Smoke removal system, smoke removal method and endoscope system

Abstract

The embodiment of the invention discloses a smog removal system, a smog removal method and an endoscope system.The smog removal system detects an endoscope image to obtain multi-modal information related to smog, and the multi-modal information comprises smog removal control parameters, smog image parameters and image post-processing parameters; the central control module can control the physical smog removal equipment to perform physical smog removal according to the smog removal control parameters, can control the image smog removal module to perform digital smog removal on the endoscope image according to the smog image parameters to obtain a smog-removed image, and performs optimization processing on the smog-removed image according to the image post-processing parameters. Therefore, the target endoscope image after the smoke is removed is obtained. On the basis, the smoke removal system provided by the embodiment of the invention not only can remove smoke physically, but also can remove smoke digitally, and smoke removal is carried out on the endoscope image through combination of physical smoke removal and digital smoke removal, so that smoke in the endoscope image can be effectively removed.

Description

Technical Field

[0001] This application relates to the field of medical device technology, and in particular to a smoke removal system and method, and an endoscope system. Background Technology

[0002] In minimally invasive surgery, medical endoscopes are indispensable tools. During the surgical procedure, operations such as cutting, peeling, and coagulating human tissues require the participation of energy instruments. In this process, a large amount of smoke is generated in the body cavity, which affects the observation of the endoscope.

[0003] The conventional solution is to deflate the insufficiency unit when the smoke during surgery becomes too dense and interferes with the procedure. However, this increases the risk of surgery or reduces its efficiency. Currently, physical smoke removal using an insufficiency unit is also an option. When smoke removal is needed, medical staff use a foot pedal to control the insufficiency unit to release the air, allowing for smoke removal based on their needs. However, this increases both the workload and the difficulty of the procedure. To avoid this additional workload, automated digital smoke removal using image processing methods has been developed. This eliminates the need for manual operation by the surgeon. However, current digital smoke removal methods are not ideal. Firstly, they are ineffective at high smoke concentrations. Secondly, the images generated after digital smoke removal often differ significantly from the actual surgical field of view, affecting the surgeon's observation and judgment. Summary of the Invention

[0004] The following is an overview of the subject matter described in detail herein. This overview is not intended to limit the scope of the claims.

[0005] This application provides a smoke removal system and method, as well as an endoscope system, which can effectively remove smoke from endoscopic images.

[0006] In a first aspect, embodiments of this application provide a smoke removal system, including:

[0007] The acquisition module is used to acquire endoscopic images of the target object obtained by the endoscope.

[0008] The central control module is used to detect and process the endoscopic images to obtain multimodal information, which includes smoke control parameters, smoke image parameters, and image post-processing parameters.

[0009] A physical smoke removal device is used to physically remove smoke from the target object;

[0010] An image desmoke module is used to digitally desmoke the endoscopic images;

[0011] The image optimization module is used to optimize the smoke-free image obtained after digital smoke removal.

[0012] The central control module is also used to control the physical smoke removal device to perform physical smoke removal according to the smoke removal control parameters, and to control the image smoke removal module to perform digital smoke removal on the endoscope image according to the smoke image parameters to obtain the smoke-removed image. The central control module is also used to control the image optimization module to optimize the smoke-removed image according to the image post-processing parameters to obtain the target endoscope image, and to control the display to show the target endoscope image.

[0013] In one embodiment of this application, the central control module includes a multimodal analysis model, and the central control module is further configured to input the endoscopic image into the multimodal analysis model for detection and processing to obtain the multimodal information.

[0014] In one embodiment of this application, the central control module is used to detect and process the endoscopic image, including:

[0015] The central control module is used to perform smoke feature analysis on the endoscopic images and determine the smoke removal control parameters;

[0016] The central control module is also used to perform smoke image analysis on the endoscopic image to determine the smoke image parameters;

[0017] The central control module is also used to perform image parameter analysis on the endoscopic image and determine the image post-processing parameters.

[0018] In one embodiment of this application, the step of performing smoke feature analysis on the endoscopic image to determine the smoke removal control parameters includes:

[0019] The smoke removal control parameters are obtained by performing image feature analysis on the endoscopic image using a smoke analysis algorithm, or by inputting the endoscopic image into a smoke feature analysis model for recognition processing.

[0020] The step of performing smoke image analysis on the endoscopic image to determine the smoke image parameters includes:

[0021] The smoke image parameters are obtained by performing image analysis on the endoscopic image using a smoke image recognition algorithm, or by inputting the endoscopic image into a smoke image recognition model for recognition processing.

[0022] The step of performing image parameter analysis on the endoscopic image to determine the image post-processing parameters includes:

[0023] The endoscope image is analyzed using an image parameter analysis algorithm to obtain the image post-processing parameters; alternatively, the endoscope image is input into an image parameter analysis model for recognition processing to obtain the image post-processing parameters.

[0024] In one embodiment of this application, the smoke control parameters include at least one of the following:

[0025] Smoke masking information indicating whether smoke masking exists;

[0026] Smoke concentration information characterizing smoke concentration;

[0027] Switch control information used to control the switch of the smoke removal equipment;

[0028] Power control information used to control the smoke removal power of the smoke removal equipment.

[0029] In one embodiment of this application, the smoke image parameters include at least one of the following:

[0030] The distribution information of smoke in the endoscopic image, wherein the distribution information includes at least location information or concentration information;

[0031] Dark channel features in the endoscopic image.

[0032] In one embodiment of this application, the smoke image parameters include dark channel features, and the central control module controls the image desmoke module to digitally desmoke the endoscopic image according to the smoke image parameters to obtain a desmoke-free image, including:

[0033] The dark channel features include atmospheric brightness values ​​and / or transmittance of the endoscopic image;

[0034] The image desmoke module processes the endoscopic image based on the dark channel features to obtain the desmoke-free image.

[0035] In one embodiment of this application, the image desmoking module includes a desmoking model, and the central control module controls the image desmoking module to digitally desmoke the endoscopic image according to the smoke image parameters to obtain a desmoked image, including:

[0036] The central control module inputs the endoscopic image and the smoke image parameters into the smoke removal model for smoke removal processing to obtain the smoke-removed image.

[0037] In one embodiment of this application, the image post-processing parameters include at least one of the following:

[0038] Brightness adjustment parameters;

[0039] Contrast adjustment parameters;

[0040] Sharpness adjustment parameters;

[0041] Color adjustment parameters;

[0042] Saturation adjustment parameters;

[0043] Tone adjustment parameters;

[0044] Regional local enhancement parameters;

[0045] Adjust the noise reduction parameters.

[0046] In one embodiment of this application, the multimodal information further includes endoscope device control parameters, which include at least one of the following:

[0047] Endoscopic light source device control parameters;

[0048] Endoscopic camera control parameters.

[0049] In one embodiment of this application, the central control module is further configured to control the endoscope device according to the endoscope device control parameters, so as to optimize the endoscope images captured by the endoscope.

[0050] In one embodiment of this application, the smoke removal system further includes a pre-processing image module and a post-processing image module. The central control module is further configured to control the pre-processing image module to perform pre-processing on the endoscopic image obtained by the acquisition module. The central control module is configured to perform detection processing on the endoscopic image after the pre-processing to obtain multimodal information. The central control module is further configured to control the image optimization module to optimize the smoke removal image according to the image post-processing parameters, and then perform post-processing on the image after processing by the post-processing image module to obtain the target endoscopic image.

[0051] Secondly, embodiments of this application provide a smoke removal method, including:

[0052] Acquire endoscopic images captured by an endoscope;

[0053] The endoscopic image is processed to obtain smoke image parameters and image post-processing parameters;

[0054] The endoscopic image is digitally desmoke-removed based on the smoke image parameters to obtain a smoke-removed image.

[0055] The smoke-free image is optimized according to the image post-processing parameters to obtain the target endoscope image;

[0056] The control display shows the target endoscopic image.

[0057] In one embodiment of this application, the step of detecting and processing the endoscopic image to obtain smoke image parameters and image post-processing parameters includes:

[0058] The endoscopic image is input into a multimodal analysis model for detection and processing to obtain the smoke image parameters and the image post-processing parameters.

[0059] In one embodiment of this application, the step of detecting and processing the endoscopic image to obtain smoke image parameters and image post-processing parameters includes:

[0060] Smoke image analysis is performed on the endoscopic images to determine the smoke image parameters;

[0061] Image parameter analysis is performed on the endoscopic images to determine the image post-processing parameters.

[0062] In one embodiment of this application, the step of performing smoke image analysis on the endoscopic image to determine the smoke image parameters includes one of the following:

[0063] The smoke image parameters are obtained by analyzing the endoscopic image using a smoke image recognition algorithm.

[0064] Alternatively, the endoscopic image can be input into a smoke image recognition model for recognition processing to obtain the smoke image parameters.

[0065] In one embodiment of this application, the smoke image parameters include at least one of the following:

[0066] The distribution information of smoke in the endoscopic image, wherein the distribution information includes at least location information or concentration information;

[0067] Dark channel features in the endoscopic image.

[0068] In one embodiment of this application, the step of performing image parameter analysis on the endoscopic image to determine the image post-processing parameters includes one of the following:

[0069] The endoscopic image is analyzed using an image parameter analysis algorithm to obtain the image post-processing parameters.

[0070] Alternatively, the endoscopic image can be input into an image parameter analysis model for recognition processing to obtain the image post-processing parameters.

[0071] In one embodiment of this application, the image post-processing parameters include at least one of the following:

[0072] Brightness adjustment parameters;

[0073] Contrast adjustment parameters;

[0074] Sharpness adjustment parameters;

[0075] Color adjustment parameters;

[0076] Saturation adjustment parameters;

[0077] Tone adjustment parameters;

[0078] Regional local enhancement parameters;

[0079] Adjust the noise reduction parameters.

[0080] In one embodiment of this application, the method further includes:

[0081] The endoscopic images are processed to obtain the control parameters of the endoscopic device;

[0082] The endoscope is controlled according to the endoscope control parameters to optimize the endoscope images captured by the endoscope.

[0083] The control parameters of the endoscopic device include at least one of the following:

[0084] Endoscopic light source device control parameters;

[0085] Endoscopic camera control parameters.

