High-speed optical module end face quality detection device and method

Abstract

The present invention discloses an end face quality detection device and method for a high-speed optical module, wherein the end face quality detection device for a high-speed optical module comprises a body, a transport component, an image acquisition component and a processor, wherein the transport component is rotatably arranged on the body, and the transport component is formed with a fixing position for fixing the optical module to be detected; the image acquisition component is rotatably arranged on the body, and the image acquisition component is used to acquire a multi-spectral image of the end face of the optical module to be detected; the processor is electrically connected to the transport component and the image acquisition component respectively, and the processor is used to adjust the relative position of the end face and the image acquisition component so that the end face is perpendicular to the imaging center line of the image acquisition component, and the processor can receive and process the multi-spectral image to identify the defect type and degree of the end face. The technical solution of the present invention can improve the accuracy, reliability and comprehensiveness of the optical fiber end face detection results.

Description

Technical Field

[0001] The present invention relates to the technical field of optical module end face defect detection, and in particular to an end face quality detection device and method for a high-speed optical module. Background Art

[0002] Optical modules are key components in fiber-optic communication systems. They convert electrical signals into optical signals or vice versa, thereby achieving high-speed, long-distance data transmission. The performance of optical modules directly affects the quality and reliability of the entire communication system, and the end face quality of optical modules is one of the important factors that determine their performance. Contaminants, scratches or other defects on the end face may cause signal attenuation, reflection or scattering, seriously affecting the transmission efficiency and service life of the optical module. Therefore, during the production and use of optical modules, it is crucial to conduct strict quality inspection of their end faces.

[0003] Existing optical module end face detection technology mainly relies on single spectrum imaging, usually using images in the visible light band for analysis. This method has several obvious disadvantages: First, some types of defects are not easily detected under the visible spectrum. For example, tiny surface cracks may be almost invisible under visible light, but will produce obvious scattering effects in the near-infrared band. Secondly, it is difficult to distinguish defects of different properties with a single spectrum. For example, surface contamination and material defects may present similar features in visible light images, leading to misjudgment. Furthermore, single spectrum detection is sensitive to changes in ambient light and is easily affected by external lighting conditions, reducing the stability and repeatability of detection. In addition, some optical coatings or treated end faces may appear uniform under visible light, masking potential quality problems that may be more easily found under other spectra. Finally, single spectrum detection is difficult to provide detailed information on the composition of the end face material, limiting the in-depth analysis of potential problems. In contrast, the high-speed optical module end face quality detection device of the present invention adopts multi-spectral imaging technology, combined with a rotatable transport component and an image acquisition component, and an intelligent processor, which can comprehensively and accurately identify various types of end face defects. This method can capture defect characteristics in different bands, distinguish problems of different natures, reduce the impact of environmental factors, detect hidden defects, and provide richer material information, thereby greatly improving the accuracy, reliability and comprehensiveness of detection. Summary of the invention

[0004] The main purpose of the present invention is to solve the technical problem that the optical fiber end face using a single spectrum detection optical module has low accuracy, reliability and comprehensiveness.

[0005] A first aspect of the present invention provides an end face quality detection device for a high-speed optical module, the end face quality detection device for a high-speed optical module comprising a body, a transport component, an image acquisition component and a processor;

[0006] The transport component is rotatably disposed on the body, and the transport component is formed with a fixing position for fixing the optical module to be detected;

[0007] The image acquisition component is rotatably disposed on the body, and the image acquisition component is used to acquire a multi-spectral image of the end face of the optical module to be detected;

[0008] The processor is electrically connected to the transport component and the image acquisition component respectively, and is used to adjust the relative position of the end face of the optical module to be detected and the image acquisition component so that the end face of the optical module to be detected is perpendicular to the imaging center line of the image acquisition component. The processor can receive and process the multispectral image to identify the type and degree of defects on the end face of the optical module to be detected.

[0009] Optionally, the transport assembly includes a rotating seat, a supporting member and a fixing member, the rotating seat is formed with a mounting surface, the supporting member is fixedly mounted on the mounting surface, the end of the supporting member away from the mounting surface is used to support the optical module to be detected, the fixing member is movably mounted on the mounting surface, and the end of the fixing member away from the mounting surface is used to be fixedly connected to the optical module to be detected; wherein,

[0010] The fixing member can move in a direction away from or close to the supporting member, so that the interval between the fixing member and the supporting member can be adjusted and adapted to optical modules to be detected of different lengths.

[0011] Optionally, the supporting member includes a frame fixed on the mounting surface and a bracket reciprocatably provided on the frame, an end of the bracket away from the mounting surface is formed as a supporting portion, and the bracket can reciprocate in a direction perpendicular to the mounting surface.

[0012] Optionally, a first bolt hole is formed on the fixing member, and a plurality of second bolt holes are formed on the mounting surface, wherein the plurality of second bolt holes are arranged at intervals in the direction of a line connecting the fixing member and the supporting member, and the end face quality detection device of the high-speed optical module includes bolts, and the bolts are respectively threadedly connected to the first bolt hole and any one of the second bolt holes to detachably fix the fixing member to the mounting surface.

[0013] Optionally, the transport assembly includes an X-axis slide, a Y-axis slide, a Z-axis slide and a rotating drive member, the X-axis slide can be movably disposed on the machine body along the X-axis direction, the Y-axis slide can be movably disposed on the X-axis slide along the Y-axis direction, the Z-axis slide can be movably disposed on the Y-axis slide along the Z-axis direction, the rotating drive member is disposed on the Z-axis slide, and the output shaft of the rotating drive member is transmission-connected to the rotating seat.

[0014] A second aspect of the present invention provides a method for detecting end face quality of a high-speed optical module, comprising:

[0015] Adjust the optical module to be detected until its end face is perpendicular to the imaging center line of the image acquisition component;

[0016] Performing multispectral imaging on the end face of the optical module to be detected to obtain a multispectral image set including a visible light band, a near infrared band, and an ultraviolet band;

[0017] Preprocessing the image of each band according to the multispectral image set;

[0018] A multi-scale feature extraction network is used to extract features from the preprocessed multispectral image set to obtain a multi-scale feature map;

[0019] Performing feature fusion on the multi-scale feature map to obtain a fused feature map;

[0020] Analyzing the fused feature map using a defect detection network to identify the location, type, and severity of the defect;

[0021] Based on the identified defect information, the end face intactness rate is calculated and a defect density map is generated.

