Method for detecting optical quality of ultra-slim endoscope based on imaging performance data

Abstract

The application discloses a superfine endoscope optical quality detection method based on imaging performance data and relates to the technical field of optical detection, and comprises the following steps: performing full field of view imaging performance test on a superfine endoscope to generate a field of view record table; locking a field of view position in the field of view record table where the field curvature offset is controlled and the point spread increases, intercepting adjacent front and rear field of view positions, and tracing the point spread expansion direction and the high frequency retention amount change direction in the adjacent front and rear field of view positions; positioning abnormal change jump points that meet the conditions of the field curvature offset being controlled and the point spread continuing to expand at the same time, and generating a flat failure record; extracting the abnormal change jump points from the flat failure record, comparing the change of the high frequency retention amount before and after the abnormal change jump points, and delimiting a clear imaging damaged position where the point spread increases and the high frequency retention amount decreases. The application can enhance the continuity and stability of clear imaging damaged identification by jointly checking the point spread expansion direction and the high frequency retention amount change direction.

Description

Technical Field

[0001] This invention relates to the field of optical inspection technology, and in particular to a method for inspecting the optical quality of ultra-fine endoscopes based on imaging performance data. Background Technology

[0002] With the development of minimally invasive diagnostic and treatment technologies, high-resolution CMOS image sensors, and ultra-miniature optical processing capabilities, ultra-thin endoscopes are gradually evolving towards smaller outer diameters, wider fields of view, higher resolution, and more consistent imaging. As the core component responsible for the imaging quality of ultra-thin endoscopes, the front-end optical components need to maintain stable image plane flatness, blur control, and high-frequency detail retention under conditions of limited aperture, short optical length, and complex fields of view. To meet the requirements of ultra-thin endoscope assembly evaluation, factory testing, and quality grading, optical quality testing based on imaging performance data has gradually become an important technical direction. Related testing typically revolves around full-field imaging performance, quantitatively characterizing the imaging state at different field-of-view positions to determine whether the ultra-thin endoscope optical components possess stable, clear, and consistent imaging capabilities.

[0003] Current optical inspection methods for ultra-fine endoscopes often focus on judging single indicators, such as field curvature, dot plots, or modulation transfer function curves. This makes it difficult to establish a continuous correlation between image plane shift, diffusion expansion, and high-frequency attenuation at the same field of view. When the field curvature is still under control but the dot plot diffusion continues to expand, existing inspection methods cannot easily identify the location of clear imaging damage, resulting in insufficient basis for optical quality level classification. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides an optical quality detection method for ultra-fine endoscopes based on imaging performance data, which solves the problem of establishing a continuous correlation between image plane shift, diffusion amplification, and high-frequency attenuation at the same field of view.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] This invention provides a method for detecting the optical quality of ultra-thin endoscopes based on imaging performance data, comprising:

[0008] Full-field imaging performance tests were performed on the ultra-fine endoscope to obtain field curvature shift, point array diffusion, and high frequency retention. The field curvature shift, point array diffusion, and high frequency retention were then grouped into the same field position to generate a field record table.

[0009] Lock the field position in the field record table where the field curvature offset is controlled and the point dispersion is expanding, extract the adjacent field positions before and after, trace the expansion direction of the point dispersion and the change direction of the high frequency holding amount in the adjacent field positions before and after, locate the abnormal change jump point that simultaneously satisfies the control of field curvature offset and the continued expansion of point dispersion, and generate a flat failure record.

[0010] Extract anomalous change jump points from the flat failure record, compare the changes in high-frequency retention before and after the anomalous change jump points, delineate the clear imaging damage locations where the point column dispersion increases and the high-frequency retention decreases, and determine the source of damage;

[0011] Based on the continuous distribution of damaged locations and the source of damage in clear imaging, the optical quality level of the ultrafine endoscope is classified, and optical detection results are generated.

[0012] As a preferred embodiment of the ultra-fine endoscope optical quality detection method based on imaging performance data described in this invention, wherein obtaining the field curvature shift, point dispersion, and high-frequency retention specifically involves...

[0013] The field of view of the ultra-fine endoscope is divided into a central field of view, a transitional field of view, and a peripheral field of view along the direction of field of view development.

[0014] The test images of the central field of view, the transition field of view, and the edge field of view are acquired and arranged in the order of field of view expansion of the central field of view, the transition field of view, and the edge field of view to form a full field of view test image.

[0015] Defocus scanning was performed on the center field-of-view test image, transition field-of-view test image and edge field-of-view test image in the full field-of-view test imaging, and the sharp peak position was recorded at different image plane positions. The field curvature curve was plotted based on the sharp peak position.

[0016] The sharp peak position corresponding to the test image of the center field of view is used as the field curvature reference position. The zero deviation distance of the center field of view relative to the field curvature reference position, as well as the deviation distance of the transition field of view and the edge field of view relative to the field curvature reference position, are extracted from the field curvature curve. The zero deviation distance and the deviation distance are combined into the field curvature offset.

[0017] Multi-wavelength point image test beams are projected into the central field of view, the transition field of view, and the edge field of view to obtain the distribution of imaging points at different wavelengths;

[0018] The outer boundaries of the imaging points at different wavelengths on the image plane are delineated, and the outer boundaries in the central field of view, the transition field of view, and the edge field of view are combined into a point series of diffusion quantities.

[0019] In the central field-of-view test imaging, transition field-of-view test imaging, and edge field-of-view test imaging, the oblique transition zone is extracted, and gray-scale sampling lines are drawn within the oblique transition zone according to the sampling direction perpendicular to the oblique boundary.

[0020] Collect the pixel grayscale values ​​of the dark areas, transition areas and bright areas that the grayscale sampling line passes through to form a grayscale transition sequence;

[0021] The grayscale transition sequence is converted into a modulation transfer function curve, and the curve value at a high spatial frequency is selected from the modulation transfer function curve.

[0022] The curve values ​​of the center field of view, transition field of view, and edge field of view are combined into a high-frequency retention value.

[0023] As a preferred embodiment of the ultra-fine endoscope optical quality detection method based on imaging performance data described in this invention, wherein: the generation of the field of view record table specifically involves...

[0024] Arrange the central field of view, transitional field of view, and edge field of view in the order of field of view unfolding to form a field of view position sequence;

[0025] Based on the field position sequence, the field curvature offset is divided into field curvature terms under the central field of view, transition field of view, and edge field of view; the point series diffusion is divided into point series terms under the central field of view, transition field of view, and edge field of view; and the high frequency retention is divided into high frequency terms under the central field of view, transition field of view, and edge field of view.

[0026] The field curvature, field point, and high-frequency terms of the field of view at the same field of view position are grouped into the same field of view position to form the field of view imaging performance terms.

[0027] Remove any field of view imaging performance items that are missing any of the field curvature, field of view point, or field of view high frequency items, and arrange the remaining field of view imaging performance items according to the field of view position sequence to generate a field of view record table.

[0028] As a preferred embodiment of the ultra-fine endoscope optical quality detection method based on imaging performance data described in this invention, wherein: the step of truncating adjacent field-of-view positions before and after the truncated position specifically involves...

[0029] Traverse the field of view imaging performance items in the field of view record table, and retain the field of view imaging performance items that have both adjacent positions in the central field of view direction and adjacent positions in the edge field of view direction as the field of view imaging performance items to be locked.

[0030] Extract the field curvature term to be locked from the imaging performance terms of the field of view to be locked, extract the center side field curvature term from the adjacent position in the center field of view direction, and extract the edge side field curvature term from the adjacent position in the edge field of view direction;

[0031] The coverage area between the central side field curvature item and the edge side field curvature item is taken as the field curvature coverage area. The field curvature item to be locked that falls into the field curvature coverage area is marked as the field curvature controlled item.

[0032] The field point column in the field curvature controlled term is taken as the controlled point column, and the field point column in the adjacent position in the central field direction is taken as the central side point column. The outer expansion boundary of the controlled point column and the central side point column is compared to form the point column expansion result.

[0033] When the point column expansion result is that the expansion is valid, the field position of the field curve controlled term is taken as the target field position;

[0034] Locate the position of the target's field of view in the field of view position sequence, and take the two field of view positions immediately before and after the current position as the adjacent field of view positions.

[0035] As a preferred embodiment of the ultra-fine endoscope optical quality detection method based on imaging performance data described in this invention, the "outward expansion" means that the outward expansion boundary of the controlled point column moves outward toward the edge field of view compared to the outward expansion boundary of the center side point column, and the diffusion range formed after the outward movement is greater than the diffusion range of the center side point column.

