A method for automatically identifying an electronic endoscope lens module

Abstract

The application discloses a kind of electronic endoscope lens module automatic identification method;Including the following steps: S1, first electronic endoscope is treated;S2, image is collected by electronic endoscope;S3, then data image information is transmitted processing;S4, image information is calculated again;S5, finally image information is post-processing;The present application is realized three hundred and sixty degrees panoramic collection data by three groups of lens module, then by processing module, the picture of three groups of electronic endoscope collection is successively branched collection discrimination, and the image signal of current effective lens module is removed interference, threshold processing and image signal increment operation etc., image processing method is realized, the collection image of detection point is discriminated and selected, and image information is processed, then by post-processing, detection information is processed, doctor is not needed Manual rotation adjustment, reduce the discomfort of patient.

Description

Technical Field

[0001] This invention belongs to the technical field of medical devices, specifically relating to a method for automatic identification of electronic endoscope lens modules. Background Technology

[0002] An electronic endoscope is a medical electro-optical instrument that integrates advanced optical, mechanical, and electrical technologies. It can be inserted into the cavities of the human body and organs for direct observation, diagnosis, and treatment. Electronic endoscopes are currently a convenient means of disease detection, providing medical personnel with strong diagnostic evidence for accurately assessing patients' conditions. However, various electronic endoscopes on the market still have various problems to meet different testing needs.

[0003] For example, the electronic endoscope system disclosed in patent announcement number CN109475281B, while generating a spectroscopic image based on the image signals of at least two systems generated by the image signal generation unit, does not address the issue that existing electronic endoscope main units sometimes need to be equipped with different lens modules. Currently, some systems use manual settings when using different lens modules, which is relatively inconvenient. Others use a one-to-one connection between the main unit and the lens module, resulting in poor compatibility, flexibility, and high cost. Therefore, it is essential to design an electronic endoscope main unit that is compatible with multiple lens modules and can automatically identify different lens modules when they are used. To this end, we propose a method for automatic identification of electronic endoscope lens modules. Summary of the Invention

[0004] The purpose of this invention is to provide a method for automatic identification of electronic endoscope lens modules. The method is easy to implement and simple to use. It realizes the identification and selection of images acquired at the detection point, and processes the image information and detection information. The host will automatically identify the current lens module and process and allocate it in the subsequent circuit.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A method for automatic identification of electronic endoscope lens modules includes the following steps:

[0007] S1. First, disinfect the electronic endoscope: disinfect the tubes of the electronic endoscope and keep the electronic endoscope sterile. Apply gel to the tubes of the electronic endoscope before inserting it.

[0008] S2. Image acquisition via electronic endoscope: When the tube of the electronic endoscope is inserted into the human body, the first lens module, the second lens module and the third lens module at the end of the tube of the electronic endoscope are used to capture detection point information.

[0009] S3. Then, the data image information is transmitted and processed: the first lens module, the second lens module and the third lens module transmit the image information to the central processing unit through a compatible interface to complete the transmission of image information;

[0010] S4. Further processing of image information: The central processing unit of the electronic endoscope is compatible with multiple lens modules through a compatible interface. When different lens modules need to be replaced according to requirements, the central processing unit automatically identifies the lens module currently in use by performing image processing on the images of different lens modules, performing successive multi-channel acquisition and identification, and performing interference removal, threshold processing, and image signal increment calculation on the image signal of the currently effective lens module.

[0011] S5. Finally, perform post-processing on the image information: process and distribute the current lens module image signal through post-processing circuits.