[0022] Optionally, the method of extracting features from the preprocessed multispectral image set using a multiscale feature extraction network to obtain a multiscale feature map includes:

[0023] Constructing a multi-scale feature extraction network including an encoder and a decoder, wherein the encoder includes a plurality of downsampling stages, and the decoder includes a plurality of upsampling stages corresponding to the downsampling stages;

[0024] Introducing a residual global perception module in each stage of the encoder and decoder, the residual global perception module divides the input feature map into multiple groups and applies space-specific convolution and channel-specific convolution within and between groups respectively;

[0025] Adding a skip connection between corresponding stages of the encoder and decoder to connect the feature map of the encoder stage with the feature map of the corresponding stage of the decoder;

[0026] The preprocessed multispectral image set is processed by the multi-scale feature extraction network, and a feature map is output from each stage of the decoder to obtain a multi-scale feature map.

[0027] Optionally, performing feature fusion on the multi-scale feature map to obtain a fused feature map includes:

[0028] Performing channel compression and excitation processing on the multi-scale feature map to obtain a channel weight map;

[0029] Generate a two-dimensional spatial attention map according to the multi-scale feature map;

[0030] Multiplying the channel weight map by the two-dimensional spatial attention map element by element to obtain an attention weighted feature map;

[0031] Performing feature fusion on the attention weighted feature map and the multi-scale feature map to obtain an intermediate feature map;

[0032] A feature pyramid network is used to perform multi-scale feature reconstruction on the intermediate feature map to obtain a fused feature map.

[0033] Optionally, the using a defect detection network to analyze the fused feature map to identify the location, type and severity of the defect includes:

[0034] Applying a region proposal network to the fused feature map to obtain a candidate defect region, wherein the region proposal network uses a circular anchor box instead of a rectangular anchor box;

[0035] Extracting regional features from the fused feature map according to the candidate defect region to obtain a regional feature map;

[0036] The regional feature map is input into a dual-branch structure network, where one branch performs defect classification and the other branch performs bounding box regression to obtain a defect type probability distribution and refined bounding box coordinates;

[0037] According to the defect type probability distribution and the refined bounding box coordinates, a non-maximum suppression algorithm is used for post-processing to obtain a final defect detection result;

[0038] Based on the final defect detection result, each detected defect is quantitatively scored according to a preset evaluation standard to obtain the location, type and severity information of the defect.

[0039] Optionally, the calculating the end face integrity rate and generating a defect density map according to the identified defect information includes:

[0040] According to the location, type and severity of the defect, the end face is divided into regions to obtain defective regions and non-defective regions;

[0041] Calculating the ratio of the area of ​​the non-defective region to the total area of ​​the end face to obtain the end face intact rate;

[0042] According to the spatial distribution and severity of the defective area, a two-dimensional thermal map is generated to obtain a defect density map.

[0043] The high-speed optical module end face quality detection device of the present invention effectively solves many defects of the existing single spectrum detection method through the innovative combination of multi-spectral imaging technology, rotatable transportation and image acquisition components and intelligent processors. Multi-spectral imaging covers visible light, near-infrared and ultraviolet bands, and can fully capture various defect characteristics. For example, tiny cracks that are difficult to detect under visible light can be clearly detected in the near-infrared band, and some pollutants are more obvious under ultraviolet light, thereby significantly improving the accuracy and comprehensiveness of detection. The rotatable transportation component allows precise adjustment of the angle of the optical module to be detected, making it suitable for different types of optical modules such as straight end faces and oblique end faces, ensuring that the end face is perpendicular to the imaging center line to obtain the best imaging effect. At the same time, the rotatable image acquisition component further enhances the flexibility of the system, can collect images from multiple angles, is suitable for complex shapes or specially treated end faces, and reduces information omissions. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the structures shown in these drawings without paying creative work.

[0045] Figure 1 It is a structural schematic diagram of an embodiment of an end face quality detection device for a high-speed optical module of the present invention;

[0046] Figure 2 for Figure 1 A partial enlarged view of the middle A;

[0047] Figure 3 for Figure 1 Schematic diagram of the structure after omitting some structures;

[0048] Figure 4 for Figure 3 A partial enlarged view of point B in the middle;

[0049] Figure 5 for Figure 3 A schematic diagram of the structure from another perspective;

[0050] Figure 6 for Figure 5 A partial enlarged view of point C in the middle;

[0051] Figure 7 The figure is a flow chart of the end face quality detection method of the high-speed optical module of the present invention.

[0052] Description of Figure Numbers:

[0053] 1. Machine body; 2. Transport assembly; 21. Rotating seat; 211. Mounting surface; 212. Second bolt hole; 22. Lifting member; 221. Frame; 222. Bracket; 222a. Supporting portion; 23. Fixing member; 231. First bolt hole; 24. X-axis slide; 25. Y-axis slide; 26. Z-axis slide; 27. Rotating drive member; 3. Image acquisition assembly; 100. Optical module to be detected.

[0054] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0055] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0056] It should be noted that if the embodiments of the present invention involve directional indications (such as up, down, left, right, front, back...), the directional indications are only used to explain the relative position relationship, movement status, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication will also change accordingly.

[0057] In addition, the descriptions of "first", "second", etc. in the present invention are only used for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, "and / or" in the full text includes three solutions. Taking A and / or B as an example, it includes technical solution A, technical solution B, and technical solution that satisfies both A and B. In addition, the technical solutions between the various embodiments can be combined with each other, which must be based on the ability of ordinary technicians in the field to implement. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0058] An embodiment of the present application provides a device for detecting the end surface quality of a high-speed optical module. Figures 1 to 6 The end face quality detection device of the high-speed optical module includes a body 1, a transport component 2, an image acquisition component 3 and a processor.

[0059] The transport component 2 is rotatably disposed on the body 1, and the transport component 2 is formed with a fixing position for fixing the optical module 100 to be detected;

[0060] The image acquisition component 3 is rotatably disposed on the body 1, and the image acquisition component 3 is used to acquire a multi-spectral image of the end face of the optical module 100 to be detected;

[0061] The processor is electrically connected to the transport component 2 and the image acquisition component 3, respectively. The processor is used to adjust the relative position of the end face of the optical module 100 to be detected and the image acquisition component 3 so that the end face of the optical module 100 to be detected is perpendicular to the imaging center line of the image acquisition component 3. The processor can receive and process the multispectral image to identify the type and degree of defects on the end face of the optical module 100 to be detected.