[0036] As a preferred embodiment of the ultra-fine endoscope optical quality detection method based on imaging performance data described in this invention, wherein: the generation of flat failure records specifically involves...

[0037] Arrange the target's field of view position and the positions of its adjacent fields of view in a field of view position sequence to form a local field of view segment;

[0038] In a local field of view segment, the field point list items under the adjacent field of view positions and the target field of view position are connected in the order of field of view expansion to form a point list boundary chain;

[0039] In the point column boundary chain, the outer expansion boundary of the adjacent field of view position and the target field of view position is compared. When the outer expansion boundary of the adjacent field of view position and the target field of view position continuously moves outward towards the edge field of view, the point column diffusion expansion direction is formed.

[0040] In a local field of view segment, the high-frequency terms of the field of view at adjacent field of view positions and at the target field of view position are connected in series according to the field of view unfolding order to form a high-frequency holding chain;

[0041] In the high-frequency holding chain, the curve values ​​of adjacent field-of-view positions and the target field-of-view position are compared. When the curve value from the target field-of-view position to the edge field-of-view direction turns downward, the direction of change of the high-frequency holding amount is formed.

[0042] The controlled terms of field curvature, the expansion direction of point-series dispersion, and the change direction of high-frequency holding quantity are merged into the same traceability entry;

[0043] When the same tracing entry contains field curvature controlled items, the point dispersion continues to expand along the point dispersion expansion direction, and the high frequency holding quantity turns to decrease along the high frequency holding quantity change direction, the target field position is determined as the anomalous change jump point.

[0044] The abnormal change jump point, local field segment, point series diffusion expansion direction and high frequency retention change direction are merged to generate a flat failure record.

[0045] As a preferred embodiment of the ultra-fine endoscope optical quality detection method based on imaging performance data described in this invention, the comparison of the changes in high-frequency retention before and after the anomalous change jump point specifically involves:

[0046] Extract the anomalous change jump points, local field segments, the expansion direction of point-series dispersion, and the change direction of high-frequency retention from the flat failure record;

[0047] Using the anomalous change jump point as the boundary, the local field of view is divided into a pre-jump field of view segment and a post-jump field of view segment. The field of view position of the pre-jump field of view segment adjacent to the anomalous change jump point is taken as the pre-jump position, and the anomalous change jump point is taken as the jump point position. The field of view position of the post-jump field of view segment adjacent to the anomalous change jump point is taken as the post-jump position.

[0048] Compare the outer expansion boundary of the point array diffusion at the pre-jump position, the jump point position, and the post-jump position. When the outer expansion boundary of the point array diffusion increases sequentially from the pre-jump position to the jump point position and then to the post-jump position, the field of view segment composed of the pre-jump position, the jump point position, and the post-jump position is defined as the point array expansion duration segment.

[0049] Compare the high-frequency sustain curve values ​​of the pre-jump position, the jump point position, and the post-jump position. When the high-frequency sustain curve value decreases sequentially from the pre-jump position to the jump point position and then to the post-jump position, the field of view segment composed of the pre-jump position, the jump point position, and the post-jump position is classified as the high-frequency descent sustained segment.

[0050] As a preferred embodiment of the ultra-fine endoscope optical quality detection method based on imaging performance data described in this invention, the location of impaired clear imaging where the diffusion of the delineated point array is amplified and the high-frequency retention is reduced specifically refers to...

[0051] Project the field position sequence of the dot-array expansion duration and the high-frequency descent duration onto the local field of view segment, and take the field position jointly covered by the dot-array expansion duration and the high-frequency descent duration as the damaged overlapping position;

[0052] The field of view positions covering the jump point in the damaged overlapping positions are defined as the clear imaging damaged positions, and the field of view positions falling into the post-jump field of view in the damaged overlapping positions are classified as the clear imaging damaged positions.

[0053] As a preferred embodiment of the ultrafine endoscope optical quality detection method based on imaging performance data described in this invention, the method for determining the source of damage involves placing the starting field position of the extended dot-array expansion continuous segment and the starting field position of the extended high-frequency descent continuous segment into the field position sequence, and determining the source of damage based on the sequential relationship between the two starting field positions in the field position sequence.

[0054] As a preferred embodiment of the ultrafine endoscope optical quality detection method based on imaging performance data described in this invention, the optical quality level includes an acceptable optical quality level, a critical optical quality level, and a substandard optical quality level.

[0055] The beneficial effects of this invention are as follows: By jointly verifying the expansion direction of the point-series dispersion and the change direction of the high-frequency retention, the continuity and stability of the identification of damage to clear imaging can be enhanced. When the field curvature offset is still under control, the expansion of the point-series dispersion and the decrease of the high-frequency retention can jointly reflect the loss of image detail. The flat failure record concentrates the flat field curvature state, the outward expansion state of the point-series dispersion, and the decrease of the high-frequency retention state into the same field of view, so that the abnormal change jump point has a clearer field of view positioning basis; the location of damage to clear imaging no longer depends on the instantaneous fluctuation of a single index, but is delineated based on the continuous change in the field of view expansion direction, which helps to reduce missed judgments and false judgments, improves the verifiability of optical quality level classification, and makes the optical detection results of the ultra-fine endoscope more accurately reflect the imaging quality of the entire field of view. Attached Figure Description

[0056] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0057] Figure 1 This is a flowchart of a method for detecting the optical quality of ultra-fine endoscopes based on imaging performance data.

[0058] Figure 2 Generate a relationship diagram for the field of view record table.

[0059] Figure 3 A graph is generated to represent the flat failure record.

[0060] Figure 4 This is a diagram showing the classification of optical quality levels. Detailed Implementation

[0061] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0062] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0063] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0064] Reference Figures 1-4 This is one embodiment of the present invention, which provides a method for detecting the optical quality of an ultra-thin endoscope based on imaging performance data, including the following steps:

[0065] S1. Perform full-field imaging performance test on the ultra-fine endoscope to obtain field curvature offset, point array diffusion, and high frequency retention. Assign the field curvature offset, point array diffusion, and high frequency retention to the same field position and generate a field record table.

[0066] S1.1. Divide the field of view along the unfolding direction of the ultra-fine endoscope into a central field of view, a transitional field of view, and a peripheral field of view, specifically as follows:

[0067] The field of view unfolding direction is determined by the radial arrangement of the center point of the ultra-fine endoscope imaging test target surface to the boundary point of the imaging test target surface. The field of view unfolding direction is derived from the corresponding position of the center point of the ultra-fine endoscope imaging test target surface and the boundary point of the imaging test target surface.

[0068] Multiple field-of-view positions arranged continuously along the field-of-view expansion direction near the center point of the imaging test target are designated as the central field of view; multiple field-of-view positions arranged continuously along the field-of-view expansion direction near the boundary point of the imaging test target are designated as the edge field of view; and multiple field-of-view positions arranged continuously along the field-of-view expansion direction between the central and edge fields of view are designated as the transition field of view. For example, when different ultra-fine endoscopes have differences in the full field of view, image height, distortion, and effective imaging circle, the central field of view, transition field of view, and edge field of view can be divided according to one of the following: the ratio of the full field of view ratio and the ratio of the image height ratio. The boundary ratio of the central field of view... For example, α and the transition field of view boundary ratio β are set as adjustable parameters, where 0 < α < β < 1. Multiple field of view positions with a normalized field of view position no greater than α are classified as the central field of view, multiple field of view positions with a normalized field of view position greater than α and no greater than β are classified as the transition field of view, and multiple field of view positions with a normalized field of view position greater than β and no greater than 1 are classified as the edge field of view. For example, when α is 0.35 and β is 0.70, in an ultra-fine endoscope with a single-sided field of view angle of 60°, 0° to 21° corresponds to the central field of view, greater than 21° to 42° corresponds to the transition field of view, and greater than 42° to 60° corresponds to the edge field of view.

[0069] S1.2 Acquire test images of the central field of view, transitional field of view, and edge field of view, and arrange the test images in the order of field of view expansion of the central field of view, transitional field of view, and edge field of view to form a full field of view test image. Specifically,

[0070] Images of the imaging test target surface in the central field of view, transition field of view, and edge field of view are acquired using a CMOS image sensor to obtain test images of the central field of view, transition field of view, and edge field of view. The test object is an ultra-fine endoscope optical component under test with a completed lens assembly and an imaging surface that can be connected to the CMOS image sensor for testing. The CMOS image sensor for testing is fixed on an optical axis displacement stage, and the photosensitive surface of the CMOS image sensor for testing serves as the imaging surface of the ultra-fine endoscope optical component under test. It is used to subsequently change the image surface position along the optical axis and acquire test images at different image surface positions.