[0012] The compatibility interface D is connected to the first lens module A, the second lens module B, the third lens module C, and the central processing unit E, respectively. The central processing unit E is connected to the image processing unit F, the compatibility interface D, and the post-processing unit G, respectively. Key components include the first lens module A, the second lens module B, the third lens module C, the compatibility interface D, the image acquisition CPU chip U1, and the image logic allocation MOS devices Q1 and Q2. Key components include... Figure 3The connection method, with TEST1, TEST2, and TEST3 representing the integration of lens modules A, B, and C and the compatible interface D, allows images to be captured by the three lens modules A, B, and C in the handle. These images are then transmitted to the host computer via the compatible interface D. Simultaneously, the image acquisition CPU chip U1 captures and identifies which image input from lens modules A, B, or C it represents. If it's a valid image from lens module A, a logic code 01 is allocated through image logic allocation MOS devices Q1 and Q2 and transmitted to the central processing unit (CPU). The CPU then activates the corresponding processing module one for lens module A. Similarly, if the image acquisition CPU chip U1 identifies a valid image from lens module B, a logic code 10 is allocated through image logic allocation MOS devices Q1 and Q2 and transmitted to the CPU. This activates the corresponding processing module two for lens module B. Similarly, if the image acquisition CPU chip U1 recognizes that the input from the handle is a valid image from lens module C, it allocates a logic code 11 through image logic allocation MOS devices Q1 and Q2, and transmits this logic code 11 to the central processing unit. The central processing unit then activates the processing module three corresponding to lens module C. The technical challenge addressed here is that when different lens modules need to be changed for a testing scenario, manual settings are often required, which is inflexible, inconvenient, and requires training for medical staff. Incorrect manual settings can also prevent normal operation, affecting the efficiency of medical staff. In some cases, there is even a one-to-one correspondence between a lens module and a host computer, requiring a different host computer to be used when changing a lens module. This is even more inconvenient, and configuring a host computer for each lens module is costly. Therefore, the contribution of this technology is that when lens modules need to be changed according to the testing scenario, automatic recognition and transmission are performed, making operation convenient, focusing more on the operational efforts of medical staff, and also significantly saving hospital costs. Furthermore, based on the automatic recognition technology, it can be expanded as needed. For example, when three lens modules A, B, and C are connected at the same time, the image acquisition CPU chip U1 automatically detects the three sets of valid image signals, automatically allocates logic codes and transmits them to the central processing unit. The central processing unit then simultaneously opens its three internal processing modules and processes and stitches the images to achieve 360-degree panoramic detection images.

[0013] Preferably, the disinfection treatment of the electronic endoscope in S1 involves immediately washing it with clean water after use to remove blood and mucus residues. During the cleaning process, the endoscope surface is cleaned with degreased cotton balls or professional lens cleaning paper. After the initial cleaning, the endoscope is soaked in a multi-enzyme cleaning solution. After soaking in the multi-enzyme cleaning solution, the endoscope should be thoroughly rinsed with clean water and dried.

[0014] Preferably, the gel in step S1 refers to a thick liquid or semi-solid preparation with gel properties made from the raw drug and excipients that can form gels. The gel applied before electronic endoscopy usually serves to lubricate, relieve pain and anesthetize. The gel includes oxybuprocaine hydrochloride gel or tetracaine hydrochloride gel.

[0015] Preferably, in step S2, the first lens module, the second lens module, and the third lens module are arranged at equal angles, and all three are tilted outwards.

[0016] Preferably, the first lens module, the second lens module, and the third lens module each include a fill light and a charge-coupled device (CCD). The CCD is used to take pictures of the detection points, and the fill light is used to provide a light source for the CCD.

[0017] Preferably, the charge-coupled device (CCD) has photoelectric conversion function, as well as signal charge storage, transfer, and readout functions, and its working process consists of four steps:

[0018] The first step is the light integration period, which is the exposure time. During this period, the pixels of the charge-coupled device (CCD) convert the incident light quanta into photogenerated charges in a proportional manner, thus completing the photo-to-electric conversion.

[0019] The second is to temporarily store the photocharge generated by each pixel in the corresponding photodiode potential well during the light integration process, thereby achieving signal charge storage.

[0020] Third, after exposure, the stored photogenerated charge is transferred to the output area along the CCD shift register to complete the charge transfer;

[0021] Fourth, in the readout amplifier, each photogenerated charge is sequentially converted into a corresponding video signal to complete the signal readout. Therefore, the charge-coupled device (CCD) can be regarded as a photoelectric converter, which transforms a spatially distributed optical image into a video voltage signal distributed in time sequence.

[0022] Preferably, the image signal de-interference in step S4 includes a model-based method or a learning-based method;

[0023] The learning-based methods include DnCNN, FFDnet, and CBDnet. DnCNN mainly targets Gaussian noise for denoising, emphasizing the role of residual learning and batch normalization (BN). FFDnet considers generalizing Gaussian noise into more complex real noise, using the noise level map as part of the network input. CBDnet mainly focuses on the noise level map part of FFDnet, adaptively obtaining the noise level map through 5 layers of FCN to achieve blind denoising.

