Bleeding Point Detection Method, Computer Device, Storage Medium, and Program Product

Abstract

The present application relates to a bleeding point detection method, a computer device, a storage medium, and a program product. The method includes: obtaining an image to be detected, determining a conversion relationship between the image to be detected and a sample image based on the image to be detected, and then converting the bleeding point position in the sample image to the image to be detected according to the conversion relationship to obtain the bleeding point position in the image to be detected. The sample image is an image including a bleeding point in a historical sequence of images of a target endoscope. Through the above method, the detection of the bleeding point position can be achieved. By utilizing the immobility of the bleeding point, the bleeding point position in the image to be detected can be obtained based on the bleeding point position in the sample image and the above conversion relationship, thereby improving the accuracy of the bleeding point position detection.

Description

Technical Field

[0001] This application relates to the field of medical technologies, and particularly to a method for detecting bleeding points, a computer device, a storage medium, and a program product. Background Art

[0002] An endoscope is a common medical device in the medical field, and is widely used in scenarios such as observing, suturing, and excising internal tissues of a target object.

[0003] When applied to excising internal tissues of a target object, a bleeding point will be formed at the excision position. After the tissue excision is completed, medical staff need to find the bleeding point for suturing. However, in actual applications, due to the excessive and rapid bleeding volume at the bleeding point, a large amount of blood will accumulate, covering the bleeding point, making it impossible for medical staff to find the bleeding point in time.

[0004] In traditional technologies, an image of internal tissues can be obtained through an image sensor at the end of an endoscope body, and a computer device performs image processing (ISP, Image Signal Processor) on the image of the internal tissues to enhance the subtle difference between the bleeding point position and the blood area in the image of the internal tissues, and then determines the bleeding point position. However, the accuracy of the bleeding point position determined by the above traditional method is relatively poor. Summary of the Invention

[0005] Based on this, it is necessary to provide a method for detecting bleeding points, a computer device, a storage medium, and a program product for the above technical problems.

[0006] In a first aspect, a method for detecting bleeding points is provided, including:

[0007] Obtain an image to be detected;

[0008] Based on the image to be detected, determine the conversion relationship between the image to be detected and a sample image; wherein, the sample image is an image including a bleeding point in a historical sequence of images of a target endoscope;

[0009] According to the conversion relationship, convert the bleeding point position in the sample image to the image to be detected to obtain the bleeding point position in the image to be detected.

[0010] In one embodiment, based on the image to be detected, determining the conversion relationship between the image to be detected and the sample image includes:

[0011] Obtain the position information of pairs of matching feature points between the image to be detected and the sample image;

[0012] Determine the affine transformation matrix between the image to be detected and the sample image according to the position information of the pairs of matching feature points;

[0013] Determine the affine transformation matrix as the conversion relationship between the image to be detected and the sample image.

[0014] In one embodiment, obtaining the position information of the matching feature point pairs between the image to be detected and the sample image includes:

[0015] Perform feature extraction on the image to be detected to obtain the multi-dimensional features of at least one first feature point in the image to be detected;

[0016] Perform feature matching between the multi-dimensional features of at least one first feature point and the multi-dimensional features of at least one second feature point in the sample image, and determine the first feature point and the second feature point with the highest feature matching degree as the matching feature point pair;

[0017] Obtain the position information of the first feature point in the matching feature point pair in the image to be detected, and the position information of the second feature point in the matching feature point pair in the sample image.

[0018] In one embodiment, obtaining the image to be detected includes:

[0019] Obtain the initial image to be detected, and determine the similarity between the initial image to be detected and the sample image;

[0020] If the similarity is greater than the similarity threshold, determine the initial image to be detected as the image to be detected.

[0021] In one embodiment, the above method further includes:

[0022] Obtain the initial images under different light source directions of the target endoscope;

[0023] Perform synthesis processing on the initial images under different light source directions to obtain the stereoscopic image at the target moment;

[0024] Arrange the stereoscopic images at the target moment in chronological order to obtain the historical sequence images.

[0025] In one embodiment, obtaining the initial images under different light source directions of the target endoscope includes:

[0026] Change the light source direction by adjusting the light-shielding position of the light-shielding sheet on the target endoscope; wherein, the light-shielding sheet is located at the end of the mirror body of the target endoscope and covers part of the light source;

[0027] Obtain the images of the operating environment of the target endoscope when the light-shielding sheet is at different light-shielding positions as the initial images.

[0028] In one embodiment, the above method further includes:

[0029] Perform differential processing on two adjacent stereo images in a historical sequence of images to obtain the difference features between the two adjacent stereo images in the historical sequence of images;

[0030] Determine the stereo image in which the bleeding point first appears according to the difference features, and use the stereo image in which the bleeding point first appears as the sample image.

[0031] In a second aspect, the present application also provides a bleeding point detection device, including:

[0032] An image acquisition module for acquiring an image to be detected; a conversion determination module for determining the conversion relationship between the image to be detected and the sample image based on the image to be detected; wherein, the sample image is an image including a bleeding point in a historical sequence of images of a target endoscope;

[0033] A position conversion module for converting the position of the bleeding point in the sample image to the image to be detected according to the conversion relationship to obtain the position of the bleeding point in the image to be detected.

[0034] In a third aspect, the present application also provides a computer device, including a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0035] Acquire an image to be detected; determine the conversion relationship between the image to be detected and the sample image based on the image to be detected; wherein, the sample image is an image including a bleeding point in a historical sequence of images of a target endoscope;

[0036] Convert the position of the bleeding point in the sample image to the image to be detected according to the conversion relationship to obtain the position of the bleeding point in the image to be detected.

[0037] In a fourth aspect, the present application also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0038] Acquire an image to be detected; determine the conversion relationship between the image to be detected and the sample image based on the image to be detected; wherein, the sample image is an image including a bleeding point in a historical sequence of images of a target endoscope;

[0039] Convert the position of the bleeding point in the sample image to the image to be detected according to the conversion relationship to obtain the position of the bleeding point in the image to be detected.

[0040] In a fifth aspect, the present application also provides a computer program product, including a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0041] Obtain the image to be detected; based on the image to be detected, determine the conversion relationship between the image to be detected and the sample image; wherein, the sample image is an image including a bleeding point in the historical sequence images of the target endoscope;

[0042] According to the conversion relationship, convert the position of the bleeding point in the sample image to the image to be detected, and obtain the position of the bleeding point in the image to be detected.