[0071] The field of view unfolding sequence is radially arranged from the center point of the ultra-fine endoscope imaging test target surface to the boundary point of the imaging test target surface, in the following order: central field of view, transitional field of view, and edge field of view. The field of view unfolding direction is determined by the direction from the center point of the ultra-fine endoscope imaging test target surface to the boundary point of the imaging test target surface. The central field of view corresponds to the starting position, the transitional field of view corresponds to the intermediate continuous position, and the edge field of view corresponds to the ending position. This is beneficial for arranging the field curvature offset, point array diffusion, and high frequency retention according to the field of view change relationship from the center to the edge, which facilitates the subsequent identification of the continuous changes in point array diffusion expansion and high frequency retention decrease.

[0072] S1.3. Perform defocus scanning on the center field-of-view test image, transition field-of-view test image, and edge field-of-view test image in the full field-of-view test imaging, record the sharp peak position at different image plane positions, and plot the field curvature curve based on the sharp peak position. Specifically,

[0073] When performing defocus scanning for central field-of-view test imaging, transitional field-of-view test imaging, and edge field-of-view test imaging, the optical axis displacement stage carrying the CMOS image sensor moves gradually along the optical axis direction. The optical components under test are mounted on the test fixture. The CMOS acts as an external detection receiver, and the displacement readings of the optical axis displacement stage serve as the source of different image plane positions.

[0074] The CMOS image sensor acquires grayscale images of the edge of the target surface line in the central field of view, transition field of view, and edge field of view at each image plane position, and arranges them according to displacement readings into a central field of view defocus scanning imaging sequence, a transition field of view defocus scanning imaging sequence, and an edge field of view defocus scanning imaging sequence.

[0075] In the central field-of-view defocus scanning imaging sequence, the transition field-of-view defocus scanning imaging sequence, and the edge field-of-view defocus scanning imaging sequence, a pre-scan is performed by a CMOS image sensor for detection along the optical axis. The pre-scan step size is larger than the formal defocus scanning step size. The pre-scan covers image plane positions that can form complete dark areas, transition areas, and bright areas. Among the multiple image plane positions obtained by the pre-scan, the gray-scale transition width of the edge of the imaging test target line is counted one by one, and the image plane position with the smallest gray-scale transition width is taken as the initial imaging reference position. The defocus scanning range is set with the initial imaging reference position as the center. The defocus scanning range covers the clear search distance on both sides of the initial imaging reference position. The clear search distance is determined according to the optical axis displacement distance corresponding to the sum of the minimum gray-scale transition width and the width change judgment amount in the pre-scan. The width change judgment amount is determined according to the fluctuation of the gray-scale transition width difference between adjacent image plane positions in the pre-scan, for example, taking twice the average absolute value of the gray-scale transition width difference between adjacent image plane positions.

[0076] The optical axis displacement stage stops gradually with a step size no greater than one-tenth of the clear search distance, for example, a step size of 0.005mm to 0.02mm, and continuously acquires multiple frames of grayscale images at each image plane position; performs dark field subtraction, brightness normalization, and median filtering on the multiple frames of grayscale images, and then takes the average image of the processed multiple frames of grayscale images to suppress isolated noise pixels and random brightness fluctuations; read the grayscale profile along the direction perpendicular to the edge of the imaging test target surface line, take the 10% grayscale position between the average grayscale of the dark area and the average grayscale of the bright area as the low limit grayscale, and take the 90% grayscale position between the average grayscale of the dark area and the average grayscale of the bright area as the high limit grayscale, and take the number of pixels that cross from the low limit grayscale position to the high limit grayscale position in the grayscale profile as the pixel width.

[0077] Among multiple image plane positions corresponding to the same field of view, the image plane position with the smallest pixel width and where the pixel width of adjacent image plane positions changes from a continuous decrease to a continuous increase is recorded as the sharp peak position. If the pixel width values ​​of two consecutive image plane positions are the same, the middle position of the two consecutive image plane positions is recorded as the sharp peak position. The sharp peak positions corresponding to multiple field of view positions in the central field of view, transitional field of view, and edge field of view are connected in the order of field of view expansion to plot the field curvature curve. The field curvature curve is used to represent the deviation relationship of the optimal imaging image plane position of different field of view positions along the field of view expansion direction.

[0078] S1.4. Using the sharp peak position corresponding to the test image of the center field of view as the field curvature reference position, extract the zero deviation distance of the center field of view relative to the field curvature reference position, as well as the deviation distances of the transition field of view and the edge field of view relative to the field curvature reference position from the field curvature curve. Combine the zero deviation distance and the deviation distance to form the field curvature offset. Specifically,

[0079] Read the sharp peak position corresponding to the test image of the center field of view in the field curvature curve, and take the optical axis displacement stage displacement reading corresponding to the sharp peak position as the reference reading of the field curvature reference position; the sharp peak position of the center field of view and the field curvature reference position correspond to the same image plane position, and the distance of the center field of view relative to the field curvature reference position is recorded as the zero deviation distance.

[0080] Next, read the displacement readings corresponding to the sharp peak position in the transition field of view and the sharp peak position in the edge field of view from the field curvature curve. Subtract these values ​​from the reference readings to obtain the deviation distances of the transition field of view relative to the field curvature reference position and the deviation distances of the edge field of view relative to the field curvature reference position. A positive difference indicates that the sharp peak position is moving away from the field curvature reference position along the positive direction of the optical axis. A negative difference indicates that the sharp peak position is moving away from the field curvature reference position along the negative direction of the optical axis. A zero difference indicates that the sharp peak position coincides with the field curvature reference position.

[0081] The zero offset distance, transition field offset distance, and edge field offset distance are arranged in the order of field unfolding of the central field of view, transition field of view, and edge field of view, and combined into the field curvature offset.

[0082] The expression for the field curvature offset is:

[0083] ;

[0084] In the formula, Indicates the first The deviation distance of each field of view position from the field curvature reference position. Indicates the first Optical axis displacement stage displacement readings corresponding to the clear peak position at each field of view location. The reference reading indicates the field curvature reference position, which is derived from the sharp peak position corresponding to the central field-of-view test image. Indicates the position of one of the central field of view, transitional field of view, and edge field of view, when When corresponding to the central field of view, .

[0085] S1.5. Project multi-wavelength point image test beams into the central field of view, transition field of view, and edge field of view respectively to obtain the distribution of imaging points at different wavelengths. Further...

[0086] By using a multi-wavelength point image test light source in conjunction with a point aperture target, point beams corresponding to, for example, F-beams (e.g., 486.1nm), d-beams (e.g., 587.6nm), and C-beams (e.g., 656.3nm) are projected onto the imaging test target positions corresponding to the central field of view, transition field of view, and edge field of view, respectively.

[0087] The multi-wavelength point image test beam is a point-shaped incident beam formed by the multi-wavelength point image test light source after being defined by a point aperture target. It is used to form different wavelength imaging point distributions on a CMOS image sensor.

[0088] S1.6. Delineate the outer boundaries of imaging points at different wavelengths on the image plane, and combine the outer boundaries in the central field of view, transition field of view, and edge field of view into a point-series dispersion measure, specifically,

[0089] In the grayscale images of the points corresponding to different wavelength imaging point distributions, continuous bright spot regions with imaging point grayscale values ​​higher than the dark background grayscale are identified according to wavelength. When multiple local peak positions exist for the same wavelength, these multiple local peak positions are recorded together as the peak position of the corresponding wavelength. The background grayscale is taken from the dark background region outside the continuous bright spot region of the corresponding wavelength that is not covered by the imaging points. The intermediate grayscale value between the background grayscale value and the highest grayscale value among the peak positions of the corresponding wavelength is taken as the half-peak boundary value of the corresponding wavelength. The half-peak boundary value is derived from the background grayscale value and the highest grayscale value among the peak positions of the points. The half-peak boundary value is used to delineate the continuous boundary positions where the grayscale value reaches the half-peak boundary value in the grayscale image of the corresponding wavelength, and forms an extended boundary in the same image plane coordinate that can surround the entire wavelength imaging point distribution.

[0090] Starting from the peak position of the corresponding wavelength, search outwards along multiple radial directions of the image plane for continuous boundary positions where the grayscale reaches the half-peak boundary value of the image. Connect the continuous boundary positions corresponding to multiple directions to form the half-peak closed boundary of the corresponding wavelength. Before forming the half-peak closed boundary, first acquire a dark field image after turning off the multi-wavelength image test light source, perform dark field subtraction on the grayscale image of the image, and determine the background grayscale and background fluctuation judgment amount by the dark background area outside the continuous bright spot area of ​​the corresponding wavelength. When continuing to screen pixels outside the half-peak closed boundary, only retain continuous pixel areas with grayscale higher than the background grayscale and exceeding the background fluctuation judgment amount, and remove isolated bright spots with a pixel count lower than the minimum connected area. The minimum connected area is set by the number of pixels in the continuous bright spot area around the peak position of the image, for example, 5% of the number of pixels in the continuous bright spot area. The outermost continuous pixel line that is still continuously connected to the half-peak closed boundary after completing dark field subtraction, background fluctuation screening, and minimum connected area removal is used as the wavelength expansion boundary of the corresponding wavelength. Superimpose multiple wavelength expansion boundaries in the same image plane coordinate to form an expansion boundary that can surround all wavelength expansion boundaries.