[0024] Preferably, the threshold processing in step S4 aims to extract target objects in the image and distinguish between background and noise. Typically, a threshold T is set to divide the pixels of the image into two categories: a group of pixels larger than T and a group of pixels smaller than T.

[0025] The most important aspect of the threshold processing is binarization.

[0026] The calculation formula for binarization is as follows:

[0027]

[0028] Where: Y is the gray value of the pixel group in the image binarization process, T is the threshold, gray is the gray value of the input image pixel, all pixels with gray value less than the threshold T have their gray value set to 0, representing black; all pixels with gray value greater than or equal to the threshold T have their gray value set to 255, representing white.

[0029] Preferably, the image signal increment in S4 is supervised data augmentation, that is, using preset data transformation rules to amplify the data on the basis of existing data, including single-sample data augmentation and multi-sample data augmentation, wherein single samples include geometric operations and color transformations.

[0030] The geometric transformation class refers to performing geometric transformations on images, including various operations such as flipping, rotating, cropping, deforming, and scaling.

[0031] The color transformation includes noise, blur, color transformation, erasure, and fill.

[0032] Preferably, the post-processing in S5 refers to the post-processing performed on the final rendered image after the normal rendering pipeline has ended;

[0033] The post-processing includes anti-aliasing, FXAA, TAA, flood glare, lens flare, color grading, white balance, saturation, contrast, gamma correction, color gain, color shift, tone mapping, depth of field, ambient occlusion, and motion blur processing.

[0034] Through the above technical measures, the most critical steps are steps S3 and S4. These steps mainly solve the problem of a host machine being compatible with the input of multiple lens modules and being able to automatically recognize them.

[0035] The main technical benefit of this step is that multiple lens modules can be freely selected without manual settings, and the recognition and transmission are automatic.

[0036] The advancement of this invention over the prior art lies in the fact that it eliminates the need for a one-to-one replacement of the main unit when changing the lens module, or the elimination of corresponding manual settings when changing the module, making it convenient and efficient.

[0037] The main difference between the technical solution of this invention and the prior art is that the prior art is configured one-to-one, with one lens module equipped with one host, or manual settings are required when a host is replaced with another lens module. The innovative technical solution, however, allows one host to be compatible with multiple lens modules, freely selecting the required lens module according to the scenario, without requiring manual settings, and automatically recognizing and transmitting the information.

[0038] Experiment Implementation: Upon power-up, with the first, second, and third lens modules disconnected, the values ​​of TEST1, TEST2, and TEST3 were acquired as follows: 0.03, 0.035, and 0.041, respectively. The logic codes CS1 and CS2 were 00, and the LEDs were red. The display showed no signal. Then, the first lens module (OV9734) was connected to the compatibility interface, while the second and third lens modules remained disconnected. The values ​​of TEST1, TEST2, and TEST3 were again acquired: 0.28, 0.034, and 0.039, respectively. The logic codes CS1 and CS2 were 01, and the display clearly showed the image from the OV9734 lens module. The LEDs were green. Next, the second lens module (OV6946) was connected to the compatibility interface, while the first and third lens modules remained disconnected. The values ​​of TEST1, TEST2, and TEST3 were again acquired: 0.029, 0.035, and 0.041, respectively. The values ​​of TEST1, TEST2, and TEST3 are 0.029, 0.033, and 0.35 respectively, with CS1 and CS2 logic codes set to 10. The display clearly shows the image of the OV6946 lens module, and the LED lights up blue. Then, the third lens module, OCHTA10, is connected to the compatible interface, while the first and second lens modules are not connected. The values ​​of TEST1, TEST2, and TEST3 are again collected, and they are 0.029, 0.033, and 0.35 respectively. The CS1 and CS2 logic codes are 11, and the display clearly shows the image of the OCHTA10 lens module, with the LED lights up yellow. Then, the first lens module, OV9734, is connected to the compatible interface again, while the second and third lens modules are not connected. The values ​​of TEST1, TEST2, and TEST3 are again collected, and they are 0.29, 0.035, and 0.038 respectively. The CS1 and CS2 logic codes are 01, and the display again clearly shows the image of the OV9734 lens module.

[0039] The above experiments showed that the design was correct, automatically identifying various lens modules and transmitting the corresponding images correctly.