[0043] The above bleeding point detection method, device, computer device, storage medium and computer program product obtain the image to be detected, determine the conversion relationship between the image to be detected and the sample image based on the image to be detected, and then convert the position of the bleeding point in the sample image to the image to be detected according to the conversion relationship, so as to obtain the position of the bleeding point in the image to be detected. Among them, the sample image is an image including a bleeding point in the historical sequence images of the target endoscope. Through the above method, the detection of the bleeding point position can be realized. By utilizing the immobility of the bleeding point, the position of the bleeding point in the image to be detected can be obtained based on the position of the bleeding point in the sample image and the above conversion relationship, thereby improving the accuracy of the bleeding point position detection. Description of the Drawings

[0044] Figure 1 It is an application environment diagram of the bleeding point detection method in an embodiment;

[0045] Figure 2 It is a flowchart of the bleeding point detection method in an embodiment;

[0046] Figure 3 It is a flowchart of determining the conversion relationship between the image to be detected and the sample image in an embodiment;

[0047] Figure 4 It is a flowchart of obtaining the position information of the matching feature point pairs in an embodiment;

[0048] Figure 5 It is a flowchart of obtaining the image to be detected in an embodiment;

[0049] Figure 6 It is a flowchart of constructing the historical sequence images in an embodiment;

[0050] Figure 7 It is a flowchart of the initial images under different light source directions in an embodiment;

[0051] Figure 8 It is a flowchart of the end portion of the body of the target endoscope in an embodiment;

[0052] Figure 9 It is a structural diagram of the light shield in an embodiment;

[0053] Figure 10 Schematic diagram of the process for determining a sample image in one embodiment;

[0054] Figure 11 Schematic diagram of the images included in the historical sequence images in one embodiment;

[0055] Figure 12 Schematic diagram of the process of a bleeding point detection method in another embodiment;

[0056] Figure 13 Schematic diagram of the illustration processing process of a bleeding point detection method in one embodiment;

[0057] Figure 14 Structural block diagram of a bleeding point detection device in one embodiment;

[0058] Figure 15 Internal structure diagram of a computer device in one embodiment. Detailed implementation manners

[0059] In order to make the objectives, technical solutions and advantages of the present application clearer and more understandable, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0060] The bleeding point detection method provided by the embodiments of the present application can be applied to an application environment as Figure 1 shown. Among them, the endoscope system includes a data acquisition end and a data processing end. The data acquisition end includes a target endoscope 102, and the data processing end includes a computer device 104 for implementing bleeding point detection. Communication is carried out between the target endoscope 102 and the computer device 104. The target endoscope 102 is used to collect the current image of the operation environment. The computer device 104 obtains the current image through the target endoscope 102 and uses it as the image to be detected. Based on the image to be detected, the conversion relationship between the image to be detected and the sample image is determined. Furthermore, according to this conversion relationship, the bleeding point position in the sample image is converted to the image to be detected to obtain the bleeding point position in the image to be detected. Among them, the sample image is an image including a bleeding point in the historical sequence images of the target endoscope. Among them, the above-mentioned computer device 104 can be a general computer device or a dedicated computer device, such as a desktop computer, a portable computer, a network server, a personal digital assistant (PDA), a mobile phone, a tablet computer, a wireless terminal device, a communication device, an embedded device, etc. The specific type of the computer device 104 is not limited in this embodiment.

[0061] In one embodiment, as Figure 2As shown, a bleeding point detection method is provided. Taking the computer device to which this method is applied as an example in Figure 1 for illustration, the method includes the following steps:

[0062] S210. Obtain the image to be detected.

[0063] Among them, the image to be detected includes the image of the tissue / inside the cavity collected by the target endoscope. The target endoscope is used to collect the image of its operating environment (for example, the inside of the simulated human tissue / cavity).

[0064] Optionally, the target endoscope and the computer device can communicate in a wired or wireless manner, and the computer device can obtain the image to be detected collected by the target endoscope. Among them, the image to be detected can be an image obtained by photometric stereo synthesis of the images collected by the target endoscope under different light source directions.

[0065] S220. Based on the image to be detected, determine the conversion relationship between the image to be detected and the sample image.

[0066] Among them, the sample image is the image including the bleeding point in the historical sequence of images of the target endoscope. The historical sequence of images of the target endoscope is the image collected by the target endoscope in the same operating environment before obtaining the image to be detected. The conversion relationship between the image to be detected and the sample image essentially reflects the coordinate conversion relationship between the image to be detected and the sample image.

[0067] Optionally, the computer device can compare the features of the image to be detected and the sample image to determine the conversion relationship between the image to be detected and the sample image.

[0068] S230. According to the conversion relationship, convert the bleeding point position in the sample image to the image to be detected to obtain the bleeding point position in the image to be detected.

[0069] Specifically, the computer device obtains the bleeding point position in the sample image, and then converts the bleeding point position in the sample image to the image to be detected according to the above conversion relationship to obtain the bleeding point position in the image to be detected.

[0070] In this embodiment, the computer device acquires the image to be detected, determines the conversion relationship between the image to be detected and the sample image based on the image to be detected, and then converts the position of the bleeding point in the sample image to the image to be detected according to the conversion relationship to obtain the position of the bleeding point in the image to be detected. Wherein, the image to be detected is the current image acquired by the target endoscope, and the sample image is the image including the bleeding point in the historical sequence of images of the target endoscope. Through the above method, the detection of the position of the bleeding point can be realized. Utilizing the immobility of the bleeding point, the position of the bleeding point in the image to be detected can be obtained based on the position of the bleeding point in the sample image and the above conversion relationship, thereby improving the accuracy of the bleeding point position detection.

[0071] In practical applications, the conversion relationship between the image to be detected and the sample image can be determined according to the matching feature point pairs between the image to be detected and the sample image. For example Figure 3 As described in, S220. Based on the image to be detected, determine the conversion relationship between the image to be detected and the sample image, then includes:

[0072] S310. Obtain the position information of the matching feature point pairs between the image to be detected and the sample image.

[0073] Wherein, the feature is a specific structure in the image, the feature point is a point or area with features, and in a specific embodiment, the feature point can be taken as an example of a point with features. The matching feature point pairs are a pair of feature points with feature matching. The above position information correspondingly includes the position information of the feature points located in the image to be detected in the image to be detected in the matching feature point pairs, and the position information of the feature points located in the sample image in the sample image in the matching feature point pairs.