[0062] Specifically, in the present embodiment, the body 1 can be a frame structure made of metal material or engineering plastics for supporting and fixing other components. The fixing position can be formed by an adjustable clamp or manipulator for stabilizing the optical module 100 to be detected. The image acquisition component 3 can include a high-resolution camera, a spectrometer or a multi-band filter set for capturing multi-spectral images. The processor can be an industrial-grade computer or an embedded system for data processing and device control.

[0063] In some embodiments, the transport component 2 and the image acquisition component 3 can be rotatable through bearings or slide rails. The processor can establish an electrical connection with the transport component 2 and the image acquisition component 3 through a data line or a wireless module to achieve data transmission and the sending of control instructions.

[0064] "Imaging centerline vertical" means that the optical axis of the image acquisition component 3 forms a 90-degree angle with the end face of the optical module 100 to be detected. This configuration ensures the clarity and accuracy of the image. The processor "adjusts the relative position" means that it can control the movement of the transport component 2 and the image acquisition component 3 to achieve the best imaging position.

[0065] It can be understood that the high-speed optical module end face quality detection device of the present invention effectively solves many defects of the existing single spectrum detection method through the innovative combination of multi-spectral imaging technology, rotatable transportation and image acquisition components 3 and intelligent processors. Multi-spectral imaging covers visible light, near-infrared and ultraviolet bands, and can fully capture various defect characteristics. For example, tiny cracks that are difficult to detect under visible light can be clearly detected in the near-infrared band, and some pollutants are more obvious under ultraviolet light, thereby significantly improving the accuracy and comprehensiveness of the detection. The rotatable transportation component 2 allows the angle of the optical module 100 to be detected to be accurately adjusted, making it suitable for different types of optical modules such as straight end faces and oblique end faces, ensuring that the end face is perpendicular to the imaging center line to obtain the best imaging effect. At the same time, the rotatable image acquisition component 3 further enhances the flexibility of the system, can collect images from multiple angles, is suitable for complex shapes or specially treated end faces, and reduces information omissions.

[0066] Optionally, the transport assembly 2 includes a rotating seat 21, a supporting member 22 and a fixing member 23, the rotating seat 21 is formed with a mounting surface 211, the supporting member 22 is fixedly mounted on the mounting surface 211, the end of the supporting member 22 away from the mounting surface 211 is used to support the optical module 100 to be detected, the fixing member 23 is movably mounted on the mounting surface 211, and the end of the fixing member 23 away from the mounting surface 211 is used to be fixedly connected to the optical module 100 to be detected; wherein,

[0067] The fixing member 23 can move in a direction away from or close to the supporting member 22 , so that the interval between the fixing member 23 and the supporting member 22 can be adjusted and adapted to different lengths of the optical modules 100 to be detected.

[0068] In this embodiment, the rotating seat 21 can be a disk, a polygonal platform or other irregularly shaped platform made of metal or high-strength engineering plastics, which is used to support and rotate the optical module 100 to be detected. The mounting surface 211 is a flat surface on the rotating seat 21 for mounting other components.

[0069] The supporting member 22 can be fixed on the mounting surface 211 by bolts, welding or embedded installation. The fixing member 23 can be connected to the mounting surface 211 by a slide groove, a guide rail, a bolt or a gear rack mechanism to achieve a movable feature. The fixing member 23 and the optical module 100 to be detected can be fixedly connected by bolts, mechanical buckles or vacuum adsorption to ensure that the optical module is stable and immobile during the detection process.

[0070] “The fixing member 23 can move in a direction away from or close to the supporting member 22” means that the fixing member 23 can reciprocate in a straight line to accommodate optical modules of different lengths. This design allows the interval between the fixing member 23 and the supporting member 22 to be adjusted as needed. This adjustable design greatly improves the applicability and flexibility of the device. It can adapt to various optical modules of different lengths on the market and can be quickly adjusted without replacing parts, thereby improving detection efficiency. At the same time, this design also ensures that optical modules of different sizes can be firmly and reliably supported and fixed, thereby ensuring the accuracy and stability of the position of the optical module during the detection process, and improving the accuracy and reliability of the detection results.

[0071] Optionally, the supporting member 22 includes a frame 221 fixed on the mounting surface 211 and a bracket 222 reciprocatably arranged on the frame 221, and the end of the bracket 222 away from the mounting surface 211 is formed as a supporting portion 222a, and the bracket 222 can reciprocate in a direction perpendicular to the mounting surface 211.

[0072] The frame 221 can be fixed on the mounting surface 211 by bolt connection, welding or integral molding. The bracket 222 can be connected to the frame 221 by a slide rail, a guide column or a screw nut mechanism to achieve reciprocating motion in the vertical direction (perpendicular to the mounting surface 211).

[0073] “The bracket 222 can reciprocate in a direction perpendicular to the mounting surface 211” means that the bracket 222 can move up and down in a plane perpendicular to the mounting surface 211. This design of the height-adjustable support member 22 significantly enhances the adaptability and accuracy of the device. It can accurately adjust the height position of the optical module to ensure that the end face of the optical module and the image acquisition component 3 maintain the best relative position, thereby obtaining the clearest image. At the same time, this design also enables the device to adapt to optical modules of different sizes and shapes, improving the versatility and efficiency of the detection process. The vertical adjustment function also helps to reduce stress during the installation of the optical module, reduces the risk of damage to the optical module, and further ensures the reliability of the detection and the integrity of the optical module.

[0074] Optionally, a first bolt hole 231 is formed on the fixing member 23, and a plurality of second bolt holes 212 are formed on the mounting surface 211, and the plurality of second bolt holes 212 are arranged at intervals in the direction of a line connecting the fixing member 23 and the supporting member 22, and the end face quality detection device of the high-speed optical module includes bolts, and the bolts are respectively threadedly connected to the first bolt hole 231 and any one of the second bolt holes 212 to detachably fix the fixing member 23 to the mounting surface 211.

[0075] The first bolt hole 231 may be a circular or elliptical through hole penetrating the fixing member 23 for accommodating a bolt. The second bolt hole 212 may be a threaded hole or a smooth through hole provided on the mounting surface 211 for use with a nut. The bolt may be a common hexagonal bolt, a hexagon socket bolt, or a quick bolt with a hand-tightening head for detachably connecting the fixing member 23 and the mounting surface 211 together.