[0086] In one embodiment of this application, the smoke image parameters include dark channel features in the endoscopic image; the step of digitally desmoke-removing the endoscopic image based on the smoke image parameters to obtain a desmoke-removed image includes:

[0087] Obtain the atmospheric brightness value and transmittance of the endoscope image included in the dark channel features;

[0088] The endoscopic image is processed based on the atmospheric brightness value and the transmittance to obtain a smoke-free image.

[0089] In one embodiment of this application, the step of digitally removing smoke from the endoscopic image based on the smoke image parameters to obtain a smoke-free image includes:

[0090] The parameters of the endoscope image and the smoke image are input into the smoke removal model for smoke removal processing to obtain the smoke-removed image.

[0091] In one embodiment of this application, the step of detecting and processing the endoscopic image to obtain smoke image parameters and image post-processing parameters includes:

[0092] The endoscopic images are processed to obtain smoke control parameters, smoke image parameters, and image post-processing parameters.

[0093] The smoke removal method also includes:

[0094] The smoke removal equipment is controlled to remove smoke according to the smoke removal control parameters.

[0095] In one embodiment of this application, the step of detecting and processing the endoscopic image to obtain smoke control parameters, smoke image parameters, and image post-processing parameters includes:

[0096] The endoscopic image is input into a multimodal analysis model for detection and processing to obtain the smoke removal control parameters, the smoke image parameters, and the image post-processing parameters.

[0097] In one embodiment of this application, the step of detecting and processing the endoscopic image to obtain smoke control parameters, smoke image parameters, and image post-processing parameters includes:

[0098] Smoke feature analysis is performed on the endoscopic images to determine the smoke removal control parameters;

[0099] Smoke image analysis is performed on the endoscopic images to determine the smoke image parameters;

[0100] Image parameter analysis is performed on the endoscopic images to determine the image post-processing parameters.

[0101] In one embodiment of this application, the step of performing smoke feature analysis on the endoscopic image to determine the smoke removal control parameters includes one of the following:

[0102] The smoke removal control parameters are obtained by performing image feature analysis on the endoscopic image using a smoke analysis algorithm.

[0103] Alternatively, the endoscopic image can be input into a smoke feature analysis model for identification and processing to obtain the smoke removal control parameters.

[0104] In one embodiment of this application, the smoke control parameters include at least one of the following:

[0105] Smoke masking information indicating whether smoke masking exists;

[0106] Smoke concentration information characterizing smoke concentration;

[0107] Switch control information for controlling the switch of the smoke removal equipment.

[0108] Power control information used to control the smoke removal power of the smoke removal equipment.

[0109] In one embodiment of this application, the step of detecting and processing the endoscopic image to obtain smoke image parameters and image post-processing parameters includes:

[0110] The endoscopic image is subjected to pre-processing; the endoscopic image after pre-processing is then subjected to detection processing to obtain the smoke image parameters and the image post-processing parameters.

[0111] And / or, the step of optimizing the smoke-free image according to the image post-processing parameters to obtain the target endoscopic image includes:

[0112] After optimizing the smoke-free image according to the image post-processing parameters, the target endoscope image is obtained through subsequent image processing.

[0113] Thirdly, embodiments of this application provide a smoke removal method, including:

[0114] Acquire endoscopic images captured by an endoscope;

[0115] The endoscopic image is processed to obtain image post-processing parameters;

[0116] The endoscopic image is digitally desmoke-free to obtain a desmoke-free image;

[0117] The smoke-free image is optimized according to the image post-processing parameters to obtain the target endoscope image;

[0118] The control display shows the target endoscopic image.

[0119] In one embodiment of this application, the step of detecting and processing the endoscopic image to obtain image post-processing parameters includes,

[0120] The endoscopic images are processed to obtain smoke control parameters and image post-processing parameters;

[0121] The smoke removal method also includes:

[0122] The smoke removal equipment is controlled to remove smoke according to the smoke removal control parameters.

[0123] In one embodiment of this application, the step of digitally desmoke-removing the endoscopic image and obtaining a desmoke-removed image includes:

[0124] The endoscopic image is input into the smoke removal model for smoke removal processing to obtain the smoke-removed image;

[0125] Alternatively, dark channel features can be extracted from the endoscopic image, including atmospheric brightness values ​​and / or transmittance, and the endoscopic image can be processed based on the dark channel features to obtain the smoke-free image.

[0126] Fourthly, embodiments of this application provide an endoscope system, comprising:

[0127] An endoscope is used to capture endoscopic images of a target object.

[0128] A monitor used to display endoscopic images;

[0129] Smoke removal equipment, used to remove smoke from the target object;

[0130] An endoscope host is connected to the endoscope. The endoscope host is used to acquire endoscopic image data captured by the endoscope and to control the smoke removal device to remove smoke. The endoscope host is also used to execute the smoke removal method as described in the second and third aspects above to obtain a target endoscopic image and display it through the display.

[0131] Fifthly, embodiments of this application provide an endoscope system, comprising:

[0132] An endoscope is used to capture endoscopic images of a target object.

[0133] A monitor used to display endoscopic images;

[0134] An endoscope host is connected to the endoscope and is used to acquire endoscopic image data captured by the endoscope. The endoscope host is also used to perform the smoke removal method as described in the second and third aspects above to obtain a target endoscopic image and display it on the display.

[0135] The embodiments of this application include at least the following beneficial effects:

[0136] On one hand, an embodiment of this application provides a smoke removal system, including an acquisition module, a physical smoke removal device, an image smoke removal module, an image optimization module, and a central control module. The acquisition module acquires endoscopic images of a target object obtained by an endoscope. The central control module processes the endoscopic images to obtain multimodal information, including smoke removal control parameters, smoke image parameters, and image post-processing parameters. The physical smoke removal device physically removes smoke from the target object. The image smoke removal module digitally removes smoke from the endoscopic images. The image optimization module optimizes the smoke-removed image obtained after digital smoke removal. The central control module also controls the physical smoke removal device to perform physical smoke removal according to the smoke removal control parameters, and controls the image smoke removal module to digitally remove smoke from the endoscopic images according to the smoke image parameters to obtain a smoke-removed image. Furthermore, the central control module controls the image optimization module to optimize the smoke-removed image according to the image post-processing parameters to obtain a target endoscopic image, and controls a display to show the target endoscopic image. By detecting endoscopic images, multimodal information related to smoke is obtained. The central control module can control the physical smoke removal device to perform physical smoke removal according to the smoke removal control parameters. It can also control the image smoke removal module to perform digital smoke removal on the endoscopic images according to the smoke image parameters, obtaining a smoke-free image. Furthermore, the smoke-free image is optimized according to image post-processing parameters to obtain a smoke-free optimized target endoscopic image. Therefore, the smoke removal system of this embodiment can perform both physical and digital smoke removal, effectively removing smoke from endoscopic images through a combination of both methods.

[0137] On the other hand, one embodiment of this application provides a smoke removal method that acquires an endoscopic image captured by an endoscope, performs detection processing on the endoscopic image to obtain smoke image parameters and image post-processing parameters. Based on the smoke image parameters, digital desmoke removal is performed on the endoscopic image to obtain a smoke-free image. Then, the smoke-free image is optimized based on the image post-processing parameters to obtain a target endoscopic image. Finally, the target endoscopic image is displayed on a monitor. By digitally desmoke-removing the endoscopic image using smoke image parameters and image post-processing parameters, i.e., by digitally desmoke-removing the endoscopic image based on the smoke image parameters to obtain a smoke-free image, and then optimizing the smoke-free image based on the image post-processing parameters, a target endoscopic image with smoke removed is obtained. Therefore, the smoke removal method of this application embodiment, by digitally desmoke-removing the endoscopic image using smoke image parameters and image post-processing parameters, can effectively remove smoke from the endoscopic image.

[0138] On the other hand, one embodiment of this application provides a smoke removal method that acquires an endoscopic image captured by an endoscope, performs detection processing on the endoscopic image to obtain image post-processing parameters, digitally desmokes the endoscopic image to obtain a smoke-free image, optimizes the smoke-free image according to the image post-processing parameters to obtain a target endoscopic image, and finally controls the display to show the target endoscopic image. By digitally desmokes the endoscopic image to obtain a smoke-free image, and optimizing the smoke-free image according to the image post-processing parameters, a target endoscopic image with smoke removed is obtained. Based on this, the smoke removal method of this application, by digitally desmokes the endoscopic image to obtain a smoke-free image and by optimizing the smoke-free image through image post-processing parameters, can effectively remove smoke from the endoscopic image. Attached Figure Description

[0139] The accompanying drawings are used to provide a further understanding of the technical solutions of this application and constitute a part of the specification. They are used together with the embodiments of this application to explain the technical solutions of this application and do not constitute a limitation on the technical solutions of this application.

[0140] Figure 1 This is a structural block diagram of a smoke removal system provided in one embodiment of this application;

[0141] Figure 2 This is a structural block diagram of a smoke removal system provided in another embodiment of this application;

[0142] Figure 3 This is a flowchart of a smoke removal method provided in one embodiment of this application;

[0143] Figure 4 This is a flowchart of steps 410 to 420 provided in one embodiment of this application;

[0144] Figure 5 This is a flowchart of steps 510 to 520 provided in one embodiment of this application;

[0145] Figure 6 This is a flowchart of steps 610 to 620 provided in one embodiment of this application;

[0146] Figure 7 This is a flowchart of steps 710 to 730 provided in one embodiment of this application;

[0147] Figure 8 This is a flowchart of a smoke removal method provided in another embodiment of this application. Detailed Implementation

[0148] The present application will be further described below with reference to the accompanying drawings and specific embodiments. The described embodiments should not be considered as limitations on the present application, and all other embodiments obtained by those skilled in the art without inventive effort are within the scope of protection of the present application.

[0149] In the following description, references are made to “some embodiments,” which describe a subset of all possible embodiments. However, it is understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0150] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0151] In minimally invasive surgery, medical endoscopes are indispensable tools. During the surgical procedure, operations such as cutting, peeling, and coagulating human tissues require the participation of energy instruments. In this process, a large amount of smoke is generated in the body cavity, which affects the observation of the endoscope.