[0091] The outer boundaries of the imaging points of different wavelengths in the central field of view, the transition field of view, and the edge field of view are delineated respectively, resulting in the outer boundaries of the central field of view, the transition field of view, and the edge field of view. These boundaries are then arranged in the order of field expansion of the central field of view, the transition field of view, and the edge field of view, and combined into a point-series diffusion quantity.

[0092] The expression for the point-series dispersion is:

[0093] ;

[0094] In the formula, Indicates the first The point array dispersion at each field of view location, Indicates the first The number of pixels enclosed by the outer boundary at each field of view location. This represents the area of ​​a single pixel in a CMOS image sensor. The larger the value, the larger the area enclosed by the outer boundary of the imaging points of different wavelengths on the image plane.

[0095] S1.7. Extract the oblique transition zone from the center field-of-view test image, the transition field-of-view test image, and the edge field-of-view test image. Within the oblique transition zone, delineate grayscale sampling lines according to the sampling direction perpendicular to the oblique boundary. Specifically,

[0096] The oblique edge boundary originates from the boundary line between the dark and bright areas formed after the oblique line edge on the imaging test target surface is imaged by an ultra-fine endoscope.

[0097] Within the central field-of-view test imaging, transition field-of-view test imaging, and edge field-of-view test imaging, the boundary line between the dark and bright areas is identified. A strip-shaped region containing the dark area, transition area, and bright area is extracted around the boundary line to form a slanted transition zone. The sampling direction is determined according to the normal direction of the slanted boundary. The grayscale sampling line extends from the dark area through the transition area to the bright area and remains perpendicular to the slanted boundary, thus obtaining the slanted transition zone of the central field of view, the slanted transition zone of the transition field of view, the slanted transition zone of the edge field of view, and the corresponding grayscale sampling line.

[0098] S1.8. Collect the pixel grayscale values ​​of the dark, transition, and bright areas traversed by the grayscale sampling line to form a grayscale transition sequence; convert the grayscale transition sequence into a modulation transfer function curve, and select the curve values ​​at high spatial frequencies from the modulation transfer function curve; combine the curve values ​​of the center field of view, transition field of view, and edge field of view into a high-frequency retention value, specifically,

[0099] It should be noted that the high spatial frequency is the nominal limiting resolution frequency of the optical component under test in the ultra-fine endoscope; if the nominal limiting resolution frequency is missing, the high spatial frequency is the spatial frequency corresponding to the Nyquist frequency of the CMOS image sensor used for testing, and the selected high spatial frequency is recorded along with the field of view recording table.

[0100] The test images formed by the CMOS image sensor, namely the central field-of-view test image, the transition field-of-view test image, and the edge field-of-view test image, read the pixel grayscale values ​​point by point from the dark area through the transition area to the bright area along the grayscale sampling lines in the central field-of-view test image, the transition field-of-view test image, and the edge field-of-view test image, respectively. The grayscale values ​​of the pixels are then read point by point according to the position of the grayscale sampling lines, forming the central field-of-view grayscale transition sequence, the transition field-of-view grayscale transition sequence, and the edge field-of-view grayscale transition sequence. The central field-of-view grayscale transition sequence, the transition field-of-view grayscale transition sequence, and the edge field-of-view grayscale transition sequence respectively represent the grayscale changes of the corresponding oblique transition zone from the dark area to the bright area.

[0101] Using the sampling positions of the grayscale sampling lines in the center field of view grayscale transition sequence, the transition field of view grayscale transition sequence, and the edge field of view grayscale transition sequence as the abscissa, and the pixel grayscale value after normalization of dark and bright area grayscale values ​​as the ordinate, edge expansion curves of the center field of view, the transition field of view, and the edge field of view are plotted. The normalized pixel grayscale values ​​at adjacent sampling positions in each of these curves are then subtracted segment by segment to obtain the center field of view line expansion curve, the transition field of view line expansion curve, and the edge field of view line expansion curve. Fourier transforms are then performed on each of these curves, and the amplitudes after the Fourier transforms are normalized to zero-frequency amplitude to obtain the modulation transfer function curves of the center field of view, the transition field of view, and the edge field of view, respectively.

[0102] Locate the abscissa position corresponding to the high spatial frequency in the modulation transfer function curve of the center field of view, and read the ordinate value of the same abscissa position as the center field of view curve value; locate the abscissa position corresponding to the high spatial frequency in the modulation transfer function curve of the transition field of view, and read the ordinate value of the same abscissa position as the transition field of view curve value; locate the abscissa position corresponding to the high spatial frequency in the modulation transfer function curve of the edge field of view, and read the ordinate value of the same abscissa position as the edge field of view curve value.

[0103] The center field-of-view curve values, transition field-of-view curve values, and edge field-of-view curve values ​​are arranged in the order of field-of-view expansion and combined into high-frequency retention values.

[0104] The expression for the high-frequency retention amount is:

[0105] ;

[0106] In the formula, Indicates the first High-frequency retention at each field of view position Indicates the first The line spread curves at each field of view location, after Fourier transform, are at high spatial frequencies. The amplitude at that point, Indicates the first The frequency domain function obtained by Fourier transforming the line spread curves at each field of view location. Indicates the first The zero-frequency amplitude of the line spread curve at each field of view position after Fourier transform. This represents the high spatial frequency selected in the modulation transfer function curve. The larger the value, the higher the value. The better the field of view position is at which high spatial frequency imaging details are preserved.

[0107] S1.9 Arrange the central field of view, transition field of view, and edge field of view according to the field of view unfolding order to form a field of view position sequence; based on the field of view position sequence, divide the field curvature offset into field curvature terms under the central field of view, transition field of view, and edge field of view; divide the point series blur into point series terms under the central field of view, transition field of view, and edge field of view; divide the high-frequency retention into high-frequency terms under the central field of view, transition field of view, and edge field of view. Specifically,

[0108] The field of view position sequence is fixed as the central field of view, transitional field of view, and edge field of view according to the field of view unfolding order; in the field curvature offset, the zero offset distance is included in the field curvature term of the central field of view, the offset distance of the transitional field of view is included in the field curvature term of the transitional field of view, and the offset distance of the edge field of view is included in the field curvature term of the edge field of view.

[0109] In the point-series diffusion, the outer boundary of the central field of view is included in the field-series item under the central field of view, the outer boundary of the transition field of view is included in the field-series item under the transition field of view, and the outer boundary of the edge field of view is included in the field-series item under the edge field of view.

[0110] In the high-frequency retention, the central field of view curve value is assigned to the high-frequency term of the field of view under the central field of view, the transition field of view curve value is assigned to the high-frequency term of the field of view under the transition field of view, and the edge field of view curve value is assigned to the high-frequency term of the field of view under the edge field of view. This completes the corresponding division of field curvature offset, point array diffusion, and high-frequency retention in the central field of view, transition field of view, and edge field of view.

[0111] S1.10. The field curvature, field point, and high-frequency fields under the same field of view position are grouped into the same field of view position to form the field of view imaging performance. The same field of view position is the position in the field of view position sequence that belongs to the same imaging area in the central field of view, transitional field of view, and edge field of view. For example, the field curvature, field point, and high-frequency fields under the central field of view correspond to the same central field of view position.

[0112] S1.11. Remove any field-of-view imaging performance items that are missing any of the field curvature, field-of-view point, or field-of-view high-frequency items, and arrange the remaining field-of-view imaging performance items according to the field-of-view position sequence to generate a field-of-view record table. Specifically,

[0113] Each field of view imaging performance item is checked to see if it simultaneously includes the field curvature, field dot series, and high-frequency field of view items. Field of view imaging performance items that simultaneously include the field curvature, field dot series, and high-frequency field of view items are retained as complete field of view imaging performance items. Field of view imaging performance items that lack any one of the field curvature, field dot series, and high-frequency field of view items are marked as missing field of view imaging performance items and removed. Complete field of view imaging performance items are arranged according to the field of view position sequence of the central field of view, transitional field of view, and edge field of view. During the arrangement process, the field curvature, field dot series, and high-frequency field of view items corresponding to the central field of view are all assigned to the central field of view position, the field curvature, field dot series, and high-frequency field of view items corresponding to the transitional field of view are all assigned to the transitional field of view position, and the field curvature, field dot series, and high-frequency field of view items corresponding to the edge field of view are all assigned to the edge field of view position, generating a field of view record table.