[0076] It is not difficult to understand that this adjustable fixing method greatly improves the flexibility and adaptability of the device. By selecting different second bolt holes 212, the operator can quickly and accurately adjust the position of the fixing member 23 to adapt to optical modules of different lengths. This design not only simplifies the adjustment process and reduces the adjustment time, but also ensures the accuracy and repeatability of the adjustment. At the same time, the detachable design facilitates maintenance and replacement, and improves the maintainability and service life of the device.

[0077] Optionally, the transport assembly 2 includes an X-axis slide 24, a Y-axis slide 25, a Z-axis slide 26 and a rotating drive member 27, the X-axis slide 24 can be movably disposed on the machine body 1 along the X-axis direction, the Y-axis slide 25 can be movably disposed on the X-axis slide 24 along the Y-axis direction, the Z-axis slide 26 can be movably disposed on the Y-axis slide 25 along the Z-axis direction, the rotating drive member 27 is disposed on the Z-axis slide 26, and the output shaft of the rotating drive member 27 is transmission-connected to the rotating seat 21.

[0078] Specifically, the X-axis slide 24 can be connected to the body 1 by means of a linear guide or a slider to achieve movement in the X direction. The Y-axis slide 25 can be installed on the X-axis slide 24 in a similar manner, and the Z-axis slide 26 is installed on the Y-axis slide 25 to form a three-dimensional motion system. The rotary drive member 27 can be fixed to the Z-axis slide 26 by a flange or a bracket. The output shaft of the rotary drive member 27 can be connected to the rotary seat 21 by means of a coupling, gear drive or belt drive to achieve power transmission.

[0079] This three-axis slide plus rotary drive design enhances the flexibility and accuracy of the transport component 2. It can accurately locate the optical module 100 to be detected in three-dimensional space, and can adjust the angle of the optical module to adapt to various types of optical module end faces, including straight end faces and oblique end faces. This full range of adjustment capabilities ensures that the optical module end face and the image acquisition component 3 always maintain the best relative position, thereby obtaining the clearest and most accurate image. At the same time, this design also improves the automation and efficiency of the device, reduces manual intervention, and improves the consistency and repeatability of detection.

[0080] Another embodiment of the present application further provides a method for detecting the end surface quality of a high-speed optical module, which is applied to the device for detecting the end surface quality of a high-speed optical module provided in the above embodiment. Figure 7 It is a flow chart of a method for detecting end face quality of a high-speed optical module.

[0081] See also Figure 7 , the end face quality inspection methods of high-speed optical modules include:

[0082] Adjust the optical module to be detected until its end face is perpendicular to the imaging center line of the image acquisition component;

[0083] Specifically, first, the transport component uses its three-axis slide system to roughly move the optical module to be inspected to a preset position. The X-axis slide controls the forward and backward movement of the optical module relative to the image acquisition component, the Y-axis slide controls the left and right movement, and the Z-axis slide adjusts the height. Subsequently, the rotary drive component comes into play, adjusting the orientation of the end face of the optical module by precisely controlling the angle of the rotating seat. For example, for an optical module with an 8-degree inclined end face, the rotary drive component needs to drive the rotating seat to rotate precisely 8 degrees so that the end face is perpendicular to the imaging center line.

[0084] During this process, the image acquisition component continuously captures real-time images of the end face of the optical module, and the processor determines whether the end face has reached a vertical state by analyzing these images. The processor can use edge detection algorithms to analyze the symmetry of the end face profile, or use laser ranging technology to accurately measure the distance from each point on the end face to the image acquisition component. If deviations are found, the processor will calculate the angle and position that need to be adjusted in real time, and send instructions to the transport component for fine-tuning.

[0085] The entire adjustment process is automated, greatly reducing human error and improving the efficiency and accuracy of detection. This precise adjustment capability enables the device to adapt to various types of optical modules, including straight end faces, oblique end faces, and even optical modules with complex curved surfaces. By ensuring that the end face is strictly perpendicular to the imaging centerline, the device lays a solid foundation for subsequent multi-spectral imaging and defect detection, greatly improving the reliability and consistency of the detection results.

[0086] Please continue reading Figure 7 , performing multispectral imaging on the end face of the optical module to be detected to obtain a multispectral image set including a visible light band, a near infrared band and an ultraviolet band;

[0087] Specifically, first, the image acquisition component is equipped with a high-performance multispectral camera, which has the ability to image in three bands: visible light, near-infrared, and ultraviolet. During the imaging process, the processor controls the image acquisition component to switch filters or light sources of different bands in sequence to obtain images in various spectral ranges. For example, for a standard FC / PC fiber optic connector, imaging is first performed in the visible light band of 400-700nm, which helps capture the overall morphology of the end face and obvious surface defects. Subsequently, the system switches to the near-infrared band of 750-1400nm, which is particularly effective for detecting tiny cracks and internal stresses, because some defects will produce obvious scattering or reflection in this band. Finally, the system performs imaging in the ultraviolet band of 200-400nm, which is very helpful for detecting surface contaminants and the integrity of certain special coatings.

[0088] The entire multispectral imaging process is precisely controlled by the processor to ensure that images in each band are acquired under optimal exposure and focus conditions. The processor also adjusts the parameters of the image acquisition components, such as exposure time, gain, and contrast, in real time to adapt to the optical characteristics of different bands. For example, when imaging in the UV band, since most materials have lower reflectivity in this band, the system may automatically increase the exposure time or increase the sensor gain. In addition, in order to ensure accurate alignment between images in different bands, the system performs rapid image registration after each band switch, which is usually achieved by identifying feature points on the end face.

[0089] The advantage of this multispectral imaging method is that it can fully capture the various features and potential defects of the end face of the optical module. Visible light images provide intuitive surface information, near-infrared images can reveal deep structural problems, and ultraviolet images are particularly suitable for detecting surface contamination and the integrity of special coatings. By comprehensively analyzing the images of these three bands, defects that are difficult to detect with traditional single spectrum detection can be detected, greatly improving the comprehensiveness and accuracy of detection. In addition, multispectral images also provide a rich data foundation for subsequent deep learning analysis, allowing the system to more accurately identify and classify various types of defects, thereby providing a more reliable basis for the quality control of optical modules.