[0040] Compared with the prior art, the beneficial effects of the present invention are:

[0041] In use, this invention integrates three lens modules within the tubing of an electronic endoscope. These three modules connect to a central processing unit via a compatible interface, enabling synchronous data transmission. The three lens modules are equidistantly distributed, facilitating 360-degree panoramic data acquisition when the electronic endoscope is inserted into the body. A processing module then sequentially acquires and identifies the images from the three lens modules, performing image processing methods such as interference removal, thresholding, and image signal increment calculation on the currently valid lens module. This allows for the identification and selection of images for the detection point and the processing of image information. Subsequent processing steps then complete the processing of the detection information. This eliminates the need for manual adjustment by the doctor, reducing patient discomfort. Furthermore, the electronic endoscope host is compatible with multiple lens modules and automatically identifies when different modules are used. The method of use is simple: users can directly switch lens modules according to the detection scenario requirements. The host will automatically identify the current lens module and process and allocate it in the subsequent circuitry. Attached Figure Description

[0042] Figure 1 A schematic diagram of the system structure for an automatic identification method of an electronic endoscope lens module;

[0043] The components include: a first lens module (OV9734), a second lens module (OV6946), and a third lens module (OCHTA10); a compatible interface is a SCSI connector with 1.27 20 pins; the image acquisition CPU is NY8B062E; and the image logic allocation and processing is AO3400.

[0044] The first lens module A (OV9734), the second lens module B (OV6946), and the third lens module C (H01A10) each have their own independent input interface in the compatible interface (SCSI) connector 1.27 (20PIN).

[0045] Figure 2 This is a flowchart illustrating a method for automatic identification of electronic endoscope lens modules.

[0046] Figure 3 This is a schematic diagram illustrating part of the principle of an automatic identification method for an electronic endoscope lens module. Detailed Implementation

[0047] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0048] Please see Figures 1-2 It is known that a method for automatic identification of an electronic endoscope lens module includes the following steps:

[0049] S1. First, disinfect the electronic endoscope: disinfect the tubes of the electronic endoscope and keep the electronic endoscope sterile. Apply gel to the tubes of the electronic endoscope before inserting it.

[0050] S2. Image acquisition via electronic endoscope: When the tube of the electronic endoscope is inserted into the human body, the detection point information is captured by the first lens module A, the second lens module B and the third lens module C at the end of the tube of the electronic endoscope.

[0051] S3. Then, the data image information is transmitted and processed: the first lens module A, the second lens module B, and the third lens module C transmit the image information to the central processing unit E through the compatible interface D, thus completing the transmission of image information.

[0052] S4. Further processing of image information: The central processing unit E of the electronic endoscope is compatible with the first lens module A, the second lens module B and the third lens module C through the compatible interface D. When different lens modules need to be replaced according to requirements, the central processing unit E automatically identifies the lens module currently in use by performing image processing F on the images of different lens modules, performing successive multi-channel acquisition and identification, and performing interference removal, threshold processing and image signal increment calculation on the image signal of the currently effective lens module.

[0053] S5. Finally, perform post-processing on the image information: G: Perform post-processing and allocation on the current lens module image signal.

[0054] The compatible interface D (SCSI connector 1.2720PIN) is connected to the first lens module A (OV9734), the second lens module B (OV6946), the third lens module C (OCHTA10), and the central processing unit E (integrated motherboard BBT_CPU01Y). The central processing unit E is connected to the image processing unit F (CPUNY8B062E), the compatible interface D, and the post-processing unit G.

[0055] Technical problems solved and effects achieved: Depending on the testing scenario, changing lens modules does not require a one-to-one replacement of the main unit, or manual settings are not required when changing modules, making it convenient and efficient. It enables the free selection, automatic identification, and transmission of various lens modules.

[0056] The image logic allocation processing is A (O3400). The first lens module A (OV9734), the second lens module B (OV6946), and the third lens module C (H01A10) each occupy an independent input interface in the compatible interface (SCSI) connector 1.27 (20PIN).

[0057] To ensure the safety of equipment use and improve aseptic operation, in this embodiment, preferably, the electronic endoscope disinfection process in S1 involves immediately washing it with clean water after use to remove blood and mucus residues. During cleaning, the endoscope surface is cleaned with degreased cotton balls or professional lens cleaning paper. After the initial cleaning, it is soaked in a multi-enzyme cleaning solution. After soaking in the multi-enzyme cleaning solution, the endoscope should be thoroughly rinsed with clean water and dried.

