in the initial detection data, and the comparison operation is performed in combination with the reference fringe parameters to obtain the wavefront shift matrix, which includes: Performing pixel-level extraction processing on the fringe peak-valley distribution in the interference fringe pattern to obtain preliminary fringe information including fringe position coordinates and intensity changes; According to the preliminary fringe information, a phase unwrapping process is performed on the fringe image to obtain a phase distribution matrix based on pixel coordinates; Performing difference comparison processing on the phase distribution matrix and the reference fringe parameters to obtain phase difference data with phase offset values ​​associated with fringe pixel coordinates; Performing interpolation fitting processing on the phase difference data to obtain a phase distortion distribution that characterizes a wavefront distortion trend; The phase distortion distribution is combined with the coordinate mapping relationship of the interference image to perform matrix conversion processing to obtain a wavefront shift matrix marked with optical coordinates and phase shift amounts.

5. The method for correcting processing errors of optical devices according to claim 3, characterized in that: The partition labeling process of the comprehensive data set using the coordinate index algorithm to obtain an error distribution map indicating the coordinate position and the corresponding error type includes: Performing coordinate registration processing on the reflection deviation, transmission deviation and phase change information contained in the comprehensive data set to obtain a multi-dimensional error point cloud indexed by a unified optical coordinate system; According to the multi-dimensional error point cloud, each pixel or sampling point is divided into different threshold values ​​to obtain a plurality of error units showing similar deviation characteristics; Performing coordinate clustering processing on the error unit to obtain a cluster center coordinate and a corresponding deviation information index table; Perform label assignment processing on each error unit according to the deviation information index table to obtain a region label list including optical coordinates and error type labels; Visual mapping is performed on the area mark list to obtain an error distribution map indicating coordinate positions and corresponding error types.

6. The method for correcting processing errors of optical devices according to claim 1, characterized in that: According to the error distribution map, grinding, polishing or local film repairing is performed on the corresponding area of ​​the optical device to obtain updated data including the local morphology after repair, film characteristics and coordinate information of the repaired area, including: Determine the area marked as surface geometric error and the area marked as film thickness anomaly in the error distribution map; Performing a cutting depth calculation process on the area marked as the surface geometric error to obtain a cutting depth parameter for the grinding work, and performing a grinding process or a polishing process on the area of ​​the surface geometric error to obtain temporary topography data after the surface flatness is improved; Performing film layer increase / decrease estimation processing on the area marked as abnormal film layer thickness to obtain film layer control instructions for local deposition or removal, and performing thin film increase / decrease operation processing on the area marked as abnormal film layer to obtain local update data containing compensated film layer thickness information; The temporary morphology data and the local update data are integrated to obtain temporary data including the local morphology after trimming and the film layer characteristics, and the coordinate index information is added to the temporary data and a unified marking process is performed to obtain updated data including the local morphology after trimming, the film layer characteristics and the coordinate information of the trimming area.

7. The method for correcting processing errors of optical devices according to claim 1, characterized in that: According to the updated data, the optical device is again subjected to LED irradiation and detection processing over the entire surface, and the comprehensive correction results obtained by marking the overall reflection change, transmission change and fringe change include: According to the trimming area coordinate information in the update data, the entire surface range of the optical device is divided into detection areas to obtain a detection distribution map covering all trimming areas and untrimmed areas; According to the detection distribution diagram, performing sub-band triggering processing on multiple LED sub-light sources to obtain a multi-angle illumination scheme adapted to each coordinate area; Performing full-surface detection processing on the optical device under the multi-angle illumination scheme to obtain secondary sampling data including reflection field, transmission field and interference fringe field information; The secondary sampling data is compared with the trimmed local morphology and film layer features marked in the update data to obtain a comprehensive correction result of the overall reflection change, transmission change and fringe change.

8. A processing error correction device for an optical device, characterized in that: The processing error correction device of the optical device comprises: A multi-source illumination module is used to perform LED illumination processing on the optical device to be corrected in multiple orientations and multiple wavebands to obtain initial detection data containing reflection distribution information, transmission distribution information and interference fringe patterns; A difference comparison module, used to perform area division and difference comparison processing on reference parameters of the optical device according to the initial detection data, and obtain an error distribution map indicating coordinate positions and corresponding error types; A local trimming module, used to perform grinding, polishing or local film repairing on the corresponding area of ​​the optical device according to the error distribution map, and obtain updated data including the local morphology after trimming, film characteristics and coordinate information of the trimming area; The global correction module is used to perform LED irradiation and detection processing on the entire surface of the optical device again according to the updated data, so as to obtain a comprehensive correction result that marks the overall reflection change, transmission change and fringe change.

9. An optical device processing error correction device, characterized in that: The processing error correction device of the optical device comprises: a memory and at least one processor, wherein the memory stores instructions; The at least one processor calls the instructions in the memory to enable the optical device processing error correction device to perform the steps of the optical device processing error correction method according to any one of claims 1 to 7.

10. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by the processor, the steps of the method for correcting processing errors of an optical device as described in any one of claims 1 to 7 are implemented.

Images