[0090] Please continue reading Figure 7 , preprocessing the image of each band according to the multispectral image set;

[0091] Specifically, the preprocessing process first performs dark field correction, which eliminates the influence of sensor inherent noise and thermal noise by subtracting the dark field image acquired under no light conditions. For example, dark field correction is particularly important for images in the near-infrared band because thermal noise is more obvious in this band. Then flat field correction is performed, using a flat field image acquired with a uniform light source to compensate for the non-uniformity of the optical system and pixel response differences. This step is critical to ensure the consistency of image brightness, especially when detecting defects in the edge area of ​​the fiber end face.

[0092] The system then uses an adaptive median filter for denoising, effectively removing impulse noise from the image while retaining edge and detail information. This method is particularly suitable for processing fiber end-face images that may contain small scratches or stains. For example, for an end-face image of an FC / PC connector, an adaptive median filter can effectively remove random noise caused by dust without blurring the fine structure of the end face. Next, the system applies the non-local mean (NLM) algorithm for further denoising, which achieves noise suppression by looking for similar pixel blocks in different areas of the image, and is particularly suitable for retaining the texture details of the image.

[0093] To enhance the contrast and detail visibility of the image, the system applies the Contrast Limited Adaptive Histogram Equalization (CLAHE) algorithm to each band of the image. This method can effectively improve the local contrast of the image while avoiding excessive enhancement of noise. The CLAHE algorithm is particularly effective for images in the UV band, because images in this band usually have low contrast. Through CLAHE processing, small contaminants or coating defects on the end face will become more obvious, facilitating subsequent defect detection.

[0094] In the final stage of preprocessing, the system performs precise registration of multispectral images. Using the visible light image as a reference, the near-infrared and ultraviolet images are spatially aligned. This step uses feature point-based registration methods such as the SIFT (Scale Invariant Feature Transform) algorithm to extract feature points, and then the RANSAC (Random Sample Consensus) algorithm is used to estimate the transformation matrix. Finally, the image is resampled through bicubic interpolation to ensure that the images of different bands correspond accurately in space. This precise registration is crucial for subsequent multispectral analysis because it allows the system to compare information from different bands at the pixel level.

[0095] The advantage of this series of preprocessing steps is that the image quality and data reliability are significantly improved. By eliminating various noises and systematic errors, enhancing image contrast, and ensuring accurate alignment of images in different bands, a solid foundation is laid for subsequent feature extraction and defect detection. This comprehensive and sophisticated preprocessing not only improves the accuracy of detection, but also enhances the system's sensitivity to tiny defects. For example, a tiny crack that is almost invisible in the original image may become clearly discernible after these preprocessing steps. In addition, this standardized preprocessing process also ensures consistency between images acquired from different batches and at different times, which is particularly important for long-term quality monitoring and trend analysis. In general, this preprocessing process greatly improves the reliability, accuracy, and sensitivity of the entire detection system, providing strong support for high-quality optical module end face detection.

[0096] Please continue reading Figure 7 , a multi-scale feature extraction network is used to extract features from the preprocessed multispectral image set to obtain a multi-scale feature map;

[0097] In one embodiment of the present invention, the method of extracting features from a preprocessed multispectral image set using a multiscale feature extraction network to obtain a multiscale feature map includes: constructing a multiscale feature extraction network including an encoder and a decoder, the encoder including multiple downsampling stages, and the decoder including multiple upsampling stages corresponding to the downsampling stages; introducing a residual global perception module in each stage of the encoder and the decoder, the residual global perception module divides the input feature map into multiple groups, and applies space-specific convolution and channel-specific convolution within and between groups respectively; adding a jump connection between corresponding stages of the encoder and the decoder, connecting the feature map of the encoder stage with the feature map of the corresponding stage of the decoder; processing the preprocessed multispectral image set through the multiscale feature extraction network, outputting a feature map from each stage of the decoder, and obtaining a multiscale feature map.

[0098] Specifically, the encoder contains multiple downsampling stages, each of which consists of two 3x3 convolutional layers and a maximum pooling layer. For example, for a 640×480 pixel input image, the first downsampling stage will halve its size to 320×240 while increasing the number of feature channels. This design of gradually reducing the spatial resolution enables the network to capture features of different scales from details to the global. Correspondingly, the decoder contains multiple upsampling stages, each of which uses transposed convolution to enlarge the spatial size of the feature map and restore it to the original resolution.

[0099] In each stage of the encoder and decoder, the present invention introduces an innovative residual global perception module (RGPM). RGPM first divides the input feature map into multiple groups, for example, dividing the feature map of 64 channels into 4 groups, each with 16 channels. Within each group, RGPM applies space-specific convolutions, which share parameters in the spatial dimension but are independent between channels. The benefit of this is that it can effectively capture spatial correlations while maintaining independence between channels. Next, RGPM applies channel-specific convolutions between groups, which share parameters in the channel dimension but are independent between spatial positions. This design allows the network to learn the relationship between channels while maintaining the independence of spatial information. Through this alternating convolution within and between groups, RGPM is able to capture local and global contextual information at the same time, greatly enhancing the network's feature extraction capabilities.

[0100] In order to better preserve and utilize information at different scales, the present invention adds jump connections between corresponding stages of the encoder and decoder. These connections connect the feature maps of a certain stage of the encoder directly to the feature maps of the corresponding stage of the decoder. For example, the feature map of the second stage of the encoder will be connected to the feature map of the penultimate stage of the decoder. This design effectively solves the gradient vanishing problem in deep networks, while also allowing high-resolution detail information to be directly passed to the later stages of the decoder. This is particularly important for fiber end face inspection because it ensures that detail information such as tiny scratches or cracks will not be lost in the deep layers of the network.

[0101] When processing a preprocessed set of multispectral images, the multiscale feature extraction network first takes images of different bands as input channels. For example, visible light, near-infrared, and ultraviolet images can form an input tensor with multiple channels. The encoder part of the network gradually extracts and compresses features, while the decoder part reconstructs and refines these features. It is worth noting that each stage of the decoder outputs a feature map that represents image features at different scales. For example, for a network with 5 downsampling / upsampling stages, 5 feature maps of different scales will be obtained in the end, corresponding to 1 / 32, 1 / 16, 1 / 8, 1 / 4, and 1 / 2 sizes of the original image.