[0114] S2. Lock the field position in the field record table where the field curvature offset is controlled and the point dispersion is expanding. Extract the adjacent field positions before and after, and trace the expansion direction of the point dispersion and the change direction of the high frequency holding amount in the adjacent field positions before and after. Locate the abnormal change jump point that simultaneously satisfies the controllable field curvature offset and the continued expansion of the point dispersion, and generate a flat failure record.

[0115] Among them, field curvature offset control means that the field curvature item to be locked falls within the field curvature coverage area between the center-side field curvature item and the edge-side field curvature item, and the field curvature offset corresponding to the field curvature item to be locked, the center-side field curvature item and the edge-side field curvature item are all not greater than the field curvature control threshold, indicating that the field curvature change at the corresponding field position has not exceeded the limit deviation.

[0116] S2.1 Traverse the field of view imaging performance items in the field of view record table, retain the field of view imaging performance items that have adjacent positions in both the central field of view direction and the edge field of view direction, and use them as field of view imaging performance items to be locked; extract the field curvature items to be locked from the field of view imaging performance items to be locked, extract the central side field curvature items from the adjacent positions in the central field of view direction, and extract the edge side field curvature items from the adjacent positions in the edge field of view direction; take the coverage area between the central side field curvature items and the edge side field curvature items as the field curvature coverage area, and mark the field of view imaging performance items to be locked that fall into the field curvature coverage area as field curvature controlled items.

[0117] Among them, the field curvature control item is not only determined by the field curvature item to be locked falling into the field curvature coverage area, but also needs to be confirmed a second time in combination with the field curvature control threshold; the field curvature item to be locked is the field curvature offset corresponding to the current field position to be judged in the field imaging performance item to be locked; the center-side field curvature item is the field curvature offset corresponding to the center-side adjacent field position in the adjacent position in the center field direction; the edge-side field curvature item is the field curvature offset corresponding to the edge-side adjacent field position in the adjacent position in the edge field direction; the field curvature control threshold is based on the depth of field of the ultrafine endoscope corresponding to the allowable image plane offset and the lens structure tolerance corresponding to the allowable The permissible image plane offset and the permissible MTF peak offset are determined, and the minimum value among the permissible image plane offset corresponding to depth of field, the permissible image plane offset corresponding to lens structure tolerance, and the permissible MTF peak offset is taken as the judgment criterion; when the field curvature offset of the center side field curvature item, the edge side field curvature item, and the field curvature item to be locked is not greater than the field curvature control threshold, the field of view imaging performance item to be locked is retained as a field curvature control item; when the field curvature offset of any of the center side field curvature item, the edge side field curvature item, and the field curvature item to be locked is greater than the field curvature control threshold, the field of view imaging performance item to be locked is removed from the field curvature control items.

[0118] S2.2. Take the field point entries in the controlled field curvature terms as controlled point entries, and take the field point entries in the adjacent positions in the central field of view direction as central side point entries. Compare the outer boundaries of the controlled point entries and the central side point entries to form the point entry expansion result. Specifically,

[0119] The outer boundaries of the controlled point column and the outer boundaries of the center side point column are placed in the same image plane coordinate system. Using the pixel position corresponding to the center point of the imaging test target as a common reference, azimuth sampling lines are set one by one according to the registered field-of-view directions from the center field of view to the edge field of view. Radial outer pixel line, tangential outer pixel line, and sagittal outer pixel line are read from each azimuth sampling line. The controlled point column is shifted outward relative to the center side point column towards the edge field of view in the radial outer pixel line, tangential outer pixel line, and sagittal outer pixel line, and the controlled point column expands outward. When the number of pixels enclosed by the boundary is greater than the number of pixels enclosed by the outer boundary of the center side point column, the point column expansion result is recorded as expansion successful; when radially outer pixel lines, tangentially outer pixel lines, and sagittal outer pixel lines in any registered field of view do not move outward synchronously, or when the number of pixels enclosed by the outer boundary of the controlled point column is not greater than the number of pixels enclosed by the outer boundary of the center side point column, the point column expansion result is recorded as expansion unsuccessful, the field of view position where the field curvature controlled item is located is removed from the target field of view position, and the capture of adjacent field of view positions before and after the field of view position where the field curvature controlled item is located is stopped.

[0120] When the outer pixel line of the controlled point column expansion boundary is located at the edge field of view of the outer pixel line of the center point column expansion boundary, and the number of pixels enclosed by the controlled point column expansion boundary is greater than the number of pixels enclosed by the center point column expansion boundary, the point column expansion result is recorded as expansion is successful. The field of view position where the field curvature controlled item is located is retained as the target field of view position, and the adjacent field of view positions before and after the target field of view position are continued to be captured.

[0121] If any of the following occurs: the outer pixel line of the controlled point column expansion boundary does not move outward towards the edge field of view, or the number of pixels enclosed by the controlled point column expansion boundary is not more than the number of pixels enclosed by the center side point column expansion boundary, the point column expansion result is recorded as expansion not valid, and the field of view position where the field curvature controlled item is located is excluded as the target field of view position, and no further interception is performed on the adjacent field of view positions before and after.

[0122] S2.3 When the point column expansion result is that the expansion is valid, the field position of the field curvature controlled item is taken as the target field position. The expansion is valid when the expansion boundary of the controlled point column item moves outward towards the edge field of view compared to the expansion boundary of the center side point column item, and the diffusion range formed after the outward movement is greater than the diffusion range of the center side point column item. The diffusion range is determined based on the pixel area surrounded by the expansion boundary of the controlled point column item.

[0123] S2.4 Locate the position of the target's field of view in the field of view position sequence, and take the two field of view positions immediately before and after the position as the adjacent field of view positions; arrange the target's field of view position and the adjacent field of view positions according to the field of view position sequence to form a local field of view segment.

[0124] S2.5. Within a local field of view segment, the field point entries at adjacent field positions and the target field position are concatenated in the order of field expansion to form a point boundary chain. The outer boundaries of adjacent field positions and the target field position are compared within this point boundary chain. When the outer boundaries of adjacent field positions and the target field position continuously shift outwards towards the edge of the field of view, the direction of point dispersion expansion is formed. Specifically…

[0125] In the point-series boundary chain, the pixel position corresponding to the center point of the imaging test target surface is used as a common reference, and the outer boundaries of the adjacent field of view positions and the target field of view position are placed in the same image plane coordinates; along the radial line pointing from the center field of view direction to the edge field of view direction, the radial positions of the outer pixels of the outer boundaries of the adjacent field of view positions in the center field of view direction, the target field of view position, and the adjacent field of view positions in the edge field of view direction are read respectively.

[0126] When the radial position of the pixel line outside the outer boundary moves away from the center point of the imaging test target in the order of adjacent field positions in the central field of view, target field of view, and adjacent field positions in the edge field of view, and the number of pixels enclosed by the outer boundary increases in the same order, it is confirmed that the outer boundary continuously moves outward in the edge field of view. The direction from the adjacent field position in the central field of view to the adjacent field position in the edge field of view is recorded as the point column diffusion expansion direction. The point column diffusion expansion direction is saved in correspondence with the local field of view segment for subsequent merging of the high-frequency maintenance change direction into the same traceability entry.

[0127] If any of the following occurs: the radial position of the outer pixel line of the outer expansion boundary does not move away from the center point of the imaging test target in sequence, or the number of pixels surrounded by the outer expansion boundary does not increase in sequence, the direction of point column diffusion expansion will not be formed. The target field of view position will be marked as the position of point column expansion discontinuity, and the abnormal change jump point judgment will be excluded. It will no longer participate in the merging of the same traceability entry.