[0074] Optionally, the computer device extracts feature points from the image to be detected to obtain each feature point in the image to be detected, and performs feature matching between each obtained feature point and each feature point in the sample image to obtain the matching feature point pairs between the image to be detected and the sample image, and then respectively obtains the position information of the feature points located in the image to be detected in the matching feature point pairs in the image to be detected, and the position information of the feature points located in the sample image in the sample image in the matching feature point pairs.

[0075] S320. Determine the affine transformation matrix between the image to be detected and the sample image according to the position information of the matching feature point pairs.

[0076] It should be noted that during the actual use of the target endoscope, there may be transformation operations such as movement, rotation, or scaling. It is difficult to ensure that the target endoscope is in the same operating state when acquiring the sample image and the image to be detected. The above affine transformation matrix can accurately reflect the correlation between the image to be detected and the sample image.

[0077] Optionally, the position information of the above-mentioned matched feature point pairs may be the coordinates of the feature points in the image. The computer device can obtain the affine transformation matrix between the image to be detected and the sample image based on the coordinates of multiple groups of feature point pairs in the image to be detected and the coordinates in the sample image.

[0078] S330. Determine the affine transformation matrix as the conversion relationship between the image to be detected and the sample image.

[0079] Specifically, after the computer device obtains the above-mentioned affine transformation matrix, it immediately determines this affine transformation matrix as the conversion relationship between the image to be detected and the sample image.

[0080] In this embodiment, the computer device obtains the position information of the matched feature point pairs between the image to be detected and the sample image, determines the affine transformation matrix between the image to be detected and the sample image according to the position information of the matched feature point pairs, and then determines the affine transformation matrix as the conversion relationship between the image to be detected and the sample image. The above-mentioned affine transformation matrix can accurately reflect the conversion relationship between the image to be detected and the sample image, which helps to accurately determine the position of the bleeding point in the image to be detected based on this conversion relationship, thereby improving the accuracy of bleeding point detection.

[0081] In an optional embodiment, the accuracy of the affine transformation matrix can be improved by increasing the number of matched feature point pairs between the image to be detected and the sample image, and then the accuracy of bleeding point detection can be improved. Based on this, as Figure 4 shown, the above-mentioned S310. Obtain the position information of the matched feature point pairs between the image to be detected and the sample image, then includes:

[0082] S410. Perform feature extraction on the image to be detected to obtain the multi-dimensional features of at least one first feature point in the image to be detected.

[0083] Among them, the multi-dimensional features are various types of feature information. Optionally, the multi-dimensional features include various types of feature information such as color, texture, size, and position information. Among them, the position information includes the distance of the feature point relative to a certain reference point, and this reference point can be another feature point or the center point of the image.

[0084] It should be noted that the feature points can be relatively prominent points in the image, such as contour points, bright points in darker areas, dark points in brighter areas, points with color differences from the surrounding areas, etc. The feature points in different images are different, and the quantity is also uncertain.

[0085] Optionally, the computer device may process the image to be detected using an image processing method based on the characteristics of the above-mentioned feature points, obtain the feature points in the image to be detected, i.e., the above-mentioned first feature points, so as to further obtain the multi-dimensional information of each first feature point. For example, the computer device may use the ORB (Oriented FAST and Rotated BRIEF) feature extraction algorithm to obtain the first feature points in the image to be detected. The computer device may also input the image to be detected into a deep learning neural network model for identifying feature points, so as to identify the first feature points in the image to be detected through this neural network model.

[0086] S420. Perform feature matching between the multi-dimensional features of at least one first feature point and the multi-dimensional features of at least one second feature point in the sample image, and determine the first feature point and the second feature point with the highest feature matching degree as the matching feature point pair.

[0087] Among them, the feature points in the sample image are the above-mentioned second feature points, and moreover, the feature types of the multi-dimensional features of the second feature points in the sample image match the feature types of the multi-dimensional features of the first feature points in the image to be detected.

[0088] Optionally, for the extraction process of the second feature points in the sample image, reference may be made to the extraction process of the first feature points in the above-mentioned image to be detected. For example, the ORB feature extraction algorithm or a deep learning neural network model may be used to obtain the second feature points in the sample image. Similarly, the first feature points in different images to be detected are different, and the second feature points in different sample images are different. However, the image to be detected and the sample image are images obtained by the target endoscope in the same operation scenario, and there are cases where there are matching first feature points and second feature points, i.e., feature point pairs, between the two.

[0089] In a specific embodiment, the extraction process of the second feature points in the sample image is a preprocessing process based on the sample image. That is, before the step of obtaining the image to be detected, the sample image is preprocessed for feature extraction to obtain the multi-dimensional features of at least one second feature point in the sample image.

[0090] Optionally, the feature matching degree is positively correlated with the number of matching feature types, that is, the more the number of matching feature types, the higher the feature matching degree; conversely, the fewer the number of matching feature types, the lower the feature matching degree. The feature matching degree is also negatively correlated with the feature deviation between various types of features, that is, the greater the feature deviation, the lower the feature matching degree; conversely, the smaller the feature deviation, the higher the feature matching degree.

[0091] Specifically, the computer device performs feature matching of multi-dimensional features between each first feature point in the image to be detected and each second feature point in the sample image to determine the matching degree between the corresponding first feature point and the second feature point. For each first feature point, the first feature point and the second feature point with the highest feature matching degree and greater than the matching degree threshold are determined as the matching feature point pair.

[0092] S430. Obtain the position information of the first feature point in the matching feature point pair in the image to be detected, and the position information of the second feature point in the matching feature point pair in the sample image.

[0093] Specifically, after the computer device determines the matching feature point pair between the image to be detected and the sample image, it obtains the coordinates of the first feature point in the matching feature point pair in the image to be detected, and the coordinates of the second feature point in the matching feature point pair in the sample image.

[0094] In this embodiment, the computer device extracts features from the image to be detected to obtain the multi-dimensional features of at least one first feature point in the image to be detected, performs feature matching on the multi-dimensional features of at least one first feature point and the multi-dimensional features of at least one second feature point in the sample image, and determines the first feature point and the second feature point with the highest feature matching degree as the matching feature point pair. Furthermore, it obtains the position information of the first feature point in the matching feature point pair in the image to be detected, and the position information of the second feature point in the matching feature point pair in the sample image. Through the above multi-dimensional feature matching method, the matching feature point pair can be accurately determined, the credibility of the determined matching feature point pair is improved, and based on the position information of the highly accurate and highly credible matching feature point pair, an affine transformation matrix with high accuracy and high credibility can be determined, thereby improving the accuracy of bleeding point detection.