[0102] The advantage of this multi-scale feature extraction method is that it can comprehensively capture various features of the fiber end face. The larger-scale feature map contains the global information of the end face, such as the overall shape and large-area defects, while the smaller-scale feature map retains fine local details, such as tiny scratches or stains. By combining information at different scales, the network can more accurately identify and locate defects of various types and sizes. In addition, the introduction of the residual global perception module greatly enhances the network's ability to capture complex features, enabling the network to better understand the relationship between images in different bands, thereby extracting richer and more meaningful features.

[0103] In general, this carefully designed multi-scale feature extraction network provides a strong foundation for subsequent defect detection and classification tasks. It can not only process image information in different bands, but also extract and integrate features at multiple scales, greatly improving the system's ability to detect various types of defects. The flexibility and powerful feature expression capabilities of this method enable the present invention to perform well in handling various complex fiber end face detection scenarios, providing a reliable quality control method for high-quality optical module production.

[0104] Please continue reading Figure 7 , performing feature fusion on the multi-scale feature map to obtain a fused feature map;

[0105] In one embodiment of the present invention, the feature fusion of the multi-scale feature map to obtain a fused feature map includes: performing channel compression and excitation processing on the multi-scale feature map to obtain a channel weight map; generating a two-dimensional spatial attention map based on the multi-scale feature map; element-by-element multiplication of the channel weight map with the two-dimensional spatial attention map to obtain an attention weighted feature map; feature fusion of the attention weighted feature map and the multi-scale feature map to obtain an intermediate feature map; and performing multi-scale feature reconstruction of the intermediate feature map using a feature pyramid network to obtain a fused feature map.

[0106] Specifically, first, the multi-scale feature maps are subjected to channel compression and excitation processing, and this step uses an improved Squeeze-and-Excitation module. Specifically, the feature maps of each scale are first subjected to global average pooling to compress the spatial dimensions into a vector, which represents the global response of each channel. Subsequently, this vector is processed through two fully connected layers and nonlinear activation functions (such as ReLU) to obtain a weight vector with the same number of input channels. This weight vector is the channel weight map, which reflects the importance of different channels. For example, for a feature map containing 64 channels, channel compression and excitation processing will generate a 64-dimensional weight vector, in which each element corresponds to the importance score of a channel.

[0107] At the same time, a two-dimensional spatial attention map is generated based on the multi-scale feature map. This step uses a spatial attention module, which first performs maximum pooling and average pooling on the feature map in the channel dimension to obtain two two-dimensional feature maps. The two feature maps are concatenated together and then pass through a convolutional layer and a sigmoid activation function to generate a two-dimensional spatial attention map. This attention map reflects the importance of different spatial positions in the feature map. This step is particularly meaningful for fiber end face inspection because it can highlight areas on the end face where defects may exist, such as the edge or center area of ​​the end face.

[0108] Next, the channel weight map is element-wise multiplied with the two-dimensional spatial attention map to obtain the attention-weighted feature map. This step actually combines channel attention and spatial attention to generate a comprehensive attention mechanism. For example, if a channel scores high in the channel weight map and a spatial position also scores high in the spatial attention map, the feature of this position on the channel will be greatly enhanced. This combination allows the network to focus on important feature channels and important spatial positions at the same time, greatly improving the expressiveness of the features.

[0109] Then, the attention-weighted feature map is fused with the original multi-scale feature map to obtain an intermediate feature map. This step is usually implemented through element-level addition or concatenation operations. This fusion not only retains the comprehensive information of the original feature map, but also highlights the important features selected by the attention mechanism, thereby maintaining the integrity of the features while improving the recognition of the features.

[0110] Finally, the feature pyramid network (FPN) is used to reconstruct the multi-scale features of the intermediate feature map to obtain the final fused feature map. The core idea of ​​FPN is to build a top-down feature pyramid and use lateral connections to fuse features of different resolutions. Specifically, FPN starts with the feature map of the highest layer (lowest resolution) and gradually builds high-resolution feature maps through upsampling and 1×1 convolution. At each layer, the upsampled feature map is fused with the intermediate feature map of the corresponding resolution. This process continues until the original resolution, and finally a series of feature maps with different resolutions but rich semantic information are obtained, which is the fused feature map.

[0111] This complex feature fusion and reconstruction process has many advantages. First, channel compression and excitation processing can adaptively adjust the importance of different channels, which is particularly useful for processing multispectral images because information in different bands may have different importance. Second, the spatial attention mechanism can help the network focus on key areas where defects may exist and improve the accuracy of detection. Combining these two attention mechanisms further enhances the network's expressiveness and pertinence. The use of feature pyramid networks ensures that the detection system can obtain semantically rich feature representations at different scales, which is crucial for detecting defects of different sizes.

[0112] In general, this feature fusion and reconstruction process greatly enhances the network's ability to detect fiber end face defects. It can not only effectively integrate multi-scale and multi-spectral information, but also intelligently focus on important features and areas while maintaining sensitivity to defects of different scales. The flexibility and powerful feature expression capabilities of this method enable the present invention to perform well in handling various complex fiber end face detection scenarios, providing a reliable and accurate quality control method for high-quality optical module production.

[0113] Please continue reading Figure 7 , analyzing the fused feature map using a defect detection network to identify the location, type and severity of the defect;

[0114] In one embodiment of the present invention, the defect detection network is used to analyze the fused feature map to identify the location, type and severity of the defect, including: applying a region proposal network to the fused feature map to obtain a candidate defect region, wherein the region proposal network uses a circular anchor box instead of a rectangular anchor box; performing regional feature extraction on the fused feature map according to the candidate defect region to obtain a regional feature map; inputting the regional feature map into a dual-branch structure network, wherein one branch performs defect classification and the other branch performs bounding box regression to obtain a defect type probability distribution and refined bounding box coordinates; post-processing using a non-maximum suppression algorithm based on the defect type probability distribution and the refined bounding box coordinates to obtain a final defect detection result; based on the final defect detection result, each detected defect is quantitatively scored according to a preset evaluation standard to obtain the location, type and severity information of the defect.