[0152] The conventional solution is to deflate the insufficiency unit when the smoke during surgery becomes too dense and interferes with the procedure. However, this increases the risk of surgery or reduces its efficiency. Currently, physical smoke removal using an insufficiency unit is also an option. When smoke removal is needed, medical staff use a foot pedal to control the insufficiency unit to release the air, allowing for smoke removal based on their needs. However, this increases both the workload and the difficulty of the procedure. To avoid this additional workload, automated digital smoke removal using image processing methods has been developed. This eliminates the need for manual operation by the surgeon. However, current digital smoke removal methods are not ideal. Firstly, they are ineffective at high smoke concentrations. Secondly, the images generated after digital smoke removal often differ significantly from the actual surgical field of view, affecting the surgeon's observation and judgment.

[0153] Based on this, the present invention proposes a smoke removal system and method, and an endoscope system, which can effectively remove smoke from endoscopic images.

[0154] This invention provides a smoke removal system that can effectively remove smoke from endoscopic images. For example... Figure 1 The diagram shows a structural block diagram of a smoke removal system. The smoke removal system 10 may include a data acquisition module 100, a physical smoke removal device 101, an image smoke removal module 102, an image optimization module 103, a central control module 104, and a display 105.

[0155] The acquisition module 100 is used to acquire endoscopic images of the target object obtained by the endoscope. The acquisition module 100 may include a lens assembly, a light source assembly, and an image sensor. The lens system is responsible for collecting and transmitting light signals. In electronic endoscopes, the lens system also includes advanced CCD / CMOS chips, which can convert light signals into electrical signals. The light source assembly includes, but is not limited to, LED lamps, xenon lamps, or lasers, used to illuminate the internal cavities so that the image sensor can capture clear images. The light source assembly is typically used in conjunction with the image sensor to provide sufficient illumination. The image sensor is responsible for converting the optical image information captured by the endoscope into electrical signals for further processing and display. The image sensor is typically located at the front end of the endoscope, receiving reflected light or fluorescence from the mucosal surface within the body cavity and converting it into digital electrical signals. The image sensor typically employs a CCD sensor or a CMOS sensor. CCD sensors are widely used in endoscopes due to their high sensitivity and good image quality, while CMOS sensors offer advantages such as small size, low power consumption, low cost, and high system integration.

[0156] The physical smoke removal device 101 is used to physically remove smoke from a target object. The physical smoke removal device 101 can be applied to an insufflator and works in conjunction with it to physically remove smoke. The physical smoke removal device 101 can employ a commonly used smoke extractor in the art, such as a controllable smoke extraction device. The physical smoke removal device 101 may include a suction device, a smoke extraction duct, a T-junction, and a circulating smoke extraction system. The suction device uses negative pressure suction to remove smoke generated during surgery, maintaining a clear surgical field. The smoke extraction duct connects the suction device and the endoscope, venting smoke from the surgical area. The T-junction can be connected to the exhaust port of the endoscope, controlling the smoke extraction and suction functions via a valve, enabling the switching between flushing and smoke extraction. The circulating smoke extraction system filters surgical smoke through a filtration system before injecting purified gas into the cavity, achieving real-time circulation and removal of smoke, ensuring a clear surgical field while reducing carbon dioxide loss. In practical applications, the intensity of physical smoke removal by the physical smoke removal device 101 can be controlled according to smoke removal control parameters, including smoke concentration information. For example, the higher the smoke concentration index, the higher the smoke removal intensity of the smoke removal device 101; conversely, the lower the smoke concentration index, the lower the smoke removal intensity of the smoke removal device 101.

[0157] The image desmog module 102 is used for digital desmog removal of endoscopic images. Employing image processing technology, the module can detect and remove smoke from endoscopic images using AI or other traditional algorithms. Digital desmog removal eliminates the need for additional hardware defogging equipment, offering fast processing speed, good real-time performance, and compatibility with different image resolutions. Specifically, the image desmog module 102 analyzes the endoscopic image, estimating the smoke concentration using principles such as Dark Channel Prior (DCP). Simultaneously, it estimates the atmospheric brightness value in the endoscopic image, typically achieved by identifying the brightest point. Dark channel features include atmospheric brightness and / or transmittance. Transmittance describes the degree to which light is attenuated by smoke particles during propagation. Using the estimated transmittance and atmospheric brightness values, a specific desmog algorithm, such as an improved U-Net network, is applied to the endoscopic image to restore its clarity, resulting in a smoke-free endoscopic image.

[0158] The image optimization module 103 is used to optimize the smoke-free image obtained after digital smoke removal. The image optimization module 103 can optimize the smoke-free image obtained after digital smoke removal according to image post-processing parameters. Image post-processing parameters include, but are not limited to, ISP control parameters, which may include exposure time, analog gain, digital gain, white balance, black level correction, etc. ISP control parameters are used to control the sensor and lens to ensure image quality. Exposure time is used to adjust the brightness and contrast of the image to ensure appropriate exposure under different lighting conditions. Analog gain and digital gain are used to enhance the brightness of the image, especially in low-light environments, by increasing the gain to improve image visibility. White balance is used to correct the color temperature of the image to ensure color consistency under different light sources. Black level correction is used to adjust the dark details of the image, reduce the impact of dark current on the image, and ensure clear dark details. Optimizing the smoke-free image obtained after digital smoke removal using image post-processing parameters can optimize the smoke-free image processing workflow, improve the saturation, contrast, sharpness, and color accuracy of the smoke-free image, thereby improving the overall quality and performance of the smoke-free image.

[0159] The central control module 104 is used to detect and process endoscopic images to obtain multimodal information, including smoke removal control parameters, smoke image parameters, and image post-processing parameters. The smoke removal control parameters include, but are not limited to: smoke obstruction information (indicating the presence of smoke obstruction), smoke concentration information (indicating smoke concentration), switch control information for controlling the on / off state of the smoke removal equipment, and power control information for controlling the smoke removal power of the equipment. The smoke image parameters include, but are not limited to: smoke distribution information in the endoscopic image and dark channel features in the endoscopic image. The image post-processing parameters include, but are not limited to: brightness adjustment parameters, contrast adjustment parameters, sharpness adjustment parameters, color adjustment parameters, saturation adjustment parameters, hue adjustment parameters, local enhancement parameters, and noise reduction adjustment parameters.

[0160] The central control module 104 is also used to control the physical smoke removal device 101 to perform physical smoke removal according to the smoke removal control parameters, and to control the image smoke removal module 102 to perform digital smoke removal on the endoscope image and obtain a smoke-removed image according to the smoke image parameters. The central control module is also used to control the image optimization module 103 to optimize the smoke-removed image according to the image post-processing parameters to obtain the target endoscope image, and to control the display 105 to display the target endoscope image. Based on this, the central control module 104 can detect the endoscope image transmitted by the acquisition module 100 and obtain multimodal information. On the one hand, it calculates the smoke concentration information and outputs the smoke removal control parameters to control the smoke removal intensity of the physical smoke removal device 101; on the other hand, it calculates and outputs the smoke image parameters and image post-processing parameters to perform digital smoke removal and adaptive optimization of the ISP parameters of the endoscope image.

[0161] In one embodiment, the central control module 104 may consist of a detection module and an image parameter control module. The detection module is used to detect and process the endoscopic image to obtain multimodal information, wherein the multimodal information includes smoke removal control parameters, smoke image parameters, and image post-processing parameters. The image parameter control module is used to control the physical smoke removal device 101 to perform physical smoke removal according to the smoke removal control parameters, and to control the image smoke removal module 102 to perform digital smoke removal on the endoscopic image according to the smoke image parameters to obtain a smoke-removed image. The central control module is also used to control the image optimization module 103 to optimize the smoke-removed image according to the image post-processing parameters to obtain a target endoscopic image, and to control the display 105 to display the target endoscopic image.

[0162] The display 105 is used to display the target endoscopic image. The display 105 can be a touch screen, liquid crystal display screen or the like installed in the smoke removal system 10, or it can be an independent display device such as a liquid crystal display or a television set, separate from the smoke removal system 10, or it can be a display screen on an electronic device such as a mobile phone or tablet computer.

[0163] An embodiment of this application provides a smoke removal system 10, which includes a data acquisition module 100, a physical smoke removal device 101, an image smoke removal module 102, an image optimization module 103, and a central control module 104. The acquisition module 100 is used to acquire endoscopic images of the target object obtained by the endoscope; the central control module 104 is used to detect and process the endoscopic images to obtain multimodal information, including smoke removal control parameters, smoke image parameters, and image post-processing parameters; the physical smoke removal device 101 is used to physically remove smoke from the target object; the image smoke removal module 102 is used to digitally remove smoke from the endoscopic images; the image optimization module 103 is used to optimize the smoke-removed image obtained after digital smoke removal; the central control module 104 is also used to control the physical smoke removal device 101 to perform physical smoke removal according to the smoke removal control parameters, and to control the image smoke removal module 102 to digitally remove smoke from the endoscopic images according to the smoke image parameters to obtain smoke-removed images. The central control module 104 is also used to control the image optimization module 103 to optimize the smoke-removed images according to the image post-processing parameters to obtain the target endoscopic image, and to control the display 105 to display the target endoscopic image. By detecting endoscopic images, multimodal information related to smoke is obtained. The central control module 104 can control the physical smoke removal device 101 to perform physical smoke removal according to smoke removal control parameters. It can also control the image smoke removal module 102 to perform digital smoke removal on the endoscopic images according to smoke image parameters and obtain a smoke-removed image. Furthermore, it can control the image optimization module 103 to optimize the smoke-removed image according to image post-processing parameters, thereby obtaining a target endoscopic image after smoke removal. Based on this, the smoke removal system of this embodiment can perform both physical and digital smoke removal. By combining physical and digital smoke removal to remove smoke from endoscopic images, smoke in endoscopic images can be effectively removed.

[0164] It should be noted that, Figure 1 The structure shown is for illustrative purposes only and may include more... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown. Figure 1 The components shown can be implemented in hardware and / or software.

[0165] In one embodiment, the central control module 104 includes a multimodal analysis model. The central control module 104 is also used to input endoscopic images into the multimodal analysis model for detection and processing to obtain multimodal information.