[0058] In order to ensure that the electronic endoscope tube does not cause discomfort or damage to human organs and tissues when inserted into the human body, in this embodiment, preferably, the gel in S1 refers to a thick liquid or semi-solid preparation with gel properties made of raw drug and excipients that can form gel. The gel applied before electronic endoscopy usually serves to lubricate, relieve pain and anesthetize. The gel includes oxybuprocaine hydrochloride gel or tetracaine hydrochloride gel.

[0059] In order to achieve panoramic acquisition and detection of the human body and facilitate switching of lenses without the need for doctors to manually rotate and adjust, in this embodiment, preferably, the first lens module A, the second lens module and the third lens module in S2 are arranged at equal angles, and the first lens module, the second lens module B and the third lens module C are all tilted outwards.

[0060] In order to achieve clear images during shooting and to make the electronic endoscope tube small enough for easy insertion into the human body, in this embodiment, preferably, the first lens module, the second lens module B and the third lens module C all include a supplementary light and a charge-coupled device (CCD). The CCD is used to take pictures of the detection points, and the supplementary light is used to provide a light source for the CCD.

[0061] To achieve a compact shooting tool while improving image clarity, a charge-coupled device (CCD) is used as the image acquisition device. In this embodiment, preferably, the CCD has photoelectric conversion function, as well as signal charge storage, transfer, and readout functions. Its working process consists of four steps:

[0062] The first step is the light integration period, which is the exposure time. During this period, the pixels of the charge-coupled device (CCD) convert the incident light quanta into photogenerated charges in a proportional manner, thus completing the photo-to-electric conversion.

[0063] The second is to temporarily store the photocharge generated by each pixel in the corresponding photodiode potential well during the light integration process, thereby achieving signal charge storage.

[0064] Third, after exposure, the stored photogenerated charge is transferred to the output area along the CCD shift register to complete the charge transfer;

[0065] Fourth, in the readout amplifier, each photogenerated charge is sequentially converted into a corresponding video signal to complete the signal readout. Therefore, the charge-coupled device (CCD) can be regarded as a photoelectric converter, which transforms a spatially distributed optical image into a video voltage signal distributed in time sequence.

[0066] In order to remove interference noise in the captured image and improve the image clarity, in this embodiment, preferably, the image signal de-interference in S4 includes a model-based method or a learning-based method.

[0067] The learning-based methods include DnCNN, FFDnet, and CBDnet. DnCNN mainly targets Gaussian noise for denoising, emphasizing the role of residual learning and batch normalization (BN). FFDnet considers generalizing Gaussian noise into more complex real noise, using the noise level map as part of the network input. CBDnet mainly focuses on the noise level map part of FFDnet, adaptively obtaining the noise level map through 5 layers of FCN to achieve blind denoising.

[0068] In order to threshold the acquired image to facilitate target recognition and feature extraction, in this embodiment, preferably, the threshold processing in S4 aims to extract the target object in the image and distinguish the background from the noise. Usually, a threshold T is set, and the pixels of the image are divided into two categories by T: the pixel group greater than T and the pixel group less than T.

[0069] The most important aspect of the threshold processing is binarization.

[0070] The calculation formula for binarization is as follows:

[0071]

[0072] Y represents the grayscale value of the pixel group in the image binarization process, T is the threshold, and gray represents the grayscale value of the input image pixel. All pixels with a grayscale value less than the threshold T have their grayscale value set to 0, representing black; all pixels with a grayscale value greater than or equal to the threshold T have their grayscale value set to 255, representing white.

[0073] In order to perform incremental processing on the acquired images and make the image information easier to identify, in this embodiment, preferably, the image signal increment in step S4 adopts supervised data augmentation, that is, using preset data transformation rules to amplify the data on the basis of existing data, including single-sample data augmentation and multi-sample data augmentation, wherein single samples include geometric operation type and color transformation type.

[0074] The geometric transformation class refers to performing geometric transformations on images, including various operations such as flipping, rotating, cropping, deforming, and scaling.

[0075] The color transformation includes noise, blur, color transformation, erasure, and fill.