[0115] Specifically, first, a region proposal network is applied to the fused feature map. The special feature of this network is that a circular anchor frame is used instead of a traditional rectangular anchor frame. This design is suitable for the circular characteristics of the fiber end face, because the fiber end face is usually circular, and the use of a circular anchor frame can more accurately locate the potential defect area. For example, for a standard single-mode fiber end face with a diameter of 125 microns, the region proposal network may generate a series of circular anchor frames of different sizes, ranging from 5 microns to 50 microns, to cover a variety of possible defect sizes. The use of this circular anchor frame greatly improves the accuracy and efficiency of the candidate defect area and reduces unnecessary background information interference.

[0116] Next, based on the obtained candidate defect regions, regional features are extracted from the fused feature map to obtain a regional feature map. This step usually uses the RoIAlign (Region of Interest Align) technology, which can accurately map irregularly shaped candidate regions to fixed-size feature maps to maintain the continuity of spatial information. For example, each candidate defect region may be mapped to a 7×7 feature map, regardless of the size of the original candidate region. This method ensures the consistency of subsequent analysis while retaining detail information.

[0117] Subsequently, the extracted regional feature map is input into an innovative dual-branch structure network. One branch of this network is responsible for defect classification, and the other branch performs bounding box regression. The classification branch passes through a series of convolutional layers and fully connected layers, and finally outputs the probability distribution of each possible defect type. For example, for the fiber end face, the defect types include scratches, cracks, contamination, bubbles, etc. At the same time, the regression branch is responsible for refining the coordinates of the bounding box, and it outputs the center coordinates of the bounding box, the radius (for a circular bounding box) or the width and height (for a rectangular bounding box) adjustment values. This dual-branch structure allows the network to simultaneously locate and identify defects, improving the efficiency and accuracy of detection.

[0118] In order to handle possible overlapping detection results, the present invention uses a non-maximum suppression (NMS) algorithm for post-processing. The NMS algorithm compares the detection confidence of the overlapping area, retains the detection result with the highest confidence, and suppresses other overlapping low-confidence detections. This step effectively eliminates redundant detection results and improves the accuracy of the final detection results. For example, if multiple detection frames overlap to mark the same scratch, NMS will retain the detection frame with the highest confidence and delete other overlapping detection frames.

[0119] Finally, based on the final defect detection results, the present invention quantitatively scores each detected defect according to preset evaluation criteria. This evaluation process takes into account multiple factors, including the size, location, shape, and performance of the defect under different spectra. For example, for a detected scratch, the system will comprehensively consider factors such as its length, depth (inferred by multi-spectral information), and distance from the fiber core. Each factor has a weight, and ultimately a comprehensive severity score is obtained. This quantitative scoring method not only provides information on the location and type of defects, but also gives the severity of the defects, providing an important basis for subsequent quality control decisions.

[0120] The advantages of this entire defect detection and evaluation process are its high accuracy, efficiency and adaptability. The use of circular anchor frames is particularly suitable for the geometric characteristics of the fiber end face, which improves the accuracy of initial detection. The dual-branch network structure allows for simultaneous defect location and classification, which improves processing efficiency. Non-maximum suppression ensures the accuracy and redundancy of the detection results. Finally, the quantitative scoring system provides an objective and comprehensive defect evaluation, which helps to develop precise quality control standards. This method can not only accurately detect various types of defects, but also evaluate their severity, providing a powerful and flexible tool for quality control in the production process of optical modules, which helps to improve product quality and production efficiency.

[0121] Please continue reading Figure 7 ,According to the identified defect information, the end face intactness rate is calculated and the defect density map is generated.

[0122] In one embodiment of the present invention, the end face intactness rate is calculated and a defect density map is generated based on the identified defect information, including: dividing the end face into regions according to the location, type and severity information of the defects to obtain defective regions and non-defective regions; calculating the ratio of the area of ​​the non-defective regions to the total area of ​​the end face to obtain the end face intactness rate; and generating a two-dimensional thermal map based on the spatial distribution and severity of the defective regions to obtain a defect density map.

[0123] Specifically, first, the end face is accurately divided into defect areas and non-defect areas based on this information. This process uses the previously obtained defect boundary box information, combined with the defect type and severity, to accurately outline the scope of each defect. For example, for a standard single-mode optical fiber end face with a diameter of 125 microns, if a scratch with a length of 10 microns is detected, the system will accurately mark the defect area on the end face image based on the location and direction of the scratch.

[0124] Subsequently, the present invention calculates the ratio of the area of ​​the non-defective region to the total area of ​​the end face to obtain the end face integrity rate. This ratio intuitively reflects the overall quality of the end face. For example, if the total area of ​​an optical fiber end face is 12271 square microns (corresponding to a circle with a diameter of 125 microns), and the total area of ​​all defective regions detected is 100 square microns, then the end face integrity rate will be 99.19%. This indicator provides a simple and clear reference standard for quality control.

[0125] Next, the present invention generates a two-dimensional heat map based on the spatial distribution and severity of the defect area, which is the so-called defect density map. In this heat map, different colors represent different defect densities or severity. For example, red may represent a highly concentrated or severe defect area, while blue represents slight or scattered defects. This visualization method not only intuitively shows the distribution of defects, but also reflects the severity of defects. For the optical fiber end face, this heat map can clearly show whether the defects are concentrated in the core area of ​​the optical fiber, which has an important impact on the transmission performance of the optical fiber.

[0126] The advantage of this comprehensive evaluation method is that it provides a multi-dimensional quality assessment. The end face integrity rate provides a quantitative indicator of the overall quality, which is convenient for quickly judging and comparing the quality of different fiber end faces. The defect density map provides more detailed and intuitive quality distribution information, which helps to identify potential problem areas and locations where defects are concentrated. The combination of these two methods can not only help quality control personnel quickly evaluate the overall quality of the fiber end face, but also guide them to focus on specific problem areas, so as to formulate more targeted improvement measures. In addition, this method also provides a powerful tool for quality trend analysis in mass production, which can help identify systematic problems in the production process, and then continuously optimize the production process and improve product quality.

[0127] The above description is only a preferred embodiment of the present invention, and does not limit the patent scope of the present invention. All equivalent structural changes made by using the contents of the present invention specification and drawings under the inventive concept of the present invention, or directly / indirectly applied in other related technical fields are included in the patent protection scope of the present invention.