[0095] In one of the embodiments, to further improve the accuracy of the determined bleeding point position, as Figure 5 shown, the above S210. Obtain the image to be detected, includes:

[0096] S510. Obtain the initial image to be detected, and determine the similarity between the initial image to be detected and the sample image.

[0097] Among them, the above initial image to be detected is an image of the operation environment obtained by the target endoscope when the user performs bleeding point detection, and can be an image collected at any acquisition position of the target endoscope during the entire operation process. In the same operation scenario, different acquisition positions correspond to different operation environments.

[0098] Optionally, the computer device can directly calculate the structural similarity measure (SSIM) between the initial image to be detected and the sample image, or can respectively represent the initial image to be detected and the sample image as vectors, and then calculate the cosine value of the vectors between the initial image to be detected and the sample image. It can also perform histogram matching on the initial image to be detected and the sample image to obtain the similarity between the initial image to be detected and the sample image.

[0099] Optionally, the computer device can also use a large number of reference images at the same acquisition position as the sample image as training samples to train a neural network model for identifying the operating environment at the corresponding acquisition position, and then input the above initial image to be detected into the neural network model, and the neural network model outputs the similarity between the initial image to be detected and the sample image.

[0100] S520. If the similarity is greater than the similarity threshold, determine the initial image to be detected as the image to be detected.

[0101] Specifically, after the computer device obtains the similarity between the initial image to be detected and the sample image, it compares the similarity with a preset similarity threshold, and then determines whether to determine the initial image to be detected as the image to be detected according to the comparison result.

[0102] Among them, if the similarity is greater than the similarity threshold, determine the initial image to be detected as the image to be detected; conversely, if the similarity is less than or equal to the similarity threshold, determine that the initial image to be detected is not the image to be detected, and the user can also be further reminded to re-obtain the initial image to be detected.

[0103] In this embodiment, the computer device obtains the initial image to be detected and determines the similarity between the initial image to be detected and the sample image. If the similarity is greater than the similarity threshold, the initial image to be detected is determined as the image to be detected. Through the above method, the screening of the image to be detected can be realized, so that the initial image to be detected collected near the bleeding point is used as the image to be detected for bleeding point detection, which improves the accuracy of the determined bleeding point position while avoiding unnecessary detection, reducing the detection time, and correspondingly improving the detection efficiency.

[0104] In one of the embodiments, the above method further includes the process of constructing a historical sequence of images. As Figure 6 shown, the above method further includes:

[0105] S610. Obtain the initial images under different light source directions of the target endoscope.

[0106] Optionally, an adjustment component is provided at the end of the body of the target endoscope, and the adjustment component is used to change the light-emitting direction of the light source. The computer device can then obtain the initial images collected by the target endoscope under different light source directions. For example, the user can trigger the control button of the adjustment component at the control end of the target endoscope to control the adjustment component to change the light-emitting direction of the light source, and control the target endoscope to collect the above initial images under different light source directions.

[0107] Optionally, the target endoscope responds to a stop instruction from the user / computer device to stop collecting the initial images, or can also automatically stop after detecting a bleeding point in the initial images.

[0108] It should be noted that the above initial images under different light source directions are collected based on the target endoscope at the same acquisition position. Multiple initial images under different light source directions can be collected at the same acquisition position as a group of initial images, and each group of initial images is used to synthesize a stereoscopic image. During the use of the target endoscope, it will move, and the computer device can obtain the initial images collected by the target endoscope under different light source directions at different acquisition positions. For example, taking the operating environment of the target endoscope as a simulated cavity of a living organism as an example, the target endoscope can collect multiple groups of initial images under different light source directions at acquisition position A, can also continue to collect multiple groups of initial images under different light source directions at acquisition position B, and can also continue to collect multiple groups of initial images under different light source directions at acquisition position C.

[0109] S620. Synthesize and process the initial images under different light source directions to obtain the stereoscopic image at the target moment.

[0110] Optionally, the target moment can be the earliest acquisition moment, the latest acquisition moment, or the middle acquisition moment among the acquisition moments corresponding to the initial images obtained under different light source directions.

[0111] Specifically, for the same acquisition position, the computer device uses photometric stereo synthesis to synthesize each group of initial images under different light source directions to obtain a stereo image at the target moment. For example, the computer device obtains two groups of initial images collected by the target endoscope at the above-mentioned acquisition position A. The first group of initial images includes a first initial image (acquisition moment t1, light source direction F1), a second initial image (acquisition moment t2, light source direction F2), and a third initial image (acquisition moment t3, light source direction F3); the second group of initial images includes a fourth initial image (acquisition moment t4, light source direction F1), a fifth initial image (acquisition moment t5, light source direction F2), and a sixth initial image (acquisition moment t6, light source direction F3). Among them, t1 < t2 < t3 < t4 < t5 < t6. The computer device then performs photometric stereo synthesis on the first group of initial images to obtain a stereo image at the target moment t1, and performs photometric stereo synthesis on the second group of initial images to obtain a stereo image at the target moment t4.

[0112] S630. Arrange the stereo images at the target moments in chronological order to obtain a historical sequence of images.

[0113] Among them, the historical sequence of images includes stereo images corresponding to multiple target moments. The historical sequence of images also includes stereo images corresponding to multiple acquisition positions.

[0114] Specifically, the computer device sorts all the obtained stereo images according to the target moment of each stereo image to obtain a historical sequence of images with a time sequence relationship.

[0115] In an optional embodiment, a light-shielding sheet is provided at the end of the endoscope body of the target endoscope. By adjusting the light-shielding position of the light-shielding sheet on the endoscope body, partial occlusion of the light source inside the endoscope body can be achieved, thereby changing the light source direction. Based on this, as Figure 7 shown, the above S610. Obtain the initial images under different light source directions of the target endoscope, including:

[0116] S710. Change the light source direction by adjusting the light-shielding position of the light-shielding sheet on the target endoscope.

[0117] Among them, the light source direction is the light-emitting direction of the light source. The light-shielding sheet is located at the end of the endoscope body of the target endoscope and covers part of the light source. As Figure 8 shown, the end of the endoscope body of the target endoscope includes a light outlet 801 for placing the light source module, and a light-shielding sheet is provided at the light outlet, and there are also openings for implementing other functions, such as a forceps port 802 for surgical instruments to pass through and a camera port 803 for placing the camera module.