[0128] S2.6. Within a local field of view segment, the high-frequency terms of the field of view at adjacent field of view positions and the target field of view position are concatenated in the order of field of view expansion to form a high-frequency holding chain. In the high-frequency holding chain, the curve values ​​of adjacent field of view positions and the target field of view position are compared. When the curve value from the target field of view position to the edge of the field of view begins to decrease, the direction of change in the high-frequency holding amount is determined. Specifically,

[0129] In the high-frequency holding chain, the curve values ​​in the high-frequency term of the field of view are read in the order of adjacent field of view positions in the center field of view direction, target field of view position, and adjacent field of view positions in the edge field of view direction. When the curve value corresponding to the target field of view position is higher than the curve value corresponding to the adjacent field of view position in the center field of view direction, and the curve value corresponding to the target field of view position is higher than the curve value corresponding to the adjacent field of view position in the edge field of view direction, it is confirmed that a high-frequency holding peak is formed from the adjacent field of view position in the center field of view direction to the target field of view position, and it is confirmed that the curve value from the target field of view position to the adjacent field of view position in the edge field of view direction begins to decrease. The target field of view position is then pointed towards the adjacent field of view position in the edge field of view direction. The direction of the field of view position is recorded as the direction of high-frequency hold-up change, and the high-frequency hold-up change direction is saved in correspondence with the local field of view segment for subsequent merging with the field curvature controlled item and the point column diffusion expansion direction into the same traceability entry. When either the curve value corresponding to the target field of view position is not higher than the curve value corresponding to the adjacent field of view position in the center field of view direction, or the curve value corresponding to the target field of view position is not higher than the curve value corresponding to the adjacent field of view position in the edge field of view direction, the high-frequency hold-up change direction is not formed, the target field of view position is marked as a high-frequency descent discontinuity position, the abnormal change jump point judgment is excluded, and it is no longer involved in the merging of the same traceability entry.

[0130] S2.7. Merge the field curvature controlled item, the expansion direction of the point-series dispersion, and the change direction of the high-frequency holding quantity into the same traceability entry. When the same traceability entry contains both the field curvature controlled item and the point-series dispersion continues to expand along the expansion direction, and the high-frequency holding quantity turns to decrease along the change direction, determine the target field of view position as the abnormal change jump point. When the same traceability entry does not contain the field curvature controlled item, and any of the following occurs: the point-series dispersion does not continue to expand along the expansion direction, and the high-frequency holding quantity does not turn to decrease along the change direction, exclude the target field of view position from the abnormal change jump point judgment. Merge the abnormal change jump point, the local field of view segment, the expansion direction of the point-series dispersion, and the change direction of the high-frequency holding quantity to generate a flat failure record.

[0131] S3. Extract anomalous change jump points from the flat failure record, compare the changes in high-frequency retention before and after the anomalous change jump points, delineate the clear imaging damage locations where the point series diffusion increases and the high-frequency retention decreases, and determine the source of damage.

[0132] S3.1 Extract the anomalous change jump point, local field of view segment, direction of diffusion expansion of point array, and direction of change of high frequency retention from the flat failure record; divide the local field of view segment into pre-jump field of view segment and post-jump field of view segment using the anomalous change jump point as the boundary, and take the field of view position of the pre-jump field of view segment adjacent to the anomalous change jump point as the pre-jump position, take the anomalous change jump point as the jump point position, and take the field of view position of the post-jump field of view segment adjacent to the anomalous change jump point as the post-jump position.

[0133] S3.2. Compare the outer expansion boundaries of the point array diffusion at the pre-jump, jump point, and post-jump positions. When the outer expansion boundary of the point array diffusion increases sequentially from the pre-jump position to the jump point and then to the post-jump position, the field of view segment composed of the pre-jump, jump point, and post-jump positions is designated as the point array expansion duration segment. Specifically,

[0134] The diffusion margins of the pre-jump, jump point, and post-jump positions are placed in the same image plane coordinate system. The pixel position corresponding to the center point of the imaging test target is used as a common reference. The radial positions of the pixel lines outside the pre-jump, jump point, and post-jump margins are read along the central field of view towards the edge field of view. The number of pixels surrounded by the pre-jump, jump point, and post-jump margins are counted simultaneously.

[0135] The pixel lines outside the pre-jump, pre-jump, and post-jump expansion boundaries move away from the center point of the imaging test target surface in sequence, and the number of pixels enclosed by the pre-jump, pre-jump, and post-jump expansion boundaries increases sequentially. It is determined that the expansion boundary of the dot matrix diffusion increases sequentially from the pre-jump to the pre-jump to the pre-jump and post-jump positions. The field of view segment composed of the pre-jump, pre-jump, and post-jump positions is taken as the starting field of view segment of the dot matrix expansion duration segment. If either the radial position does not move away from the center point of the imaging test target surface in sequence or the number of pixels does not increase sequentially, the dot matrix expansion duration segment is not formed, and the field of view segment composed of the pre-jump, pre-jump, and post-jump positions is excluded from the determination result of the dot matrix expansion duration segment. It is no longer compared with the high-frequency descent duration segment and is not used to form the damaged overlap position.

[0136] It should be noted that the dot-matrix expansion duration is not fixed to three points: pre-jump, jump, and post-jump. Instead, it is based on the initial field of view segment and, with the jump point as the center, continues to read the dot-matrix diffusion boundary of adjacent field of view positions along the central and edge field of view directions. When the pixel line outside the outer boundary of the adjacent field of view position continues to move away from the center point of the imaging test target surface, and the number of pixels surrounded by the outer boundary continues to increase, the adjacent field of view position is incorporated into the dot-matrix expansion duration. When the pixel line outside the outer boundary of the adjacent field of view position does not continue to move away from the center point of the imaging test target surface, the adjacent field of view position is incorporated into the dot-matrix expansion duration. If the number of pixels enclosed by the outer boundary does not continue to increase, the extension of the dot column expansion continuous segment is stopped, and the start and end positions of the dot column expansion continuous segment are determined. The start and end positions of the high-frequency descent continuous segment are determined by the continuous decrease of the curve value in the high-frequency term of the field of view. The start and end positions of the dot column expansion continuous segment are determined by the dot column diffusion outer boundary and the number of pixels enclosed by the outer boundary. The judgment criteria for the two continuous segments are different, and their lengths and start and end positions are not necessarily the same. The common coverage comparison only retains the field of view positions that fall into both the dot column expansion continuous segment and the high-frequency descent continuous segment, which are used to form the damaged overlap position.

[0137] S3.3. Compare the high-frequency sustain curve values ​​at the pre-jump, jump point, and post-jump positions. When the high-frequency sustain curve value decreases sequentially from the pre-jump position to the jump point and then to the post-jump position, the field of view segment composed of the pre-jump, jump point, and post-jump positions is designated as the high-frequency descent duration segment. Specifically,

[0138] The high-frequency maintenance curve values ​​corresponding to the pre-jump, jump point, and post-jump positions are arranged in the order of field of view expansion to form the high-frequency maintenance curve values ​​for the pre-jump, jump point, and post-jump positions. When the high-frequency maintenance curve value for the pre-jump position is higher than that for the jump point, and the high-frequency maintenance curve value for the jump point is higher than that for the post-jump position, the high-frequency maintenance curve value is considered to decrease sequentially from the pre-jump position to the jump point and then to the post-jump position. The field of view segment composed of the pre-jump, jump point, and post-jump positions is classified as a high-frequency decreasing continuous segment. If any of the following occurs, the high-frequency maintenance curve value for the pre-jump position is not higher than that for the jump point, or the high-frequency maintenance curve value for the jump point is not higher than that for the post-jump position, it is not classified as a high-frequency decreasing continuous segment. The field of view segment composed of the pre-jump, jump point, and post-jump positions is excluded from the determination of the high-frequency decreasing continuous segment and is no longer compared with the point expansion continuous segment for joint coverage, nor is it used to form the damaged overlapping position.

[0139] S3.4 Project the dot-matrix expansion duration and the high-frequency descent duration onto the field-of-view sequence of the local field of view segment. Take the field-of-view position jointly covered by the dot-matrix expansion duration and the high-frequency descent duration as the damaged overlap position. Specifically,

[0140] The jump positions, jump points, and jump ends contained in the dot-series expansion duration are mapped one by one to the field-of-view position sequence of the local field of view segment, and the corresponding field-of-view positions are marked as dot-series expansion coverage positions; the jump positions, jump points, and jump ends contained in the high-frequency descent duration are mapped one by one to the field-of-view position sequence of the local field of view segment, and the corresponding field-of-view positions are marked as high-frequency descent coverage positions; when the same field-of-view position belongs to both the dot-series expansion coverage position and the high-frequency descent coverage position, the same field-of-view position is regarded as the damaged overlapping position; when there is no field-of-view position that belongs to both the dot-series expansion coverage position and the high-frequency descent coverage position, no damaged overlapping position is formed, and flat failure records are not included in the clear imaging damaged position delineation.