Claims

1. A method for detecting the end face quality of a high-speed optical module, characterized in that: The end face quality detection method of the high-speed optical module is applied to the end face quality detection device of the high-speed optical module, and the end face quality detection device of the high-speed optical module comprises: Body (1); A transport component (2) is rotatably disposed on the body (1), and the transport component (2) is formed with a fixing position for fixing the optical module (100) to be detected; An image acquisition component (3) is rotatably disposed on the body (1), and the image acquisition component (3) is used to acquire a multi-spectral image of the end face of the optical module (100) to be detected; a processor, electrically connected to the transport component (2) and the image acquisition component (3), respectively, the processor being used to adjust the relative position of the end face of the optical module (100) to be detected and the image acquisition component (3) so that the end face of the optical module (100) to be detected is perpendicular to the imaging center line of the image acquisition component (3), and the processor is capable of receiving and processing the multispectral image to identify the type and degree of defects on the end face of the optical module (100) to be detected; The end face quality detection method of the high-speed optical module comprises: Adjust the optical module to be detected until its end face is perpendicular to the imaging center line of the image acquisition component; Performing multispectral imaging on the end face of the optical module to be detected to obtain a multispectral image set including a visible light band, a near infrared band, and an ultraviolet band; Preprocessing the image of each band according to the multispectral image set; A multi-scale feature extraction network is used to extract features from a pre-processed multi-spectral image set to obtain a multi-scale feature map; the multi-scale feature extraction network is used to extract features from a pre-processed multi-spectral image set to obtain a multi-scale feature map, including: constructing a multi-scale feature extraction network including an encoder and a decoder, the encoder including multiple downsampling stages, and the decoder including multiple upsampling stages corresponding to the downsampling stages; introducing a residual global perception module in each stage of the encoder and the decoder, the residual global perception module divides the input feature map into multiple groups, and applies space-specific convolution and channel-specific convolution within and between groups respectively; adding a jump connection between the corresponding stages of the encoder and the decoder, connecting the feature map of the encoder stage with the feature map of the corresponding stage of the decoder; processing the pre-processed multi-spectral image set through the multi-scale feature extraction network, outputting a feature map from each stage of the decoder, and obtaining a multi-scale feature map; Performing feature fusion on the multi-scale feature map to obtain a fused feature map; performing feature fusion on the multi-scale feature map to obtain a fused feature map, including: performing channel compression and excitation processing on the multi-scale feature map to obtain a channel weight map; generating a two-dimensional spatial attention map according to the multi-scale feature map; performing element-by-element multiplication of the channel weight map and the two-dimensional spatial attention map to obtain an attention weighted feature map; performing feature fusion on the attention weighted feature map and the multi-scale feature map to obtain an intermediate feature map; performing multi-scale feature reconstruction on the intermediate feature map using a feature pyramid network to obtain a fused feature map; The fused feature map is analyzed by using a defect detection network to identify the location, type and severity of the defect; the analysis of the fused feature map by using a defect detection network to identify the location, type and severity of the defect includes: applying a region proposal network to the fused feature map to obtain a candidate defect region, wherein the region proposal network uses a circular anchor frame instead of a rectangular anchor frame; performing regional feature extraction on the fused feature map according to the candidate defect region to obtain a regional feature map; inputting the regional feature map into a dual-branch structure network, wherein one branch performs defect classification and the other branch performs bounding box regression to obtain a defect type probability distribution and refined bounding box coordinates; post-processing is performed using a non-maximum suppression algorithm based on the defect type probability distribution and the refined bounding box coordinates to obtain a final defect detection result; based on the final defect detection result, each detected defect is quantitatively scored according to a preset evaluation standard to obtain information on the location, type and severity of the defect; Based on the identified defect information, the end face intactness rate is calculated and a defect density map is generated.

2. The end face quality detection method of a high-speed optical module according to claim 1, characterized in that: The transport component (2) comprises a rotating seat (21), a supporting member (22) and a fixing member (23); the rotating seat (21) is formed with a mounting surface (211); the supporting member (22) is fixedly mounted on the mounting surface (211); an end of the supporting member (22) away from the mounting surface (211) is used to support the optical module (100) to be detected; the fixing member (23) is movably mounted on the mounting surface (211); an end of the fixing member (23) away from the mounting surface (211) is used to be fixedly connected to the optical module (100) to be detected; wherein: The fixing member (23) can move in a direction away from or close to the supporting member (22), so that the interval between the fixing member (23) and the supporting member (22) can be adjusted and adapted to optical modules (100) to be detected of different lengths.

3. The end face quality detection method of a high-speed optical module according to claim 2, characterized in that: The supporting member (22) comprises a frame body (221) fixed on the mounting surface (211) and a bracket (222) reciprocatably arranged on the frame body (221), wherein the end of the bracket (222) away from the mounting surface (211) is formed as a supporting portion (222a), and the bracket (222) is reciprocatably movable in a direction perpendicular to the mounting surface (211).

4. The end face quality detection method of a high-speed optical module according to claim 2, characterized in that: A first bolt hole (231) is formed on the fixing member (23), and a plurality of second bolt holes (212) are formed on the mounting surface (211). The plurality of second bolt holes (212) are arranged at intervals in the direction of a line connecting the fixing member (23) and the supporting member (22). The end face quality detection device of the high-speed optical module comprises bolts, and the bolts are respectively threadedly connected to the first bolt hole (231) and any one of the second bolt holes (212) so as to detachably fix the fixing member (23) on the mounting surface (211).

5. The end face quality detection method of a high-speed optical module according to claim 2, characterized in that: The transport assembly (2) comprises an X-axis slide (24), a Y-axis slide (25), a Z-axis slide (26) and a rotating drive member (27); the X-axis slide (24) is movably arranged on the machine body (1) along the X-axis direction; the Y-axis slide (25) is movably arranged on the X-axis slide (24) along the Y-axis direction; the Z-axis slide (26) is movably arranged on the Y-axis slide (25) along the Z-axis direction; the rotating drive member (27) is arranged on the Z-axis slide (26); and the output shaft of the rotating drive member (27) is drivingly connected to the rotating seat (21).

6. The end face quality detection method of a high-speed optical module according to claim 1, characterized in that: The method of calculating the end face integrity rate and generating a defect density map according to the identified defect information includes: According to the location, type and severity of the defect, the end face is divided into regions to obtain defective regions and non-defective regions; Calculating the ratio of the area of ​​the non-defective region to the total area of ​​the end face to obtain the end face intact rate; According to the spatial distribution and severity of the defective area, a two-dimensional thermal map is generated to obtain a defect density map.

Images