[0118] Optionally, as Figure 9As shown, the above light-shielding sheet can be of the 4-division type or the 8-division type. Among them, the 4-division type light-shielding sheet means that 1 / 4 of the area is reserved for light output and 3 / 4 of the area is blocked, and the 8-division type light-shielding sheet means that 1 / 8 of the area is reserved for light output and 7 / 8 of the area is blocked. In this embodiment, the light-shielding sheet adapts to the shape of the light outlet and is circular.

[0119] Optionally, the computer device can adjust the light-shielding position of the light-shielding sheet on the target endoscope based on the user's control operation to change the light source direction. Optionally, the light-shielding sheet can be controlled to rotate multiple times to complete a 360° rotation. Each rotation changes the light source direction once, and it can rotate clockwise or counterclockwise. For example, taking the Figure 9 4-division type in it as an example, rotate the light-transmitting area (white area) of the light-shielding sheet to position ①, corresponding to the light source direction S1, rotate the light-transmitting area of the light-shielding sheet to position ②, corresponding to the light source direction S2, rotate the light-transmitting area of the light-shielding sheet to position ③, corresponding to the light source direction S3, and then rotate the light-transmitting area of the light-shielding sheet to position ④, corresponding to the light source direction S4. At this time, the target endoscope completes a 360° rotation.

[0120] S720. Obtain images of the operating environment of the target endoscope when the light-shielding sheet is in different light-shielding positions as initial images.

[0121] Specifically, when the light-shielding sheet of the target endoscope is in different light-shielding positions, the computer device obtains the images of the current operating environment collected by the target endoscope as initial images. Continuing the above example, the target endoscope obtains the image of the current operating environment when the light-transmitting area of the light-shielding sheet is rotated to position ①, corresponding to the light source direction S1, as the first initial image, obtains the image of the current operating environment when the light-transmitting area of the light-shielding sheet is rotated to position ②, corresponding to the light source direction S2, as the second initial image, obtains the image of the current operating environment when the light-transmitting area of the light-shielding sheet is rotated to position ③, corresponding to the light source direction S3, as the third initial image, and obtains the image of the current operating environment when the light-transmitting area of the light-shielding sheet is rotated to position ④, corresponding to the light source direction S4, as the fourth initial image.

[0122] It should be noted that the to-be-detected images and the initial to-be-detected images in the foregoing embodiments are all three-dimensional images obtained by photometric stereo synthesis from the initial images under different light source directions. The specific processes of collecting the initial images and photometric synthesis are also the same as the above processes.

[0123] In this embodiment, the computer device acquires initial images under different light source directions of the target endoscope, performs synthesis processing on the initial images under different light source directions to obtain a stereoscopic image at the target moment, and then arranges the stereoscopic images at the target moment in chronological order to obtain a historical sequence of images. The above historical sequence of images can cover the entire process of the application of the target endoscope, record the whole process from the absence of bleeding points to the emergence of bleeding points in the operating environment, which helps to determine the position of the bleeding point based on the historical sequence of images subsequently. Moreover, each frame of the image in the historical sequence of images is a stereoscopic image synthesized by the luminosities of the initial images under different light source directions, covering multi-dimensional information, which helps to determine the conversion relationship between the image to be detected and the sample image subsequently, and further improves the accuracy of bleeding point detection.

[0124] To further improve the accuracy of the determined bleeding point position, the above sample image can be the image when the bleeding point first appears. In order to determine the image when the bleeding point first appears in the historical sequence of images, as Figure 10 shown, the above method further includes:

[0125] S1010. Perform differential processing on two adjacent stereoscopic images in the historical sequence of images to obtain the difference features between the two adjacent stereoscopic images in the historical sequence of images.

[0126] Optionally, when the historical sequence of images includes a set of stereoscopic images corresponding to multiple acquisition positions (each acquisition position corresponds to a set of stereoscopic images), the computer device performs differential processing on two adjacent stereoscopic images belonging to the same set of stereoscopic images in the historical sequence of images, that is, weakens the similar parts between the two frames of stereoscopic images and highlights the changing parts between the two frames of stereoscopic images, so as to obtain the difference features between the two adjacent stereoscopic images.

[0127] As Figure 11 shown, there are multiple frames of stereoscopic images at the same acquisition position in the historical sequence of images, and each frame of stereoscopic image is synthesized by the luminosities of multiple initial images. In order to determine the stereoscopic image (i.e., the sample image) when the bleeding point first appears from the above historical sequence of images, the computer device performs differential processing on two adjacent stereoscopic images, that is, between the second frame and the first frame, the third frame and the second frame, the fourth frame and the third frame, and so on, to obtain the difference features between the two adjacent stereoscopic images.

[0128] S1020. Determine the stereoscopic image when the bleeding point first appears according to the difference features, and use the stereoscopic image when the bleeding point first appears as the sample image.

[0129] It should be noted that for the same acquisition position, the operating environment of the target endoscope generally does not change significantly. Therefore, there are no differential features between two adjacent frames of stereoscopic images. However, there will be differential features between two adjacent frames of stereoscopic images after a bleeding point appears in the operating environment.

[0130] Specifically, after obtaining the differential features between two adjacent frames of stereoscopic images, the computer device determines whether a bleeding point has appeared during the acquisition time interval from the previous frame of stereoscopic image to the next frame of stereoscopic image based on these differential features. And when it is determined that a bleeding point has appeared, the computer device determines the next frame of stereoscopic image in the two adjacent frames of stereoscopic images as the stereoscopic image where the bleeding point first appears as the above sample image.

[0131] Optionally, if the above differential feature is that red pixel points appear in the next frame of stereoscopic image relative to the previous frame of stereoscopic image or the pixel color has changed and the changed area meets the preset condition, the computer device can determine that a bleeding point has appeared. If the above differential feature is that the texture of a partial area in the next frame of stereoscopic image is deepened relative to the previous frame of stereoscopic image, the computer device can also determine that a bleeding point has appeared. Continuing with the above example, the computer device can determine that a bleeding point has appeared based on the differential features between the 5th frame and the 4th frame, and can determine the 5th frame as the above sample image.