[0166] The multimodal analysis model is a deep learning model, such as a convolutional neural network (CNN) or a U-Net network. The central control module 104 includes the multimodal analysis model. By inputting endoscopic images into the multimodal analysis model, the central control module 104 analyzes and processes the endoscopic images, completing multimodal recognition in one step to obtain multimodal information, namely, obtaining smoke removal control parameters, smoke image parameters, and image post-processing parameters in one go. Based on this, the central control module 104 can calculate the smoke removal control parameters and control the smoke removal intensity of the physical smoke removal device 101 according to these parameters. For example, the smoke removal control parameter uses the smoke concentration index; the higher the smoke concentration index, the higher the smoke removal intensity of the smoke removal device 101; conversely, the lower the smoke concentration index, the lower the smoke removal intensity of the smoke removal device 101. The central control module 104 also calculates and outputs smoke image parameters and image post-processing parameters to perform adaptive optimization of digital smoke removal and ISP parameters of the endoscopic images.

[0167] In one embodiment, the central control module 104 is used to detect and process endoscopic images, including:

[0168] The central control module 104 is used to perform smoke feature analysis on endoscopic images and determine smoke control parameters.

[0169] The central control module 104 is also used to perform smoke image analysis on the endoscopic images and determine the smoke image parameters;

[0170] The central control module 104 is also used to perform image parameter analysis on endoscopic images and determine image post-processing parameters.

[0171] The central control module 104 can perform multimodal recognition on endoscopic images in three image analysis steps. Specifically, the central control module 104 performs smoke feature analysis on the acquired endoscopic images to determine smoke removal control parameters. The central control module 104 also performs smoke image analysis on the endoscopic images to determine smoke image parameters. Furthermore, the central control module 104 performs image parameter analysis on the endoscopic images to determine image post-processing parameters. Based on this, the central control module 104 can obtain smoke removal control parameters through smoke feature analysis and control the smoke removal intensity of the physical smoke removal device 101 according to these parameters. For example, the smoke removal control parameters use a smoke concentration index; the higher the smoke concentration index, the higher the smoke removal intensity of the smoke removal device 101; conversely, the lower the smoke concentration index, the lower the smoke removal intensity of the smoke removal device 101. Simultaneously, the central control module 104 also performs smoke feature analysis and outputs smoke image parameters, as well as image parameter analysis and outputs image post-processing parameters, to perform digital smoke removal and adaptive optimization of the endoscopic images.

[0172] In one embodiment, smoke feature analysis is performed on endoscopic images to determine smoke control parameters, including:

[0173] The smoke control parameters can be obtained by performing image feature analysis on the endoscopic image using a smoke analysis algorithm, or by inputting the endoscopic image into a smoke feature analysis model for recognition processing.

[0174] Based on this, smoke control parameters can be obtained by directly analyzing the image features of endoscopic images using smoke analysis algorithms, or by using smoke feature analysis models to identify and process endoscopic images to obtain smoke control parameters.

[0175] In one embodiment, smoke image analysis is performed on the endoscopic image to determine smoke image parameters, including:

[0176] Smoke image parameters can be obtained by analyzing endoscopic images using a smoke image recognition algorithm, or by inputting endoscopic images into a smoke image recognition model for recognition processing.

[0177] Based on this, smoke image parameters can be obtained by directly analyzing endoscopic images using smoke image recognition algorithms, or by using smoke image recognition models to process endoscopic images and obtain smoke image parameters.

[0178] In one embodiment, image parameter analysis is performed on the endoscopic image to determine image post-processing parameters, including:

[0179] Image parameter analysis algorithms are used to analyze the endoscopic images to obtain image post-processing parameters. Alternatively, the endoscopic images can be input into an image parameter analysis model for recognition processing to obtain image post-processing parameters.

[0180] Based on this, image post-processing parameters can be obtained directly from endoscopic images through image parameter analysis algorithms, or image post-processing parameters can be obtained from endoscopic images through recognition processing using image parameter analysis models.

[0181] In one embodiment, the smoke control parameters include at least one of the following:

[0182] Smoke masking information indicating whether smoke masking exists;

[0183] Smoke concentration information characterizing smoke concentration;

[0184] Switch control information used to control the switching on and off of smoke removal equipment;

[0185] Power control information used to control the smoke removal power of smoke removal equipment.

[0186] In one embodiment, the smoke image parameters include at least one of the following:

[0187] Distribution information of smoke in endoscopic images, wherein the distribution information includes at least location information or concentration information; or / and

[0188] Dark channel features in endoscopic images.

[0189] In one embodiment, the smoke image parameters include the distribution information of smoke in the endoscopic image, wherein the distribution information includes at least location information or concentration information. In this embodiment, the smoke distribution information can be obtained by calculating dark channel features using an atmospheric light prior model; alternatively, it can be obtained by predicting features from large datasets using deep learning and a data model.

[0190] In one embodiment, the smoke image parameters include dark channel features. The central control module 104 controls the image desmoke module 102 to digitally desmoke the endoscopic image based on the smoke image parameters and obtain a desmoke-free image, including:

[0191] The image desmoke module 102 processes the endoscope image based on the atmospheric brightness value and / or transmittance of the endoscope image included in the dark channel features to obtain a desmoke image.

[0192] In this embodiment of the application, endoscopic image processing utilizes a dark channel method to remove smoke from endoscopic images. The image desmoking module 102 processes the endoscopic image based on dark channel features to obtain a desmoked image, thereby removing smoke from the endoscopic image and improving its clarity and visibility. Dark channel features in an endoscopic image refer to the presence of a maximum intensity value for at least one color channel (red, green, or blue) within a local region of the endoscopic image. This local maximum intensity value can be used as an estimation result of atmospheric light. In the dark channel prior dehazing algorithm, dark channel features are used to estimate the fog density in the image, thereby restoring image clarity. In the endoscopic image, atmospheric light can be estimated by searching for local maximum intensity values, determining the atmospheric brightness value of the endoscopic image, and thus constructing a dark channel image. The atmospheric brightness value is the light intensity value in the image due to atmospheric scattering, which can be estimated by finding the brightest point in the image. Specifically, the transmittance can be calculated by obtaining the atmospheric brightness value from the dark channel features. Transmittance describes the degree to which light is attenuated by smoke particles during propagation. The image is reconstructed using estimated transmittance and atmospheric brightness values, thus obtaining a smoke-free image. These steps effectively remove smoke from endoscopic images, improving image quality and providing surgeons with a clearer surgical field of vision. This method not only enhances surgical safety but also improves image processing performance in computer-assisted surgery.

[0193] In one embodiment, the image desmoke module includes a desmoke model, and the central control module 104 controls the image desmoke module to digitally desmoke the endoscopic image according to the smoke image parameters to obtain a desmoke-free image, including:

[0194] The central control module 104 inputs the endoscopic image and smoke image parameters into the smoke removal model for smoke removal processing to obtain a smoke-removed image.

[0195] In this embodiment of the application, endoscopic image processing utilizes a smoke removal model to remove smoke from endoscopic images. Specifically, the central control module 104 inputs endoscopic image and smoke image parameters into the smoke removal model for smoke removal processing, resulting in a smoke-free image. The smoke image parameters may include the distribution information of smoke in the endoscopic image and dark channel features in the endoscopic image. The smoke removal model may include a YOLO-based smoke detection model, a U-Net-based smoke purification algorithm, and other deep neural network-based endoscopic image smoke purification models. The smoke removal model uses different mechanisms and algorithms, such as dark channel priors, attention mechanisms, and adversarial learning, to process endoscopic images for smoke removal, improving the clarity and visibility of the endoscopic images, thereby providing doctors with a more accurate surgical field of view.

[0196] In one embodiment, the image post-processing parameters include, but are not limited to: brightness adjustment parameters, contrast adjustment parameters, sharpness adjustment parameters, color adjustment parameters, saturation adjustment parameters, hue adjustment parameters, local enhancement parameters, and noise reduction adjustment parameters. Optimizing the smoke-free image obtained after digital smoke removal using these image post-processing parameters can streamline the smoke-free image processing workflow, improve the saturation, contrast, sharpness, and color accuracy of the smoke-free image, thereby enhancing the overall quality and performance of the smoke-free image.

[0197] In one embodiment of this application, the multimodal information further includes endoscope device control parameters, which include, but are not limited to, endoscope light source device control parameters and endoscope camera control parameters. The endoscope light source device control parameters are used to adjust the endoscope light source; the endoscope camera control parameters are used to adjust the endoscope camera, for example, focusing and field of view.

[0198] The central control module 104 is used to detect and process endoscopic images and obtain endoscopic device control parameters. That is, in addition to the smoke image parameters and image post-processing parameters obtained based on endoscopic image detection being used for digital smoke removal, in this embodiment, the endoscopic device control parameters obtained based on endoscopic image detection can also be used to control other physical machines besides the physical smoke removal device 101 to perform physical smoke removal.

[0199] In one embodiment, the central control module 104 is further configured to control the endoscope according to the endoscope control parameters, so as to optimize the endoscope images captured by the endoscope.

[0200] In one embodiment of this application, the central control module 104 can also control the endoscope device to perform physical optimization according to the endoscope device control parameters. For example, when the central control module 104 detects that the endoscope image field of view is not clear enough, it generates endoscope device control parameters, such as light source parameters and focus parameters, and performs physical optimization of the endoscope device through the endoscope device control parameters, so as to make the endoscope image captured by the endoscope clearer.

[0201] In one embodiment, the smoke removal system further includes a pre-processing image module 106 and a post-processing image module 107. The central control module 104 is also used to control the pre-processing image module 106 to perform pre-processing on the endoscope images obtained by the acquisition module 100. The central control module 104 is used to perform detection processing on the endoscope images after pre-processing to obtain multimodal information. The central control module 104 is also used to control the image optimization module 103 to optimize the smoke removal images according to the image post-processing parameters, and then perform post-processing on the image image module 107 to obtain the target endoscope image.

[0202] In one embodiment of this application, such as Figure 2 The diagram shown is a structural block diagram of another smoke removal system. This smoke removal system 10 may include a data acquisition module 100, a physical smoke removal device 101, an image smoke removal module 102, an image optimization module 103, a central control module 104, a display 105, a front-end image processing module 106, and a back-end image processing module 107. The functions of the data acquisition module 100, the physical smoke removal device 101, the image smoke removal module 102, the image optimization module 103, the central control module 104, and the display 105 are the same as described above and will not be repeated here.