[0141] S3.5. Delineate the field of view positions covering the jump point in the overlapping damaged areas as the clear imaging damaged areas, and classify the field of view positions falling into the post-jump field of view segments in the overlapping damaged areas as the clear imaging damaged areas. Specifically,

[0142] The locations of overlapping damaged areas are mapped to jump points. When the overlapping damaged area includes a jump point, the field of view corresponding to the jump point is designated as the location of the damaged area with clear imaging. The locations of overlapping damaged areas are mapped to the field of view after the jump. When the overlapping damaged area falls into the field of view after the jump, the field of view falling into the field of view after the jump is still classified as the location of the damaged area with clear imaging. When the overlapping damaged area does not include a jump point and does not fall into the field of view after the jump, the location of the damaged area with clear imaging is not designated, and the flat failure record is not included in the determination of the source of damage.

[0143] It should be noted that the pre-jump position is used as a reference field of view before the anomalous change jump point. Its function is to confirm whether there is a change starting point between the pre-jump position and the jump point position where the point array diffusion increases and is accompanied by a decrease in the high frequency retention. The clear imaging damaged position is defined by the jump point position and the field of view position in the post-jump field of view segment that continuously meets the damaged overlap condition. Therefore, even if the pre-jump position is a damaged overlap position, it is only used as an adjacent reference before the change and is not included in the clear imaging damaged position. When it is necessary to continue to trace the damaged starting point in the direction of the central field of view, the pre-jump position is used again as a candidate jump point position to perform the anomalous change jump point judgment.

[0144] S3.6 Determining the source of damage involves placing the starting field-of-view position of the extended dot-matrix expansion duration and the starting field-of-view position of the extended high-frequency descent duration into a field-of-view position sequence. The source of damage is determined based on the order of these two starting field-of-view positions within the sequence. If the starting field-of-view position of the dot-matrix expansion duration is earlier than the starting field-of-view position of the high-frequency descent duration, the source of damage is determined to be an increase in dot-matrix dispersion. If the starting field-of-view position of the high-frequency descent duration is earlier than the starting field-of-view position of the dot-matrix expansion duration, the source of damage is determined to be a decrease in high-frequency retention. If the starting field-of-view positions of the dot-matrix expansion duration and the high-frequency descent duration are the same, the source of damage is determined to be the combined effect of an increase in dot-matrix dispersion and a decrease in high-frequency retention.

[0145] S4. Based on the continuous distribution of damaged locations and the source of damage in clear imaging, classify the optical quality level of the ultrafine endoscope and generate optical detection results.

[0146] S4.1 Place the damaged areas of clear imaging into the field of view position sequence, and check the adjacent relationships in the order of central field of view, transitional field of view, and edge field of view.

[0147] When the locations of damaged areas in a clear image are directly adjacent in the sequence of field of view positions, they are considered to be in a continuous distribution. When the locations of damaged areas in a clear image are separated by undamaged field of view positions, they are considered to be in a discontinuous distribution.

[0148] It should be noted that before classifying optical quality levels, a detection threshold verification is performed on the locations where clear imaging is impaired. This threshold verification includes verification of the percentage of point array diffusion area exceeding the limit, verification of the decrease in high-frequency retention, verification of the absolute value of field curvature shift, and verification of repeatability. The threshold for the percentage of point array diffusion area exceeding the limit is determined based on the allowable point array area within the same field of view and the point array area distribution of qualified samples from the same batch. The threshold for the decrease in high-frequency retention is determined based on the allowable high-frequency retention attenuation from the center field of view to the edge field of view. The absolute threshold for field curvature shift is determined based on the allowable image plane shift corresponding to the depth of field of the ultra-fine endoscope and the allowable image plane shift corresponding to the lens structural tolerance. The repeatability threshold is determined based on the results of multiple imaging tests at the same field of view. The fluctuation difference of point-series diffusion and the fluctuation difference of high-frequency retention are determined. When the location of the damaged clear image simultaneously meets the following conditions, such as the point-series diffusion area exceeding the limit ratio reaching the threshold, the high-frequency retention decrease reaching the high-frequency retention decrease threshold, the absolute value of field curvature shift not exceeding the absolute value of field curvature shift, and the repeatability consistency not exceeding the repeatability consistency threshold, it enters the optical quality level classification. When the location of the damaged clear image does not meet any of the following conditions, such as the point-series diffusion area exceeding the limit ratio verification, the high-frequency retention decrease verification, the absolute value of field curvature shift verification, and the repeatability consistency verification, it is not directly classified as a critical optical quality level or an unqualified optical quality level, but is marked as a field of view location to be retested.

[0149] When the location of the image impairment is not clearly delineated, the ultra-fine endoscope is classified into an acceptable optical quality level.

[0150] When there is only one field of view where clear imaging is impaired, and the source of the damage is either an increase in point array diffusion or a decrease in high frequency retention, the ultrafine endoscope is classified as a critical optical quality level.

[0151] When the location of the damaged area in clear imaging is only one field of view and the damage is caused by the combined effect of increased point-series diffusion and decreased high-frequency retention, the ultrafine endoscope is classified as having an unqualified optical quality level. When the locations of the damaged areas in clear imaging are continuously distributed in the field of view sequence and extend into the post-jump field of view segment, the ultrafine endoscope is classified as having an unqualified optical quality level.

[0152] When there are multiple consecutive distributions of damaged clear imaging locations, the segment with the longest clear imaging damage extending towards the edge of the field of view is taken as the grade determination segment. When the extension lengths are the same, the segment with clear imaging damage closest to the edge of the field of view is taken as the grade determination segment. When there is only one field of view location in the grade determination segment and the source of damage is either an increase in point array diffusion or a decrease in high frequency retention, the ultrafine endoscope is classified into a critical optical quality level.

[0153] It should be noted that the locations of damaged clear images originate from anomalous change jump points and post-jump field of view segments in the entire flat failure record, and not just from a single local field of view segment. When there are more than two anomalous change jump points in the field of view position sequence, the locations of damaged clear images are delineated according to the local field of view segments corresponding to the anomalous change jump points, and then all damaged clear images are placed in the same field of view position sequence for adjacency verification. When there are intervals between undamaged field of view positions between the damaged clear images corresponding to different anomalous change jump points, multiple continuous distributions are formed.

[0154] When the grading segment has only one field of view and the damage originates from the combined effect of increased point dispersion and decreased high-frequency retention, the ultrafine endoscope is classified as having an unqualified optical quality grade. When the grading segment is continuously distributed and extends into the post-jump field of view segment, the ultrafine endoscope is classified as having an unqualified optical quality grade.

[0155] The location of the damage, its continuous distribution, the source of the damage, and the optical quality level are recorded in the same record to generate optical detection results.

[0156] This embodiment also provides a computer device applicable to the ultra-fine endoscope optical quality detection method based on imaging performance data, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the ultra-fine endoscope optical quality detection method based on imaging performance data as proposed in the above embodiment.

[0157] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0158] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the method for detecting the optical quality of an ultrafine endoscope based on imaging performance data as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0159] In summary, this invention enhances the continuity and stability of identifying damage to clear imaging by jointly verifying the expansion direction of dot-matrix dispersion and the change direction of high-frequency retention. When the field curvature shift is still under control, the expansion of dot-matrix dispersion and the decrease in high-frequency retention can jointly reflect the ongoing loss of image detail. The flat failure record concentrates the flat field curvature state, the dot-matrix expansion state, and the high-frequency decrease state into the same field of view, making the abnormal change jump point have a clearer field of view positioning basis. The location of damage to clear imaging no longer depends on the instantaneous fluctuation of a single indicator, but is delineated based on the continuous change in the field of view expansion direction, which helps to reduce missed and false judgments, improves the verifiability of optical quality level classification, and makes the optical detection results of ultrafine endoscopes more accurately reflect the imaging quality of the entire field of view.

[0160] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for detecting the optical quality of an ultrafine endoscope based on imaging performance data, characterized in that: include, Full-field imaging performance tests were performed on the ultra-fine endoscope to obtain field curvature shift, point array diffusion, and high frequency retention. The field curvature shift, point array diffusion, and high frequency retention were then grouped into the same field position to generate a field record table. Lock the field position in the field record table where the field curvature offset is controlled and the point dispersion is expanding. Extract the adjacent field positions before and after, and trace the expansion direction of the point dispersion and the change direction of the high frequency holding amount in the adjacent field positions before and after. Locate the abnormal change jump point that simultaneously satisfies the conditions of controlled field curvature offset, continued expansion of point dispersion, and a decrease in high frequency holding amount, and generate a flat failure record. Extract anomalous change jump points from the flat failure record, compare the changes in high-frequency retention and point array dispersion before and after the anomalous change jump points, delineate the clear imaging damage location where point array dispersion increases and high-frequency retention decreases, and determine the source of damage. Based on the continuous distribution of damaged locations and the source of damage in clear imaging, the optical quality level of the ultrafine endoscope is classified, and optical detection results are generated.