[0132] In this embodiment, the computer device performs differential processing on two adjacent frames of stereoscopic images in the historical sequence of images to obtain the differential features between the two adjacent frames of stereoscopic images in the historical sequence of images, and then determines the stereoscopic image where the bleeding point first appears according to the differential features, and uses the stereoscopic image where the bleeding point first appears as the sample image. During the application process of the target endoscope, the whole process from the initial absence of bleeding points to the appearance of bleeding points in the operation scenario will be experienced. Correspondingly, the historical sequence of images includes images without bleeding points and images with bleeding points at the same acquisition position. By using the above differential processing method, the stereoscopic image where the bleeding point first appears, that is, the above sample image, can be determined among multiple frames of stereoscopic images at the same acquisition position in the historical sequence of images. Since there are few interfering elements in the stereoscopic image where the bleeding point first appears, subsequent bleeding point detection based on this sample image can improve the accuracy of the determined bleeding point position.

[0133] In one embodiment, as Figure 12 shown, a bleeding point detection method is further provided, including the following steps:

[0134] S1210. Obtain an initial image to be detected, and determine the similarity between the initial image to be detected and the sample image; wherein, the sample image is an image including a bleeding point in the historical sequence of images.

[0135] S1220. If the similarity is greater than the similarity threshold, determine the initial image to be detected as the image to be detected.

[0136] S1230. According to the affine transformation matrix between the image to be detected and the sample image, convert the position of the bleeding point in the sample image to the image to be detected, so as to obtain the position of the bleeding point in the image to be detected.

[0137] Optionally, the specific process of the computer device for bleeding point detection and determining the bleeding point position is as follows:

[0138] The doctor uses the target endoscope to perform a medical operation (such as a suturing operation) in a corresponding operation environment (such as simulating the inside of a human tissue / cavity). During the medical operation, the target endoscope can, under the action of the doctor's operation instruction or the control instruction conveyed by the computer device, obtain multiple groups of initial images in different light source directions at the acquisition position of the target endoscope. The computer device acquires the initial images collected by the target endoscope, and photometrically synthesizes each group of initial images into a frame of stereoscopic image. All the obtained stereoscopic images are arranged according to the acquisition time to form a historical sequence of images. Perform differential processing on two adjacent frames of stereoscopic images in the historical sequence of images to obtain the difference features between two adjacent frames of stereoscopic images in the historical sequence of images; determine the stereoscopic image in which the bleeding point first appears according to the difference features, and use the stereoscopic image in which the bleeding point first appears as the sample image. At the same time, obtain the neural network model for identifying the operation environment in the image, and the affine transformation matrix between the image to be detected and the sample image. The specific process can refer to the above related embodiments and will not be elaborated here.

[0139] After the doctor finishes the above medical operation, it is necessary to find the bleeding point generated during the operation. Combining Figure 13 , the computer device acquires a frame of initial image to be detected through the target endoscope. This initial image to be detected is also a stereoscopic image photometrically synthesized based on multiple groups of initial images to be detected in different light source directions. The computer device acquires this initial image to be detected, and uses the neural network model to identify this initial image to be detected, obtain the similarity between the initial image to be detected and the sample image, and when the similarity is greater than the similarity threshold, determine the initial image to be detected as the image to be detected. Furthermore, according to the affine transformation matrix between the image to be detected and the sample image, convert the position of the bleeding point in the sample image to the image to be detected, so as to obtain the position of the bleeding point in the image to be detected, and display the bleeding point in real time in the image to be detected.

[0140] It should be noted that the specific processes of the relevant steps in this embodiment can be seen in the above related embodiments and will not be elaborated here. Through the above method, the detection of the bleeding point can be realized, and the accuracy of the determined bleeding point position is improved.

[0141] It should be understood that although the steps in the flowcharts involved in the above-described embodiments are sequentially shown according to the indications of the arrows, these steps are not necessarily executed sequentially in the order indicated by the arrows. Unless there is a clear indication in this article, the execution of these steps has no strict order limit, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-described embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be executed alternately or in turn with at least a part of other steps or steps or stages in other steps.

[0142] In one embodiment, as Figure 14 shown, a bleeding point detection device is provided, including an image acquisition module 1401, a conversion determination module 1402, and a position conversion module 1403, where:

[0143] The image acquisition module 1401 is configured to acquire an image to be detected; the conversion determination module 1402 is configured to determine the conversion relationship between the image to be detected and the sample image based on the image to be detected; wherein, the sample image is an image including a bleeding point in the historical sequence images of the target endoscope;

[0144] The position conversion module 1403 is configured to convert the bleeding point position in the sample image to the image to be detected according to the conversion relationship, so as to obtain the bleeding point position in the image to be detected.

[0145] In one of the embodiments, the conversion determination module 1402 is specifically configured to:

[0146] Obtain the position information of the matching feature point pairs between the image to be detected and the sample image; determine the affine transformation matrix between the image to be detected and the sample image according to the position information of the matching feature point pairs; and determine the affine transformation matrix as the conversion relationship between the image to be detected and the sample image.

[0147] In one of the embodiments, the conversion determination module 1402 is specifically configured to:

[0148] Perform feature extraction on the image to be detected to obtain the multi-dimensional features of at least one first feature point in the image to be detected; perform feature matching on the multi-dimensional features of at least one first feature point and the multi-dimensional features of at least one second feature point in the sample image, and determine the first feature point and the second feature point with the highest feature matching degree as the matching feature point pairs; obtain the position information of the first feature point in the matching feature point pair in the image to be detected, and the position information of the second feature point in the matching feature point pair in the sample image.

[0149] In one embodiment, the image acquisition module 1401 is specifically configured to:

[0150] Acquire an initial image to be detected, and determine the similarity between the initial image to be detected and a sample image; if the similarity is greater than the similarity threshold, determine the initial image to be detected as the image to be detected.

[0151] In one embodiment, the image acquisition module 1401 is further configured to:

[0152] Acquire initial images under different light source directions of the target endoscope; perform a synthesis process on the initial images under different light source directions to obtain a stereoscopic image at the target moment; arrange the stereoscopic images at the target moment in chronological order to obtain a historical sequence of images.

[0153] In one embodiment, the image acquisition module 1401 is specifically configured to:

[0154] Change the light source direction by adjusting the light-shielding position of the light-shielding plate on the target endoscope; wherein, the light-shielding plate is located at the end of the mirror body of the target endoscope and covers part of the light source; acquire images of the operating environment of the target endoscope when the light-shielding plate is in different light-shielding positions as the initial images.

[0155] In one embodiment, the image acquisition module 1401 is further configured to:

[0156] Perform a differential process on two adjacent stereoscopic images in the historical sequence of images to obtain the difference features between the two adjacent stereoscopic images in the historical sequence of images; determine the stereoscopic image in which the bleeding point first appears according to the difference features, and use the stereoscopic image in which the bleeding point first appears as the sample image.