[0203] The pre-processing image module 106 is used to perform pre-processing on the endoscopic images acquired by the acquisition module 100. Pre-processing refers to a series of processing steps performed on the raw image before it is further analyzed or displayed. These steps aim to improve image quality, reduce noise, correct distortion, and prepare image data for subsequent processing. Pre-processing includes, but is not limited to, denoising, geometric correction, and bad pixel correction. Pre-processing plays a crucial role in improving image quality, ensuring that the image data provides accurate color and good visual effects after being transmitted to the central control module 104. Exemplarily, the pre-processing image module 106 may be a pre-processing ISP (Image Signal Processing) module.

[0204] The post-processing image module 107 is used to perform post-processing on the smoke-removal image optimized by the image optimization module 103. Post-processing refers to further processing steps performed on the image after initial analysis or processing. These steps aim to improve the visual effect of the image, enhance features, or prepare it for specific application purposes, such as output or display. Post-processing includes, but is not limited to, image enhancement, color correction, gamma correction, and other image processing. These post-processing steps ensure that the image can better reproduce scene details under different optical conditions, greatly improving image quality and visual effect. For example, the post-processing image module 107 can be a post-ISP image processing module.

[0205] In one embodiment of this application, the central control module 104 of the aforementioned smoke removal system 10 can be implemented by software, hardware, firmware, or a combination thereof. It can use at least one of the following: application-specific integrated circuit (ASIC), digital signal processor (DSP), digital signal processing device (DSPD), programmable logic device (PLD), field-programmable gate array (FPGA), central processing unit (CPU), controller, microcontroller, and microprocessor, so that the processor 105 can execute the corresponding steps of the smoke removal method in the various embodiments of this application.

[0206] The embodiments of the smoke removal system provided in this application can also perform the following smoke removal methods, and specific embodiments are described below using smoke removal methods.

[0207] like Figure 3 As shown in the figure, this application provides a smoke removal method, which includes:

[0208] Step 110: Obtain endoscopic images captured by the endoscope.

[0209] In this step, the doctor or technician inserts an endoscope through the body's natural cavities or a small surgical incision to reach the area requiring examination or treatment. The endoscope emits a light source to illuminate the internal cavity, allowing the image sensor to capture a clear image. The optical image captured by the image sensor (such as a CCD or CMOS) at the endoscope tip is converted into an electrical signal. Located at the front of the camera system, the image sensor is responsible for collecting images and converting light signals into electrical signals, which are then transmitted to the camera system via the endoscope's circuitry to obtain the endoscopic image.

[0210] In one embodiment, the following can be used: Figure 1 The acquisition module 100 acquires endoscopic images captured by the endoscope. The acquisition module 100 may include a lens assembly, a light source assembly, and an image sensor. The lens system is responsible for collecting and transmitting light signals. In electronic endoscopes, the lens system also includes advanced CCD / CMOS chips that convert light signals into electrical signals. The light source assembly includes, but is not limited to, LED lamps, xenon lamps, or lasers, used to illuminate the internal cavities so that the image sensor can capture clear images. The light source assembly is typically used in conjunction with the image sensor to provide sufficient illumination. The image sensor is responsible for converting the optical image information captured by the endoscope into electrical signals for further processing and display. The image sensor is typically located at the front end of the endoscope and receives reflected light and / or fluorescence from the mucosal surface within the body cavity, converting it into digital electrical signals. The image sensor typically employs a CCD sensor or a CMOS sensor. CCD sensors are widely used in endoscopes due to their high sensitivity and good image quality, while CMOS sensors offer advantages such as small size, low power consumption, low cost, and high system integration.

[0211] Step 120: Detect and process the endoscopic image to obtain smoke image parameters and image post-processing parameters.

[0212] In this step, the endoscopic image is processed to obtain multimodal information, including smoke image parameters and image post-processing parameters. The smoke image parameters include, but are not limited to: (1) the distribution information of smoke in the endoscopic image, including location or concentration information; and (2) the dark channel features in the endoscopic image, including atmospheric brightness and / or transmittance. The image post-processing parameters include, but are not limited to: brightness adjustment parameters, contrast adjustment parameters, sharpness adjustment parameters, color adjustment parameters, saturation adjustment parameters, hue adjustment parameters, local enhancement parameters, and noise reduction adjustment parameters.

[0213] Step 130: Perform digital desmoke removal on the endoscopic image based on the smoke image parameters to obtain a smoke-removed image.

[0214] In this step, image processing technology is employed. AI or other traditional algorithms can be used to detect and remove smoke from endoscopic images. Digital desmoking eliminates the need for additional hardware defogging equipment, offering fast processing speed, good real-time performance, and compatibility with different image resolutions. Digital desmoking can be performed on endoscopic images based on smoke image parameters to obtain a smoke-free image. These parameters can include dark channel features in the endoscopic image. Specifically, taking dark channel desmoking as an example, the smoke concentration in the endoscopic image can be estimated by analyzing the image, such as using the Dark Channel Prior (DCP) principle. Simultaneously, the atmospheric brightness value in the endoscopic image is estimated, typically by identifying the brightest point in the image. Processing is then performed based on the atmospheric brightness value included in the dark channel features and the transmittance of each pixel in the image. Transmittance describes the degree to which light is attenuated during propagation due to smoke particles. Using the estimated transmittance and atmospheric light intensity, a specific desmoking algorithm, such as an improved U-Net network, is used to desmoke the endoscopic image, restoring its clarity and resulting in a smoke-free endoscopic image.

[0215] Step 140: Optimize the smoke-free image according to the image post-processing parameters to obtain the target endoscope image.

[0216] In this step, the de-smoked image obtained after digital de-smoking can be optimized based on image post-processing parameters. These parameters include, but are not limited to, ISP control parameters, which may include exposure time, analog gain, digital gain, white balance, and black level correction. ISP control parameters are used to control the sensor and lens to ensure image quality. Exposure time is used to adjust the image's brightness and contrast, ensuring appropriate exposure under different lighting conditions. Analog and digital gain are used to enhance image brightness, especially in low-light environments, increasing visibility by increasing gain. White balance is used to correct the image's color temperature, ensuring color consistency under different light sources. Black level correction is used to adjust shadow details, reducing the impact of dark current and ensuring clear shadow details. Optimizing the de-smoked image obtained after digital de-smoking using image post-processing parameters optimizes the de-smoked image processing workflow, improving the saturation, contrast, sharpness, and color accuracy of the de-smoked image, thereby improving the overall quality and performance of the de-smoked image.

[0217] Step 150: Control the display to show the target endoscope image.

[0218] In this step, a monitor can be used to display the target endoscopic image. The monitor can be a touch screen or LCD screen installed in the smoke removal system, or it can be a standalone display device such as an LCD screen or television set independent of the smoke removal system, or it can be a display screen on an electronic device such as a mobile phone or tablet.

[0219] The smoke removal method provided in this application involves acquiring an endoscopic image captured by an endoscope, performing detection processing on the endoscopic image to obtain smoke image parameters and image post-processing parameters. Based on the smoke image parameters, digital desmoke removal is performed on the endoscopic image to obtain a smoke-free image. Then, the smoke-free image is optimized based on the image post-processing parameters to obtain a target endoscopic image. Finally, the target endoscopic image is displayed on a monitor. The method involves digitally desmoke-removing the endoscopic image using smoke image parameters and image post-processing parameters, thereby obtaining a smoke-free target endoscopic image. Therefore, the smoke removal method in this application effectively removes smoke from endoscopic images by digitally desmoke-removing the endoscopic image using smoke image parameters and image post-processing parameters.

[0220] In one embodiment, step 120 includes:

[0221] Endoscopic images are input into a multimodal analysis model for detection and processing to obtain smoke image parameters and image post-processing parameters.

[0222] The multimodal analysis model is a deep learning model, such as a convolutional neural network (CNN) or a U-Net network. By inputting endoscopic images into the multimodal analysis model, the endoscopic images are analyzed and processed, enabling multimodal recognition to be completed in one step to obtain multimodal information, namely, obtaining smoke image parameters and image post-processing parameters in one go. The smoke image parameters include, but are not limited to: the distribution information of smoke in the endoscopic image and the dark channel features in the endoscopic image. The image post-processing parameters include, but are not limited to: brightness adjustment parameters, contrast adjustment parameters, sharpness adjustment parameters, color adjustment parameters, saturation adjustment parameters, hue adjustment parameters, local enhancement parameters, and noise reduction adjustment parameters. Based on the smoke image parameters and image post-processing parameters, digital smoke removal and adaptive optimization of ISP control parameters of the endoscopic image can be achieved.

[0223] In one embodiment, such as Figure 4 As shown, step 120 includes:

[0224] Step 410: Perform smoke image analysis on the endoscopic image to determine the smoke image parameters;

[0225] Step 420: Perform image parameter analysis on the endoscopic image to determine the image post-processing parameters.

[0226] In this step, multimodal recognition of the endoscopic images can be performed in stages. Specifically, smoke image analysis is performed on the acquired endoscopic images to determine smoke image parameters. Image parameter analysis is also performed on the endoscopic images to determine image post-processing parameters. Smoke image parameters include, but are not limited to: the distribution information of smoke in the endoscopic image and the dark channel features in the endoscopic image. Image post-processing parameters include, but are not limited to: brightness adjustment parameters, contrast adjustment parameters, sharpness adjustment parameters, color adjustment parameters, saturation adjustment parameters, hue adjustment parameters, local enhancement parameters, and noise reduction adjustment parameters. Based on this, by analyzing and obtaining the smoke image parameters and image post-processing parameters, digital smoke removal and adaptive optimization of ISP parameters of the endoscopic images can be achieved.

[0227] In one embodiment, step 410 includes one of the following:

[0228] Smoke image parameters are obtained by analyzing endoscopic images using a smoke image recognition algorithm;

[0229] Alternatively, the endoscopic image can be input into a smoke image recognition model for recognition processing to obtain smoke image parameters.

[0230] Based on this, smoke image parameters can be obtained by directly analyzing endoscopic images using smoke image recognition algorithms, or by using smoke image recognition models to process endoscopic images and obtain smoke image parameters.