[0076] In order to enable subsequent processing of the image and facilitate clear viewing and identification of the image information, in this embodiment, preferably, the post-processing in step S5 refers to the post-processing of the final rendered image after the normal rendering pipeline has ended.

[0077] The post-processing includes anti-aliasing, FXAA, TAA, flood glare, lens flare, color grading, white balance, saturation, contrast, gamma correction, color gain, color shift, tone mapping, depth of field, ambient occlusion, and motion blur processing.

[0078] according to Figure 3 As can be seen, a schematic diagram of a partial principle of an automatic identification method for electronic endoscope lens modules shows that VCC2 is a stable DC power supply from a switching power supply, GND is the common ground of the circuit, D2 is a multi-color indicator light, U1 is the image acquisition CPU, Q1 and Q2 are MOS matching devices, and TEST1, TEST2, and TEST3 come from the first lens module, the second lens module, the third lens module, and the compatible interface assembly, respectively. When the electronic endoscope is in working condition, the image acquisition CPU chip U1 cyclically detects TEST1, TEST2, and TEST3 and performs internal calculations. When a valid signal is detected, it simultaneously determines which channel module the signal comes from, and the indicator light illuminates the corresponding signal color. Then, the CPU chip U1 outputs a drive signal to drive Q1 and Q2 to perform automatic logic matching. CS1 and CS2 complete the logic matching value and transmit it to the central processing unit. The central processing unit then starts the corresponding internal processing module for the currently valid module signal.

[0079] Working principle and usage process of this invention:

[0080] Step 1: First, disinfect the electronic endoscope: Disinfect the tubing of the electronic endoscope and keep the electronic endoscope sterile. Apply gel to the tubing of the electronic endoscope before insertion.

[0081] The second step is to acquire images using an electronic endoscope: When the tube of the electronic endoscope is inserted into the human body, the first lens module A, the second lens module B, and the third lens module C at the end of the tube of the electronic endoscope are used to capture detection point information.

[0082] The third step involves transmitting and processing the image data: the first lens module A, the second lens module B, and the third lens module C transmit the image information to the central processing unit E through the compatible interface D, thus completing the transmission of the image information.

[0083] The fourth step involves further processing the image information: The central processing unit E of the electronic endoscope is compatible with the first lens module A, the second lens module B, and the third lens module C through the compatible interface D. When different lens modules need to be replaced as required, the central processing unit E automatically identifies the lens module currently in use by performing image processing F on the images of different lens modules, successively acquiring and identifying the images, and performing interference removal, threshold processing, and image signal increment calculation on the image signals of the currently effective lens modules.

[0084] Fifth step: Finally, perform post-processing on the image information G: process and distribute the current lens module image signal through post-processing circuits.

[0085] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for automatic identification of electronic endoscope lens modules, characterized in that, It includes the following steps: S1. First, disinfect the electronic endoscope: disinfect the tubes of the electronic endoscope and keep the electronic endoscope sterile. Apply gel to the tubes of the electronic endoscope before inserting it. S2. Image acquisition via electronic endoscope: When the tube of the electronic endoscope is inserted into the human body, the first lens module, the second lens module and the third lens module at the end of the tube of the electronic endoscope are used to capture detection point information. S3. Then, the data image information is transmitted and processed: the first lens module, the second lens module and the third lens module transmit the image information to the central processing unit through a compatible interface to complete the transmission of image information; S4. Further processing of image information: The central processing unit of the electronic endoscope is compatible with multiple lens modules through a compatible interface. When different lens modules need to be replaced according to requirements, the central processing unit automatically identifies the lens module currently in use by performing image processing on the images of different lens modules, performing successive multi-channel acquisition and identification, and performing interference removal, threshold processing, and image signal increment calculation on the image signal of the currently effective lens module. S5. Finally, post-processing of image information: Post-processing and allocation of the current lens module image signal; the post-processing and allocation of the post-processing circuit refers to the post-processing of the final rendered image after the normal rendering pipeline is completed; including anti-aliasing, FXAA, TAA, flood glare, lens flare, color grading, white balance, saturation, contrast, gamma correction, color gain, color shift, tone mapping, depth of field, ambient light occlusion and motion blur processing; The compatibility interface (D) is connected to the first lens module (A), the second lens module (B), the third lens module (C), and the central processing unit (E), respectively. The central processing unit (E) is connected to the image processing (F), the compatibility interface (D), and the post-processing (G), respectively.