[0157] Each module in the above bleeding point detection device can be implemented in whole or in part by software, hardware, and their combination. The above modules can be embedded in the processor of the computer device in hardware form or independent of it, or stored in the memory of the computer device in software form, so that the processor can call and execute the operations corresponding to the above respective modules.

[0158] In one embodiment, a computer device is provided. The computer device can be a terminal, and its internal structure diagram can be as Figure 15As shown. The computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected via a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a mobile cellular network, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements a bleeding point detection method. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or buttons, a trackball, or a touchpad provided on the housing of the computer device, or an external keyboard, touchpad, or mouse, etc.

[0159] Those skilled in the art can understand that Figure 15 the structure shown in is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have a different component layout.

[0160] In one embodiment, a computer device is provided, including a memory and a processor. A computer program is stored in the memory. When the processor executes the computer program, the following steps are implemented:

[0161] Obtain the image to be detected; based on the image to be detected, determine the conversion relationship between the image to be detected and the sample image; wherein, the sample image is an image including a bleeding point in the historical sequence of images of the target endoscope; according to the conversion relationship, convert the position of the bleeding point in the sample image to the image to be detected to obtain the position of the bleeding point in the image to be detected.

[0162] In one of the embodiments, when the processor executes the computer program, the following steps are also implemented:

[0163] Obtain the position information of the matching feature point pairs between the image to be detected and the sample image; determine the affine transformation matrix between the image to be detected and the sample image according to the position information of the matching feature point pairs; determine the affine transformation matrix as the conversion relationship between the image to be detected and the sample image.

[0164] In one of the embodiments, when the processor executes the computer program, the following steps are also implemented:

[0165] Feature extraction is performed on the image to be detected to obtain the multi-dimensional features of at least one first feature point in the image to be detected; the multi-dimensional features of at least one first feature point are feature-matched with the multi-dimensional features of at least one second feature point in the sample image, and the first feature point and the second feature point with the highest feature matching degree are determined as the matching feature point pair; the position information of the first feature point in the matching feature point pair in the image to be detected and the position information of the second feature point in the matching feature point pair in the sample image are obtained.

[0166] In one embodiment, when the processor executes the computer program, the following steps are further implemented:

[0167] An initial image to be detected is obtained, and the similarity between the initial image to be detected and the sample image is determined; if the similarity is greater than the similarity threshold, the initial image to be detected is determined as the image to be detected.

[0168] In one embodiment, when the processor executes the computer program, the following steps are further implemented:

[0169] Initial images under different light source directions of the target endoscope are obtained; the initial images under different light source directions are synthesized to obtain a stereoscopic image at the target moment; the stereoscopic images at the target moment are arranged in chronological order to obtain a historical sequence of images.

[0170] In one embodiment, when the processor executes the computer program, the following steps are further implemented:

[0171] The light source direction is changed by adjusting the light-shielding position of the light-shielding sheet on the target endoscope; wherein, the light-shielding sheet is located at the end of the body of the target endoscope and covers part of the light source; images of the operating environment of the target endoscope when the light-shielding sheet is at different light-shielding positions are obtained as the initial images.

[0172] In one embodiment, when the processor executes the computer program, the following steps are further implemented:

[0173] Differential processing is performed on two adjacent stereoscopic images in the historical sequence of images to obtain the difference features between two adjacent stereoscopic images in the historical sequence of images; the stereoscopic image in which the bleeding point first appears is determined according to the difference features, and the stereoscopic image in which the bleeding point first appears is used as the sample image.

[0174] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented:

[0175] Obtain the image to be detected; based on the image to be detected, determine the conversion relationship between the image to be detected and the sample image; wherein, the sample image is an image including a bleeding point in the historical sequence images of the target endoscope; convert the position of the bleeding point in the sample image to the image to be detected according to the conversion relationship, and obtain the position of the bleeding point in the image to be detected.

[0176] In one embodiment, when the computer program is executed by the processor, the following steps are further implemented:

[0177] Obtain the position information of the pairs of matching feature points between the image to be detected and the sample image; determine the affine transformation matrix between the image to be detected and the sample image according to the position information of the pairs of matching feature points; and determine the affine transformation matrix as the conversion relationship between the image to be detected and the sample image.

[0178] In one embodiment, when the computer program is executed by the processor, the following steps are further implemented:

[0179] Extract features from the image to be detected to obtain the multi-dimensional features of at least one first feature point in the image to be detected; perform feature matching on the multi-dimensional features of at least one first feature point and the multi-dimensional features of at least one second feature point in the sample image, and determine the first feature point and the second feature point with the highest feature matching degree as the pair of matching feature points; obtain the position information of the first feature point in the pair of matching feature points in the image to be detected, and the position information of the second feature point in the pair of matching feature points in the sample image.

[0180] In one embodiment, when the computer program is executed by the processor, the following steps are further implemented:

[0181] Obtain the initial image to be detected, and determine the similarity between the initial image to be detected and the sample image; if the similarity is greater than the similarity threshold, determine the initial image to be detected as the image to be detected.

[0182] In one embodiment, when the computer program is executed by the processor, the following steps are further implemented:

[0183] Obtain the initial images under different light source directions of the target endoscope; perform synthesis processing on the initial images under different light source directions to obtain the stereoscopic image at the target moment; arrange the stereoscopic images at the target moment in chronological order to obtain the historical sequence images.

[0184] In one embodiment, when the computer program is executed by the processor, the following steps are further implemented:

[0185] Change the light source direction by adjusting the light-shielding position of the light-shield on the target endoscope; wherein, the light-shield is located at the end of the body of the target endoscope and covers part of the light source; obtain images of the operating environment of the target endoscope when the light-shield is in different light-shielding positions as initial images.

[0186] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0187] Perform differential processing on two adjacent stereo images in the historical sequence of images to obtain the difference features between the two adjacent stereo images in the historical sequence of images; determine the stereo image in which the bleeding point first appears according to the difference features, and use the stereo image in which the bleeding point first appears as the sample image.

[0188] In one embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0189] Obtain the image to be detected; based on the image to be detected, determine the conversion relationship between the image to be detected and the sample image; wherein, the sample image is an image including a bleeding point in the historical sequence of images of the target endoscope; convert the position of the bleeding point in the sample image to the image to be detected according to the conversion relationship to obtain the position of the bleeding point in the image to be detected.