[0231] In one embodiment, step 420 includes one of the following:

[0232] Image parameter analysis algorithms are used to analyze the image parameters of endoscopic images to obtain image post-processing parameters.

[0233] Alternatively, the endoscopic image can be input into an image parameter analysis model for recognition and processing to obtain image post-processing parameters.

[0234] Based on this, image post-processing parameters can be obtained directly from endoscopic images through image parameter analysis algorithms, or image post-processing parameters can be obtained from endoscopic images through recognition processing using image parameter analysis models.

[0235] In one embodiment, such as Figure 5 As shown, the smoke removal method also includes:

[0236] Step 510: Detect and process the endoscopic image to obtain the control parameters of the endoscopic device;

[0237] Step 520: Control the endoscope according to the endoscope control parameters to optimize the endoscopic images captured by the endoscope.

[0238] In this step, the control parameters of the endoscopic equipment include, but are not limited to, the control parameters of the endoscopic light source device and the control parameters of the endoscopic camera. Specifically, the control parameters of the endoscopic light source device are used to adjust the endoscopic light source; the control parameters of the endoscopic camera are used to adjust the endoscopic camera, for example, focusing and field of view.

[0239] By detecting and processing endoscopic images, control parameters for the endoscopic device can also be obtained. That is, in addition to the smoke image parameters and post-processing parameters obtained from endoscopic image detection being used for digital smoke removal, in this embodiment, the control parameters obtained from endoscopic image detection can also be used to control the endoscopic device to perform physical smoke removal.

[0240] It can also control the endoscope to perform physical optimization based on the endoscope's control parameters. For example, when the endoscope image field of view is not clear enough, endoscope control parameters, such as light source parameters and focus parameters, are generated. The endoscope is then physically optimized using the endoscope control parameters, thereby making the endoscope image captured by the endoscope clearer.

[0241] In one embodiment, such as Figure 6 As shown, the smoke image parameters include dark channel features in the endoscopic image, and step 130 includes:

[0242] Step 610: Obtain the atmospheric brightness value and transmittance of the endoscopic image included in the dark channel features;

[0243] Step 620: Process the endoscopic image based on the atmospheric brightness value and transmittance to obtain a smoke-free image.

[0244] In this step, the embodiment of this application, in the processing of endoscopic images, can use the dark channel method to remove smoke from endoscopic images. Specifically, based on dark channel features, the endoscopic image is processed according to atmospheric brightness and transmittance to obtain a smoke-free image, thereby removing smoke from the endoscopic image and improving its clarity and visibility. Dark channel features in endoscopic images refer to the presence of at least one maximum intensity value for a color channel (red, green, or blue) within a local region of the endoscopic image. This local maximum intensity value can be used as an estimate of atmospheric light. In the dark channel prior dehazing algorithm, dark channel features are used to estimate the fog density in the image, thereby restoring image clarity. In endoscopic images, atmospheric light can be estimated by searching for local maximum intensity values, determining the atmospheric brightness value of the endoscopic image, and thus constructing the dark channel image. Atmospheric brightness is the light intensity value in the image caused by atmospheric scattering, which can be estimated by finding the brightest point in the image. Specifically, transmittance can be obtained from the dark channel features; transmittance describes the degree to which light is attenuated by smoke particles during propagation. The image is reconstructed using estimated transmittance and atmospheric brightness values, resulting in a smoke-free image. These steps effectively remove smoke from endoscopic images, improving image quality and providing surgeons with a clearer surgical field of view. This method not only enhances surgical safety but also improves image processing performance in computer-assisted surgery. It should be noted that dark channel features include, but are not limited to, atmospheric brightness values ​​and / or transmittance of the endoscopic image; other features may also be present besides atmospheric brightness and transmittance. By acquiring dark channel features and processing the endoscopic image based on these features, a smoke-free image can be obtained.

[0245] In one embodiment, step 130 includes:

[0246] The parameters of the endoscopic image and the smoke image are input into the smoke removal model for smoke removal processing to obtain the smoke-removed image.

[0247] In this step, the embodiment of this application utilizes AI methods to remove smoke from endoscopic images during image processing. Specifically, a smoke removal model is used to remove smoke from the endoscopic images. The central control module 104 inputs the endoscopic image and smoke image parameters into the smoke removal model for processing, resulting in a smoke-free image. The smoke image parameters can include the distribution information of smoke in the endoscopic image and the dark channel features in the endoscopic image. The smoke removal model can include a YOLO-based smoke detection model, a U-Net-based smoke purification algorithm, and other deep neural network-based endoscopic image smoke purification models. The smoke removal model uses different mechanisms and algorithms, such as dark channel priors, attention mechanisms, and adversarial learning, to process endoscopic images for smoke removal, improving the clarity and visibility of the endoscopic images, thereby providing doctors with a more accurate surgical field of view.

[0248] In one embodiment, step 120 includes:

[0249] Step 121: Detect and process the endoscopic image to obtain smoke control parameters, smoke image parameters, and image post-processing parameters;

[0250] Smoke removal methods also include:

[0251] The smoke removal equipment is controlled according to the smoke removal control parameters to remove smoke.

[0252] In this step, the physical smoke removal equipment can be controlled according to the smoke removal control parameters to perform physical smoke removal, and the endoscopic image can be digitally desmoke-removed according to the smoke image parameters to obtain a smoke-removed image. The smoke-removed image can then be optimized according to the image post-processing parameters to obtain the target endoscopic image, which is then displayed. Based on this, multimodal information is obtained by detecting the endoscopic image, including smoke removal control parameters, smoke image parameters, and image post-processing parameters. Based on this, on the one hand, the smoke concentration information is calculated, and the smoke removal control parameters are output to control the smoke removal intensity of the physical smoke removal equipment; on the other hand, the smoke image parameters and image post-processing parameters are calculated and output to perform digital smoke removal of the endoscopic image and adaptive optimization of the ISP parameters.

[0253] It should be noted that the smoke control parameters include, but are not limited to: smoke obstruction information indicating the presence of smoke obstruction, smoke concentration information indicating smoke concentration, switch control information for controlling the switching on and off of the smoke removal equipment, and power control information for controlling the smoke removal power of the smoke removal equipment.

[0254] For physical smoke removal, for example, the smoke concentration information, such as the smoke concentration index, can be calculated, and the smoke removal control parameters can be output based on the smoke concentration index to control the smoke removal intensity of the physical smoke removal device 101. For example, the higher the smoke concentration index, the higher the smoke removal intensity of the smoke removal device 101; conversely, the lower the smoke concentration index, the lower the smoke removal intensity of the smoke removal device 101.

[0255] In one embodiment, step 121 includes:

[0256] Endoscopic images are input into a multimodal analysis model for detection and processing to obtain smoke control parameters, smoke image parameters, and image post-processing parameters.

[0257] In this step, the multimodal analysis model is a deep learning model, such as a convolutional neural network (CNN) or a U-Net network. By inputting the endoscopic image into the multimodal analysis model, the endoscopic image is analyzed and processed, allowing for multimodal recognition to be completed in one step to obtain multimodal information, namely, smoke removal control parameters, smoke image parameters, and image post-processing parameters. Based on this, the endoscopic image is input into the multimodal analysis model for detection processing. The multimodal analysis model outputs the smoke removal control parameters, which include, but are not limited to, smoke concentration information. Taking the smoke concentration index as an example, the smoke removal intensity of the physical smoke removal equipment is controlled according to the smoke concentration index. For example, the higher the smoke concentration index, the higher the smoke removal intensity of the smoke removal equipment; conversely, the lower the smoke concentration index, the lower the smoke removal intensity of the smoke removal equipment. At the same time, the multimodal analysis model also outputs smoke image parameters and image post-processing parameters for digital smoke removal and adaptive optimization of the endoscopic image.

[0258] In one embodiment, such as Figure 7 As shown, step 121 includes:

[0259] Step 710: Perform smoke feature analysis on the endoscopic images to determine smoke control parameters;

[0260] Step 720: Perform smoke image analysis on the endoscopic image to determine the smoke image parameters;

[0261] Step 730: Perform image parameter analysis on the endoscopic image to determine the image post-processing parameters.

[0262] This step involves three image analysis steps to perform multimodal recognition on the endoscopic images. Specifically, smoke feature analysis can be performed on the acquired endoscopic images to determine smoke removal control parameters. Smoke image analysis can also be performed on the endoscopic images to determine smoke image parameters. Furthermore, image parameter analysis can be performed on the endoscopic images to determine image post-processing parameters. Based on this, smoke removal control parameters can be obtained through smoke feature analysis, and the smoke removal intensity of the physical smoke removal equipment can be controlled according to these parameters. Taking the smoke concentration index as an example, the higher the smoke concentration index, the higher the smoke removal intensity of the smoke removal equipment; conversely, the lower the smoke concentration index, the lower the smoke removal intensity. Simultaneously, based on the smoke image parameters obtained from smoke image analysis and the image post-processing parameters obtained from image parameter analysis, digital smoke removal and adaptive optimization of the endoscopic images are performed.

[0263] In one embodiment, step 710 includes one of the following:

[0264] Smoke analysis algorithms were used to perform image feature analysis on endoscopic images to obtain smoke control parameters.

[0265] Alternatively, the endoscopic images can be input into a smoke feature analysis model for identification and processing to obtain smoke control parameters.

[0266] Based on this, smoke removal control parameters can be obtained directly by performing image feature analysis on endoscopic images using smoke analysis algorithms, or by performing recognition processing on endoscopic images using smoke feature analysis models. These smoke removal control parameters include, but are not limited to: smoke obstruction information indicating the presence of smoke, smoke concentration information indicating smoke concentration, switch control information for controlling the on / off state of the smoke removal equipment, and power control information for controlling the smoke removal power of the equipment.

[0267] In one embodiment, step 120 includes:

[0268] The endoscopic image is pre-processed, and the pre-processed endoscopic image is then detected to obtain smoke image parameters and post-processing parameters.

[0269] And / or, step 140 includes:

[0270] After optimizing the smoke-free image based on the image post-processing parameters, subsequent image processing is performed to obtain the target endoscope image.