2. The method for detecting the optical quality of an ultrafine endoscope based on imaging performance data as described in claim 1, characterized in that: The obtained field curvature offset, point dispersion, and high-frequency retention are specifically as follows: The field of view of the ultra-fine endoscope is divided into a central field of view, a transitional field of view, and a peripheral field of view along the direction of field of view development. The test images of the central field of view, the transition field of view, and the edge field of view are acquired and arranged in the order of field of view expansion of the central field of view, the transition field of view, and the edge field of view to form a full field of view test image. Defocus scanning was performed on the center field-of-view test image, transition field-of-view test image and edge field-of-view test image in the full field-of-view test imaging, and the sharp peak position was recorded at different image plane positions. The field curvature curve was plotted based on the sharp peak position. The sharp peak position corresponding to the test image of the center field of view is used as the field curvature reference position. The zero deviation distance of the center field of view relative to the field curvature reference position, as well as the deviation distance of the transition field of view and the edge field of view relative to the field curvature reference position, are extracted from the field curvature curve. The zero deviation distance and the deviation distance are combined into the field curvature offset. Multi-wavelength point image test beams are projected into the central field of view, the transition field of view, and the edge field of view to obtain the distribution of imaging points at different wavelengths; The outer boundaries of the imaging points at different wavelengths on the image plane are delineated, and the outer boundaries in the central field of view, the transition field of view, and the edge field of view are combined into a point series of diffusion quantities. In the central field-of-view test imaging, transition field-of-view test imaging, and edge field-of-view test imaging, the oblique transition zone is extracted, and gray-scale sampling lines are drawn within the oblique transition zone according to the sampling direction perpendicular to the oblique boundary. Collect the pixel grayscale values ​​of the dark areas, transition areas and bright areas that the grayscale sampling line passes through to form a grayscale transition sequence; The grayscale transition sequence is converted into a modulation transfer function curve, and the curve value at a high spatial frequency is selected from the modulation transfer function curve. The curve values ​​of the center field of view, transition field of view, and edge field of view are combined into a high-frequency retention value.

3. The method for detecting the optical quality of an ultrafine endoscope based on imaging performance data as described in claim 2, characterized in that: The generation of the field of view record table specifically involves... Arrange the central field of view, transitional field of view, and edge field of view in the order of field of view unfolding to form a field of view position sequence; Based on the field position sequence, the field curvature offset is divided into field curvature terms under the central field of view, transition field of view, and edge field of view; the point series diffusion is divided into point series terms under the central field of view, transition field of view, and edge field of view; and the high frequency retention is divided into high frequency terms under the central field of view, transition field of view, and edge field of view. The field curvature, field point, and high-frequency terms of the field of view at the same field of view position are grouped into the same field of view position to form the field of view imaging performance terms. Remove any field of view imaging performance items that are missing any of the field curvature, field of view point, or field of view high frequency items, and arrange the remaining field of view imaging performance items according to the field of view position sequence to generate a field of view record table.

4. The method for detecting the optical quality of an ultrafine endoscope based on imaging performance data as described in claim 3, characterized in that: Specifically, the extraction of adjacent field-of-view positions before and after the cropping is as follows: Traverse the field of view imaging performance items in the field of view record table, and retain the field of view imaging performance items that have both adjacent positions in the central field of view direction and adjacent positions in the edge field of view direction as the field of view imaging performance items to be locked. Extract the field curvature term to be locked from the imaging performance terms of the field of view to be locked, extract the center side field curvature term from the adjacent position in the center field of view direction, and extract the edge side field curvature term from the adjacent position in the edge field of view direction; The coverage area between the central side field curvature item and the edge side field curvature item is taken as the field curvature coverage area. The field curvature item to be locked that falls into the field curvature coverage area is marked as the field curvature controlled item. The field point column in the field curvature controlled term is taken as the controlled point column, and the field point column in the adjacent position in the central field direction is taken as the central side point column. The outer expansion boundary of the controlled point column and the central side point column is compared to form the point column expansion result. When the point column expansion result is that the expansion is valid, the field position of the field curve controlled term is taken as the target field position; Locate the position of the target's field of view in the field of view position sequence, and take the two field of view positions immediately before and after the current position as the adjacent field of view positions.

5. The method for detecting the optical quality of an ultrafine endoscope based on imaging performance data as described in claim 4, characterized in that: The term "outward expansion" means that the outward expansion boundary of the controlled point column moves outward toward the edge field of view compared to the outward expansion boundary of the center side point column, and the resulting diffusion range is greater than the diffusion range of the center side point column.

6. The method for detecting the optical quality of an ultrafine endoscope based on imaging performance data as described in claim 5, characterized in that: The generation of flat failure records specifically involves... Arrange the target's field of view position and the positions of its adjacent fields of view in a field of view position sequence to form a local field of view segment; In a local field of view segment, the field point list items under the adjacent field of view positions and the target field of view position are connected in the order of field of view expansion to form a point list boundary chain; In the point column boundary chain, the outer expansion boundary of the adjacent field of view position and the target field of view position is compared. When the outer expansion boundary of the adjacent field of view position and the target field of view position continuously moves outward towards the edge field of view, the point column diffusion expansion direction is formed. In a local field of view segment, the high-frequency terms of the field of view at adjacent field of view positions and at the target field of view position are connected in series according to the field of view unfolding order to form a high-frequency holding chain; In the high-frequency holding chain, the curve values ​​of adjacent field-of-view positions and the target field-of-view position are compared. When the curve value from the target field-of-view position to the edge field-of-view direction turns downward, the direction of change of the high-frequency holding amount is formed. The controlled terms of field curvature, the expansion direction of point-series dispersion, and the change direction of high-frequency holding quantity are grouped into the same traceability entry; When the same tracing entry contains field curvature controlled items, the point dispersion continues to expand along the point dispersion expansion direction, and the high frequency holding quantity turns to decrease along the high frequency holding quantity change direction, the target field position is determined as the anomalous change jump point. The abnormal change jump points, local field segments, the expansion direction of the point series dispersion, and the change direction of the high frequency retention are merged to generate a flat failure record.

7. The method for detecting the optical quality of an ultrafine endoscope based on imaging performance data as described in claim 6, characterized in that: The comparison of the changes in high-frequency retention and point-series dispersion before and after the anomalous change jump point is specifically as follows: Extract the anomalous change jump points, local field segments, the expansion direction of point-series dispersion, and the change direction of high-frequency retention from the flat failure record; Using the anomalous change jump point as the boundary, the local field of view is divided into a pre-jump field of view segment and a post-jump field of view segment. The field of view position of the pre-jump field of view segment adjacent to the anomalous change jump point is taken as the pre-jump position, and the anomalous change jump point is taken as the jump point position. The field of view position of the post-jump field of view segment adjacent to the anomalous change jump point is taken as the post-jump position. Compare the outer expansion boundary of the point array diffusion at the pre-jump position, the jump point position, and the post-jump position. When the outer expansion boundary of the point array diffusion increases sequentially from the pre-jump position to the jump point position and then to the post-jump position, the field of view segment composed of the pre-jump position, the jump point position, and the post-jump position is defined as the point array expansion duration segment. Compare the high-frequency sustain curve values ​​of the pre-jump position, the jump point position, and the post-jump position. When the high-frequency sustain curve value decreases sequentially from the pre-jump position to the jump point position and then to the post-jump position, the field of view segment composed of the pre-jump position, the jump point position, and the post-jump position is classified as the high-frequency descent sustained segment.

8. The method for detecting the optical quality of an ultrafine endoscope based on imaging performance data as described in claim 7, characterized in that: The locations where the clear imaging is impaired due to increased dispersion and decreased high-frequency retention are specifically defined as follows: Project the field position sequence of the dot-array expansion duration and the high-frequency descent duration onto the local field of view segment, and take the field position jointly covered by the dot-array expansion duration and the high-frequency descent duration as the damaged overlapping position; The field of view positions covering the jump point in the damaged overlapping positions are defined as the clear imaging damaged positions, and the field of view positions falling into the post-jump field of view in the damaged overlapping positions are classified as clear imaging damaged positions.

9. The method for detecting the optical quality of an ultrafine endoscope based on imaging performance data as described in claim 8, characterized in that: The method for determining the source of damage involves placing the starting field-of-view position of the extended point expansion segment and the starting field-of-view position of the extended high-frequency descent segment into the field-of-view position sequence, and determining the source of damage based on the sequential relationship between the two starting field-of-view positions in the field-of-view position sequence.

10. The method for detecting the optical quality of an ultrafine endoscope based on imaging performance data as described in claim 9, characterized in that: The optical quality grades include acceptable optical quality grades, critical optical quality grades, and unacceptable optical quality grades.

Images