[0190] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0191] Obtain the position information of the matching feature point pairs between the image to be detected and the sample image; determine the affine transformation matrix between the image to be detected and the sample image according to the position information of the matching feature point pairs; and determine the affine transformation matrix as the conversion relationship between the image to be detected and the sample image.

[0192] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0193] Extract features from the image to be detected to obtain the multi-dimensional features of at least one first feature point in the image to be detected; perform feature matching between the multi-dimensional features of at least one first feature point and the multi-dimensional features of at least one second feature point in the sample image, and determine the first feature point and the second feature point with the highest feature matching degree as the matching feature point pairs; obtain the position information of the first feature point in the matching feature point pair in the image to be detected, and the position information of the second feature point in the matching feature point pair in the sample image.

[0194] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0195] Obtain an initial image to be detected, and determine the similarity between the initial image to be detected and a sample image; if the similarity is greater than the similarity threshold, determine the initial image to be detected as the image to be detected.

[0196] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0197] Obtain initial images under different light source directions of the target endoscope; perform synthesis processing on the initial images under different light source directions to obtain a stereoscopic image at the target moment; arrange the stereoscopic images at the target moment in chronological order to obtain a historical sequence of images.

[0198] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0199] Change the light source direction by adjusting the light-shielding position of the light-shielding sheet on the target endoscope; wherein, the light-shielding sheet is located at the end of the body of the target endoscope and covers part of the light source; obtain images of the operating environment of the target endoscope when the light-shielding sheet is in different light-shielding positions as the initial images.

[0200] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0201] Perform differential processing on two adjacent stereoscopic images in the historical sequence of images to obtain the difference features between two adjacent stereoscopic images in the historical sequence of images; determine the stereoscopic image in which the bleeding point first appears according to the difference features, and use the stereoscopic image in which the bleeding point first appears as the sample image.

[0202] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, database, or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memories. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The databases involved in the embodiments provided in the present application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the embodiments provided in the present application can be general-purpose processors, central processors, graphics processors, digital signal processors, programmable logic devices, data processing logics based on quantum computing, etc., without limitation.

[0203] The technical features of the above embodiments can be combined arbitrarily. For the sake of concise description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.

[0204] The above-described embodiments merely represent several implementation manners of the present application. Their descriptions are relatively specific and detailed, but they should not be construed as limiting the patent scope of the present application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the appended claims.

Claims

1. A method for detecting bleeding points, characterized in that, Including: Obtain initial images under different light source directions of a target endoscope; Perform synthesis processing on the initial images under the different light source directions to obtain a stereoscopic image at a target moment; Arrange the stereoscopic images at the target moment in chronological order to obtain a historical sequence of images; Obtain an initial image to be detected and determine the similarity between the initial image to be detected and a sample image; the sample image is an image including a bleeding point in the historical sequence of images; If the similarity is greater than a similarity threshold, determine the initial image to be detected as an image to be detected; the image to be detected is an image with a bleeding point covered by blood; Based on the image to be detected, determine the conversion relationship between the image to be detected and the sample image; According to the conversion relationship, convert the position of the bleeding point in the sample image to the image to be detected to obtain the position of the bleeding point in the image to be detected; The obtaining of the initial images under different light source directions of the target endoscope includes: Change the light source direction by adjusting the light-shielding position of a light-shield on the target endoscope; wherein, the light-shield is located at the end of the mirror body of the target endoscope and covers part of the light source; Obtain an image of the operating environment of the target endoscope when the light-shield is at different light-shielding positions as the initial image.

2. The method according to claim 1, wherein The determining of the conversion relationship between the image to be detected and the sample image based on the image to be detected includes: Obtain the position information of pairs of feature points that match between the image to be detected and the sample image; Determine an affine transformation matrix between the image to be detected and the sample image according to the position information of the pairs of matching feature points; Determine the affine transformation matrix as the conversion relationship between the image to be detected and the sample image.

3. The method according to claim 2, wherein The obtaining of the position information of pairs of feature points that match between the image to be detected and the sample image includes: Perform feature extraction on the image to be detected to obtain multi-dimensional features of at least one first feature point in the image to be detected; Perform feature matching between the multi-dimensional features of the at least one first feature point and the multi-dimensional features of at least one second feature point in the sample image, and determine the first feature point and the second feature point with the highest feature matching degree as the pairs of matching feature points; Obtain the position information of the first feature point in the pair of matching feature points in the image to be detected, and the position information of the second feature point in the pair of matching feature points in the sample image.

4. The method according to claim 3, wherein The method further includes: Use a fast feature point extraction and description feature extraction algorithm or a deep learning neural network model to obtain the first feature points in the image to be detected and the second feature points in the sample image.

5. The method according to any one of claims 1 to 3, characterized in that The similarity is determined based on the structural similarity measure or vector cosine value between the initial image to be detected and the sample image.

6. The method according to any one of claims 1 to 3, characterized in that, The method further includes: Perform differential processing on two adjacent frames of stereoscopic images in the historical sequence of images to obtain the difference features between two adjacent frames of stereoscopic images in the historical sequence of images; Determine a stereoscopic image of the first appearance of a bleeding point according to the differential feature, and use the stereoscopic image of the first appearance of the bleeding point as the sample image.

7. A bleeding point detection device, characterized in that, The device includes: An image acquisition module, configured to acquire initial images under different light source directions of a target endoscope; perform synthesis processing on the initial images under different light source directions to obtain a stereoscopic image at a target moment; arrange the stereoscopic images at the target moment in chronological order to obtain a historical sequence of images; acquire an initial image to be detected, and determine the similarity between the initial image to be detected and the sample image; if the similarity is greater than a similarity threshold, determine the initial image to be detected as an image to be detected; the sample image is an image including a bleeding point in the historical sequence of images; the image to be detected is an image with the bleeding point covered by blood; A conversion determination module, configured to determine a conversion relationship between the image to be detected and the sample image based on the image to be detected; A position conversion module, configured to convert the position of the bleeding point in the sample image to the image to be detected according to the conversion relationship to obtain the position of the bleeding point in the image to be detected; The image acquisition module is further configured to: change the light source direction by adjusting the light-shielding position of a light-shielding sheet on the target endoscope; wherein, the light-shielding sheet is located at the end of the body of the target endoscope and covers part of the light source; acquire images of the operating environment of the target endoscope when the light-shielding sheet is at different light-shielding positions as the initial images.

8. A computer device, comprising an end memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images