[0271] In this step, pre-processing can be performed on the acquired endoscopic images. Pre-processing refers to a series of processing steps performed on the raw image before it is further analyzed or displayed. These steps aim to improve image quality, reduce noise, correct distortion, and prepare image data for subsequent processing. Pre-processing includes, but is not limited to, denoising, geometric correction, and bad pixel correction. Pre-processing plays a crucial role in improving image quality, ensuring that the image data provides accurate color and good visual effects after being transmitted to the central control module. For example, pre-processing can be pre-ISP image processing.

[0272] In this step, post-processing can also be performed on the smoke-removal image optimized by image optimization module 103. Post-processing refers to further processing steps performed on the image after initial analysis or processing. These steps aim to improve the visual effect of the image, enhance features, or prepare it for specific application purposes, such as output or display. Post-processing includes, but is not limited to, image enhancement, color correction, gamma correction, and other image processing. Post-processing steps ensure that the image can better reproduce scene details under different optical conditions, greatly improving image quality and visual effect. For example, post-processing can be post-ISP image processing.

[0273] like Figure 8 As shown in the embodiments of this application, another smoke removal method is also provided, which includes:

[0274] Step 810: Obtain endoscopic images captured by the endoscope.

[0275] In this step, the doctor or technician inserts an endoscope through the body's natural cavities or a small surgical incision to reach the area requiring examination or treatment. The endoscope's light source illuminates the internal cavity, allowing the image sensor to capture a clear image. The optical image captured by the image sensor (such as a CCD or CMOS) at the endoscope's tip is converted into an electrical signal. Located at the front of the camera system, the image sensor is responsible for collecting images and converting light signals into electrical signals, which are then transmitted to the camera system via the endoscope's circuitry to obtain the endoscopic image.

[0276] Step 820: Detect and process the endoscopic image to obtain image post-processing parameters.

[0277] In this step, the endoscopic image is processed to obtain image post-processing parameters. These parameters include, but are not limited to: brightness adjustment parameters, contrast adjustment parameters, sharpness adjustment parameters, color adjustment parameters, saturation adjustment parameters, hue adjustment parameters, local enhancement parameters, and noise reduction adjustment parameters.

[0278] Step 830: Perform digital desmearing on the endoscopic image to obtain a desmeared image.

[0279] In this step, digital desmoking of the endoscopic image can be achieved using AI technology, without relying on the aforementioned smoke image parameters, thus obtaining a smoke-free image. For example, digital desmoking of the endoscopic image can be performed based on a deep learning-based smoke detection model, an AI image signal processor, or an artificial intelligence algorithm, such as an image dehazing algorithm, to obtain a smoke-free image.

[0280] Step 840: Optimize the smoke-free image according to the image post-processing parameters to obtain the target endoscope image.

[0281] In this step, the de-smoked image obtained after digital de-smoking can be optimized based on image post-processing parameters. These parameters include, but are not limited to, ISP control parameters, which may include exposure time, analog gain, digital gain, white balance, and black level correction. ISP control parameters are used to control the sensor and lens to ensure image quality. Exposure time is used to adjust the image's brightness and contrast, ensuring appropriate exposure under different lighting conditions. Analog and digital gain are used to enhance image brightness, especially in low-light environments, increasing visibility by increasing gain. White balance is used to correct the image's color temperature, ensuring color consistency under different light sources. Black level correction is used to adjust shadow details, reducing the impact of dark current and ensuring clear shadow details. Optimizing the de-smoked image obtained after digital de-smoking using image post-processing parameters optimizes the de-smoked image processing workflow, improving the saturation, contrast, sharpness, and color accuracy of the de-smoked image, thereby improving the overall quality and performance of the de-smoked image.

[0282] Step 850: Control the display to show the target endoscope image.

[0283] In this step, a monitor can be used to display the target endoscopic image. The monitor can be a touch screen or LCD screen installed in the smoke removal system, or it can be a standalone display device such as an LCD screen or television set independent of the smoke removal system, or it can be a display screen on an electronic device such as a mobile phone or tablet.

[0284] In one embodiment, step 820 includes:

[0285] Endoscopic images are detected and processed to obtain smoke removal control parameters and image post-processing parameters;

[0286] Smoke removal methods also include:

[0287] The smoke removal equipment is controlled according to the smoke removal control parameters to remove smoke.

[0288] In this step, the physical smoke removal equipment can be controlled to physically remove smoke based on smoke removal control parameters, and the smoke-removed image can be optimized based on image post-processing parameters to obtain and display the target endoscope image. Based on this, multimodal information, including smoke removal control parameters and image post-processing parameters, is obtained by detecting the endoscope image. Based on this, on the one hand, the smoke concentration information is calculated, and the smoke removal control parameters are output to control the smoke removal intensity of the physical smoke removal equipment; on the other hand, adaptive optimization of the endoscope image is performed based on the image post-processing parameters.

[0289] It should be noted that the smoke control parameters include, but are not limited to: smoke obstruction information indicating the presence of smoke obstruction, smoke concentration information indicating smoke concentration, switch control information for controlling the switching on and off of the smoke removal equipment, and power control information for controlling the smoke removal power of the smoke removal equipment.

[0290] For physical smoke removal, for example, the smoke concentration information, such as the smoke concentration index, can be calculated, and the smoke removal control parameters can be output based on the smoke concentration index to control the smoke removal intensity of the physical smoke removal device 101. For example, the higher the smoke concentration index, the higher the smoke removal intensity of the smoke removal device 101; conversely, the lower the smoke concentration index, the lower the smoke removal intensity of the smoke removal device 101.

[0291] In one embodiment, step 830 includes:

[0292] The endoscopic image is input into the smoke removal model for smoke removal processing to obtain the smoke-removed image;

[0293] Alternatively, dark channel features can be extracted from the endoscopic image, including atmospheric brightness and / or transmittance. The endoscopic image can then be processed based on the transmittance and atmospheric brightness to obtain a smoke-free image.

[0294] In one embodiment, this application embodiment utilizes AI methods to remove smoke from endoscopic images during image processing. Specifically, a smoke removal model can be used to remove smoke from endoscopic images. The central control module 104 inputs endoscopic image and smoke image parameters into the smoke removal model for processing, resulting in a smoke-free image. The smoke image parameters may include the distribution information of smoke in the endoscopic image and dark channel features. The smoke removal model may include a YOLO-based smoke detection model, a U-Net-based smoke purification algorithm, and other deep neural network-based endoscopic image smoke purification models. The smoke removal model uses different mechanisms and algorithms, such as dark channel priors, attention mechanisms, and adversarial learning, to process endoscopic images for smoke removal, improving the clarity and visibility of the endoscopic images, thereby providing doctors with a more accurate surgical field of view.

[0295] In one embodiment, this application embodiment utilizes a dark channel method to remove smoke from endoscopic images during image processing. This involves processing the endoscopic image based on dark channel features and atmospheric brightness values ​​to obtain a smoke-free image, thereby removing smoke from the endoscopic image and improving its clarity and visibility. Dark channel features in an endoscopic image refer to the presence of a maximum intensity value for at least one color channel (red, green, or blue) within a local region of the image. This local maximum intensity value can be used as an estimate of atmospheric light. In the dark channel prior dehazing algorithm, dark channel features are used to estimate the fog density in the image, thereby restoring image clarity. In the endoscopic image, atmospheric light can be estimated by searching for local maximum intensity values, determining the atmospheric brightness value of the endoscopic image, and thus constructing a dark channel image. Atmospheric brightness value is the light intensity value in the image due to atmospheric scattering, which can be estimated by finding the brightest point in the image. Specifically, transmittance can be obtained from the dark channel features; transmittance describes the degree to which light is attenuated by smoke particles during propagation. The image is reconstructed using estimated transmittance and atmospheric brightness values, thus obtaining a smoke-free image. These steps effectively remove smoke from endoscopic images, improving image quality and providing surgeons with a clearer surgical field of vision. This method not only enhances surgical safety but also improves image processing performance in computer-assisted surgery.

[0296] This application also provides an endoscope system, including:

[0297] An endoscope is used to capture endoscopic images of a target object.

[0298] A monitor used to display endoscopic images;

[0299] Smoke removal equipment is used to remove smoke from target objects;

[0300] An endoscope host is connected to an endoscope. The endoscope host is used to acquire endoscopic image data captured by the endoscope and to control the smoke removal device to remove smoke. The endoscope host is also used to execute the smoke removal method provided in any of the above embodiments to obtain a target endoscopic image and display it on a display.

[0301] This application also provides an endoscope system, including:

[0302] An endoscope is used to capture endoscopic images of a target object.

[0303] A monitor used to display endoscopic images;

[0304] An endoscope host is connected to an endoscope and is used to acquire endoscopic image data captured by the endoscope. The endoscope host is also used to execute the smoke removal method for endoscopic images provided in any of the above embodiments to obtain a target endoscopic image, and to display it on a monitor.

[0305] This application provides a computer storage medium storing a computer program applied to an ultrasonic imaging device. When the computer program is executed by a processor, it implements the smoke removal method as described in the above embodiment.

[0306] This application provides a computer program product or computer program that includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the smoke removal method as described in the above embodiments.

[0307] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or indirect coupling or communication connection between apparatuses or units, and may be electrical, mechanical, or other forms.

[0308] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0309] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0310] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0311] It should also be understood that the various implementation methods provided in this application can be combined arbitrarily to achieve different technical effects.

[0312] The above provides a detailed description of the preferred embodiments of this application. However, this application is not limited to the above-described embodiments. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of this application. All such equivalent modifications or substitutions are included within the scope defined by the claims of this application.

Claims

1. A smoke removal system, characterized in that, include: The acquisition module is used to acquire endoscopic images of the target object obtained by the endoscope. The central control module is used to detect and process the endoscopic images to obtain multimodal information, including smoke removal control parameters, smoke image parameters, and image post-processing parameters. A physical smoke removal device is used to physically remove smoke from the target object; An image desmoke module is used to digitally desmoke the endoscopic images; The image optimization module is used to optimize the smoke-free image obtained after digital smoke removal. The central control module is also used to control the physical smoke removal device to perform physical smoke removal according to the smoke removal control parameters, and to control the image smoke removal module to perform digital smoke removal on the endoscope image and obtain a smoke-removed image according to the smoke image parameters. The central control module is also used to optimize the smoke-removed image according to the image post-processing parameters to obtain a target endoscope image, and to control the display to show the target endoscope image.