2. The method for automatic identification of an electronic endoscope lens module according to claim 1, characterized in that: The disinfection process of the electronic endoscope in S1 involves immediately washing it with clean water after use to remove blood and mucus residue. During the cleaning process, the endoscope surface is cleaned with degreased cotton balls or professional lens cleaning paper. After the initial cleaning, it is soaked in a multi-enzyme cleaning solution. After soaking in the multi-enzyme cleaning solution, the endoscope should be thoroughly rinsed with clean water and dried.

3. The method for automatic identification of an electronic endoscope lens module according to claim 1, characterized in that: The gel in step (S1) refers to a thick liquid or semi-solid preparation with gel properties made from the raw drug and excipients that can form gels. The gel applied before electronic endoscopy usually serves to lubricate, relieve pain and provide anesthesia. The gel includes oxybuprocaine hydrochloride gel or tetracaine hydrochloride gel.

4. The method for automatic identification of an electronic endoscope lens module according to claim 1, characterized in that: In step (S2), the first lens module, the second lens module, and the third lens module are arranged at equal angles, and all three are tilted outwards.

5. The method for automatic identification of an electronic endoscope lens module according to claim 4, characterized in that: The first lens module, the second lens module, and the third lens module all include a fill light and a charge-coupled device (CCD). The CCD is used to take pictures of the detection points, and the fill light is used to provide a light source for the CCD.

6. The method for automatic identification of an electronic endoscope lens module according to claim 5, characterized in that: The charge-coupled device (CCD) has photoelectric conversion function, as well as signal charge storage, transfer, and readout functions. Its working process consists of four steps: The first step is the light integration period, which is the exposure time. During this period, the pixels of the charge-coupled device (CCD) convert the incident light quanta into photogenerated charges in a proportional manner, thus completing the photo-to-electric conversion. The second is to temporarily store the photocharge generated by each pixel in the corresponding photodiode potential well during the light integration process, thereby achieving signal charge storage. Third, after exposure, the stored photogenerated charge is transferred to the output area along the CCD shift register to complete the charge transfer; Fourth, in the readout amplifier, each photogenerated charge is sequentially converted into a corresponding video signal to complete the signal readout. Therefore, the charge-coupled device (CCD) can be regarded as a photoelectric converter, which transforms a spatially distributed optical image into a video voltage signal distributed in time sequence.

7. The method for automatic identification of an electronic endoscope lens module according to claim 1, characterized in that: The image signal de-interference process in (S4) includes model-based methods or learning-based methods; The learning-based methods include DnCNN, FFDnet, and CBDnet. DnCNN mainly targets Gaussian noise for denoising, emphasizing the role of residual learning and batch normalization (BN). FFDnet considers generalizing Gaussian noise into more complex real noise, using the noise level map as part of the network input. CBDnet mainly focuses on the noise level map part of FFDnet, adaptively obtaining the noise level map through 5 layers of FCN to achieve blind denoising.

8. The method for automatic identification of an electronic endoscope lens module according to claim 1, characterized in that: The thresholding process in step (S4) aims to extract target objects in the image and distinguish between the background and noise. Typically, a threshold T is set to divide the pixels of the image into two categories: a group of pixels greater than or equal to T and a group of pixels less than T. The most important aspect of the threshold processing is binarization. The calculation formula for binarization is as follows: Y represents the grayscale value of the pixel group in the image binarization process, T is the threshold, and gray represents the grayscale value of the input image pixel. All pixels with a grayscale value less than the threshold T have their grayscale value set to 0, representing black; all pixels with a grayscale value greater than or equal to the threshold T have their grayscale value set to 255, representing white.

9. The method for automatic identification of an electronic endoscope lens module according to claim 1, characterized in that: The image signal increment in step (S4) is supervised data augmentation, which uses preset data transformation rules to amplify the data based on the existing data. It includes single-sample data augmentation and multi-sample data augmentation. The single sample includes geometric operation type and color transformation type. The geometric transformation class refers to performing geometric transformations on images, including various operations such as flipping, rotating, cropping, deforming, and scaling. The color transformation includes noise, blur, color transformation, erasure, and fill.

Images