2. The smoke removal system according to claim 1, characterized in that, The central control module includes a multimodal analysis model, and the central control module is also used to input the endoscopic image into the multimodal analysis model for detection and processing to obtain the multimodal information.

3. The smoke removal system according to claim 1, characterized in that, The central control module is used to detect and process the endoscopic images, including: The central control module is used to perform smoke feature analysis on the endoscopic images and determine the smoke removal control parameters; The central control module is also used to perform smoke image analysis on the endoscopic image to determine the smoke image parameters; The central control module is also used to perform image parameter analysis on the endoscopic image and determine the image post-processing parameters.

4. The smoke removal system according to claim 3, characterized in that, The step of performing smoke feature analysis on the endoscopic image to determine the smoke removal control parameters includes: The smoke removal control parameters are obtained by performing image feature analysis on the endoscopic image using a smoke analysis algorithm, or by inputting the endoscopic image into a smoke feature analysis model for recognition processing. The step of performing smoke image analysis on the endoscopic image to determine the smoke image parameters includes: The smoke image parameters are obtained by performing image analysis on the endoscopic image using a smoke image recognition algorithm, or by inputting the endoscopic image into a smoke image recognition model for recognition processing. The step of performing image parameter analysis on the endoscopic image to determine the image post-processing parameters includes: The endoscope image is analyzed using an image parameter analysis algorithm to obtain the image post-processing parameters; alternatively, the endoscope image is input into an image parameter analysis model for recognition processing to obtain the image post-processing parameters.

5. The smoke removal system according to any one of claims 1 to 4, characterized in that, The smoke removal control parameters include at least one of the following: Smoke masking information indicating whether smoke masking exists; Smoke concentration information characterizing smoke concentration; Switch control information used to control the switch of the smoke removal equipment; Power control information used to control the smoke removal power of the smoke removal equipment.

6. The smoke removal system according to any one of claims 1 to 4, characterized in that, The smoke image parameters include at least one of the following: The distribution information of smoke in the endoscopic image, wherein the distribution information includes at least location information or concentration information; and / or... Dark channel features in the endoscopic image.

7. The smoke removal system according to any one of claims 1 to 4, characterized in that, The smoke image parameters include dark channel features. The central control module controls the image desmoke module to digitally desmoke the endoscopic image based on the smoke image parameters and obtain a desmoke-free image, including: The dark channel features include atmospheric brightness values ​​and / or transmittance of the endoscopic image; The image desmoke module processes the endoscopic image based on the dark channel features to obtain the desmoke-free image.

8. The smoke removal system according to any one of claims 1 to 4, characterized in that, The image desmoke module includes a desmoke model. The central control module controls the image desmoke module to digitally desmoke the endoscopic image based on the smoke image parameters and obtain a desmoke-free image, including: The central control module inputs the endoscopic image and the smoke image parameters into the smoke removal model for smoke removal processing to obtain the smoke-removed image.

9. The smoke removal system according to any one of claims 1 to 4, characterized in that, The image post-processing parameters include at least one of the following: Brightness adjustment parameters; Contrast adjustment parameters; Sharpness adjustment parameters; Color adjustment parameters; Saturation adjustment parameters; Tone adjustment parameters; Regional local enhancement parameters; Adjust the noise reduction parameters.

10. The smoke removal system according to claim 1, characterized in that, The multimodal information also includes endoscopic device control parameters, which include at least one of the following: Endoscopic light source device control parameters; Endoscopic camera control parameters.

11. The smoke removal system according to claim 10, characterized in that, The central control module is also used to control the endoscope according to the endoscope control parameters, so as to optimize the endoscope images captured by the endoscope.

12. The smoke removal system according to any one of claims 1 to 4, characterized in that, The smoke removal system further includes a pre-processing image module and a post-processing image module. The central control module is also used to control the pre-processing image module to perform pre-processing on the endoscopic images obtained by the acquisition module. The central control module is used to perform detection processing on the endoscopic images after the pre-processing to obtain multimodal information. The central control module is also used to control the image optimization module to optimize the smoke removal images according to the image post-processing parameters, and then perform post-processing on the images by the post-processing image module to obtain the target endoscopic image.

13. A method for removing smoke, characterized in that, include: Acquire endoscopic images captured by an endoscope; The endoscopic image is processed to obtain smoke image parameters and image post-processing parameters; The endoscopic image is digitally desmoke-removed based on the smoke image parameters to obtain a smoke-removed image. The smoke-free image is optimized according to the image post-processing parameters to obtain the target endoscope image; The control display shows the target endoscopic image.

14. The smoke removal method according to claim 13, characterized in that, The detection and processing of the endoscopic image to obtain smoke image parameters and image post-processing parameters includes: The endoscopic image is input into a multimodal analysis model for detection and processing to obtain the smoke image parameters and the image post-processing parameters.

15. The smoke removal method according to claim 13, characterized in that, The detection and processing of the endoscopic image to obtain smoke image parameters and image post-processing parameters includes: Smoke image analysis is performed on the endoscopic images to determine the smoke image parameters; Image parameter analysis is performed on the endoscopic images to determine the image post-processing parameters.

16. The smoke removal method according to claim 15, characterized in that, The step of performing smoke image analysis on the endoscopic image to determine the smoke image parameters includes one of the following: The smoke image parameters are obtained by analyzing the endoscopic image using a smoke image recognition algorithm. Alternatively, the endoscopic image can be input into a smoke image recognition model for recognition processing to obtain the smoke image parameters.

17. The smoke removal method according to any one of claims 13 to 16, characterized in that, The smoke image parameters include at least one of the following: The distribution information of smoke in the endoscopic image, wherein the distribution information includes at least location information or concentration information; Dark channel features in the endoscopic image.

18. The smoke removal method according to claim 15, characterized in that, The step of performing image parameter analysis on the endoscopic image to determine the image post-processing parameters includes one of the following: The endoscopic image is analyzed using an image parameter analysis algorithm to obtain the image post-processing parameters. Alternatively, the endoscopic image can be input into an image parameter analysis model for recognition processing to obtain the image post-processing parameters.

19. The smoke removal method according to any one of claims 13 to 15, 18, characterized in that, The image post-processing parameters include at least one of the following: Brightness adjustment parameters; Contrast adjustment parameters; Sharpness adjustment parameters; Color adjustment parameters; Saturation adjustment parameters; Tone adjustment parameters; Regional local enhancement parameters; Adjust the noise reduction parameters.

20. The smoke removal method according to claim 13, characterized in that, The method further includes: The endoscopic images are processed to obtain the control parameters of the endoscopic device; The endoscope is controlled according to the endoscope control parameters to optimize the endoscope images captured by the endoscope. The control parameters of the endoscopic device include at least one of the following: Endoscopic light source device control parameters; Endoscopic camera control parameters.

21. The smoke removal method according to any one of claims 13 to 16, characterized in that, The smoke image parameters include dark channel features in the endoscopic image; the step of digitally desmoke-removing the endoscopic image based on the smoke image parameters to obtain a desmoke-removed image includes: Obtain the atmospheric brightness value and transmittance of the endoscope image included in the dark channel features; The endoscopic image is processed based on the atmospheric brightness value and the transmittance to obtain the smoke-free image.

22. The smoke removal method according to any one of claims 13 to 16, characterized in that, The step of digitally removing smoke from the endoscopic image based on the smoke image parameters to obtain a smoke-free image includes: The parameters of the endoscope image and the smoke image are input into the smoke removal model for smoke removal processing to obtain the smoke-removed image.

23. The smoke removal method according to any one of claims 13 to 16, characterized in that, The detection and processing of the endoscopic image to obtain smoke image parameters and image post-processing parameters includes: The endoscopic images are processed to obtain smoke control parameters, smoke image parameters, and image post-processing parameters. The smoke removal method also includes: The smoke removal equipment is controlled to remove smoke according to the smoke removal control parameters.

24. The smoke removal method according to claim 23, characterized in that, The detection and processing of the endoscopic image to obtain smoke control parameters, smoke image parameters, and image post-processing parameters includes: The endoscopic image is input into a multimodal analysis model for detection and processing to obtain the smoke removal control parameters, the smoke image parameters, and the image post-processing parameters.

25. The smoke removal method according to claim 23, characterized in that, The detection and processing of the endoscopic image to obtain smoke control parameters, smoke image parameters, and image post-processing parameters includes: Smoke feature analysis is performed on the endoscopic images to determine the smoke removal control parameters; Smoke image analysis is performed on the endoscopic images to determine the smoke image parameters; Image parameter analysis is performed on the endoscopic images to determine the image post-processing parameters.

26. The smoke removal method according to claim 25, characterized in that, The step of performing smoke feature analysis on the endoscopic image to determine the smoke removal control parameters includes one of the following: The smoke removal control parameters are obtained by performing image feature analysis on the endoscopic image using a smoke analysis algorithm. Alternatively, the endoscopic image can be input into a smoke feature analysis model for identification and processing to obtain the smoke removal control parameters.

27. The smoke removal method according to any one of claims 23 to 25, characterized in that, The smoke removal control parameters include at least one of the following: Smoke masking information indicating whether smoke masking exists; Smoke concentration information characterizing smoke concentration; Switch control information for controlling the switch of the smoke removal equipment. Power control information used to control the smoke removal power of the smoke removal equipment.

28. The smoke removal method according to claim 13, characterized in that, The detection and processing of the endoscopic image to obtain smoke image parameters and image post-processing parameters includes: The endoscopic image is subjected to pre-processing; the endoscopic image after pre-processing is then subjected to detection processing to obtain the smoke image parameters and the image post-processing parameters. And / or, the step of optimizing the smoke-free image according to the image post-processing parameters to obtain the target endoscopic image includes: After optimizing the smoke-free image according to the image post-processing parameters, the target endoscope image is obtained through subsequent image processing.

29. An endoscope system, characterized in that, include: An endoscope is used to capture endoscopic images of a target object. A monitor used to display endoscopic images; Smoke removal equipment, used to remove smoke from the target object; An endoscope host is connected to the endoscope. The endoscope host is used to acquire endoscopic image data captured by the endoscope and to control the smoke removal device to remove smoke. The endoscope host is also used to execute the smoke removal method according to any one of claims 13 to 28 to obtain a target endoscopic image and display it through the display.

Images