CT feature extraction method of digital image system for CT machine

Abstract

The invention relates to the technical field of image processing, in particular to a CT feature extraction method of a digital image system for a CT machine. The method comprises the following steps: acquiring a plurality of to-be-processed images, and performing image segmentation on the plurality of to-be-processed images to obtain a plurality of initial areas included in each to-be-processed image; boundary feature analysis is carried out on all initial regions included in the plurality of to-be-processed images to obtain a plurality of target region positions; analyzing pixel features of each target area position in each to-be-processed image to obtain a plurality of fusion weights corresponding to each target area position; and according to the plurality of fusion weights corresponding to each target area position, fusing the plurality of images to be processed to obtain a fused image. According to the invention, the fusion effect of the multi-angle CT digital images of the same human body part can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image processing, and in particular to a CT feature extraction method for a digital image system of a CT machine. BACKGROUND

[0002] Since the development of a computer X-ray tomography machine, i.e., a CT machine, the CT machine has been widely applied in the field of clinical medicine.

[0003] The CT machine performs scanning on an X-ray beam emitted from a certain part of a human body. When the X-ray beam is directed to human tissue, part of the X-ray beam is absorbed by the human tissue, and another part of the X-ray beam passes through the human body and is received by a detector device on the CT machine, thereby generating an electrical signal.

[0004] Different human tissues have different densities, and the penetration ability of X-rays on different tissues is also different. Compared with traditional X-ray photography, the CT machine can distinguish very small differences in X-ray absorption values. Therefore, the detector can receive X-rays of different intensities after passing through different human tissues, distinguish X-ray absorption values, and distinguish more bones, fat, and other soft tissues in the human body. The electrical signals generated by the X-rays absorbed at different positions in the scanning layer are converted into gray values by computer calculation, and the CT digital image of the scanned layer is obtained.

[0005] In order to reduce the radiation impact of the CT machine on patients, a new type of light quantum digital CT machine uses a small power X-ray source. However, the low-power X-ray source has low penetration, which may cause the digital image obtained by scanning to be blurred in some areas. Therefore, the CT machine is commonly used to take CT digital images of the pathological part of a patient from multiple directions and angles, and to fuse the corresponding positions of multiple images to obtain a CT digital image that can more accurately reflect the patient's condition.

[0006] However, in the application, it is found that when fusing multiple-angle CT digital images of the same part of a human body, the existing technology commonly uses a mean weighted fusion method to fuse the images after weighting different images based on image features such as overall brightness and blurriness. Due to the small power of the X-ray, there are images with inconsistent clarity of internal tissues. The use of uniform weighting cannot make good use of the high clarity of the local area in the image, resulting in an unsatisfactory effect of the final fused digital image.

[0007] That is, the effect of fusing multiple-angle CT digital images of the same part of a human body based on the existing technology is poor. SUMMARY

[0008] In order to solve the technical problem of poor effect of fusing multiple-angle CT digital images of the same part of a human body based on the existing technology, the purpose of the present application is to provide a CT feature extraction method for a digital image system of a CT machine. The technical solution adopted is as follows: In a first aspect, one embodiment of the present application provides a CT feature extraction method for a digital image system of a CT machine, the method comprising: obtaining a plurality of to-be-processed images, wherein the plurality of to-be-processed images are a plurality of images obtained by image rotation on a plurality of initial images according to standard angles, and the plurality of initial images are a plurality of CT digital images obtained by scanning a target human body part from different angles; performing image segmentation on the plurality of to-be-processed images respectively to obtain a plurality of initial regions included in each to-be-processed image; performing boundary feature analysis on all initial regions included in the plurality of to-be-processed images to obtain a plurality of target region positions, wherein the plurality of target region positions represent a plurality of human body tissues in the target human body part; performing analysis on pixel features of each target region position in each to-be-processed image to obtain a plurality of fusion weights corresponding to each target region position, wherein the plurality of fusion weights corresponding to each target region position correspond to the plurality of to-be-processed images one by one; performing fusion on the plurality of to-be-processed images according to the plurality of fusion weights corresponding to each target region position to obtain a fused image.

[0009] In one embodiment, the performing boundary feature analysis on all initial regions included in the plurality of to-be-processed images to obtain a plurality of target region positions comprises: performing analysis on a plurality of initial boundary points corresponding to the plurality of to-be-processed images to determine a plurality of final boundary point positions, wherein the initial boundary points are image points located on a region boundary of an initial region in a to-be-processed image; determining the plurality of target region positions based on the plurality of final boundary point positions.

[0010] In one embodiment, the performing analysis on a plurality of initial boundary points on a region boundary of each initial region to determine a plurality of final boundary point positions comprises: calculating a ratio of a first quantity and a second quantity of a target boundary point position to obtain a quantity index corresponding to the target boundary point position, wherein the target boundary point position represents a position of any one of all initial boundary points included in the plurality of to-be-processed images, the first quantity represents a quantity of initial boundary points corresponding to the target boundary point position in the plurality of to-be-processed images, and the second quantity represents a quantity of all image points corresponding to the target boundary point position in the plurality of to-be-processed images; obtaining a plurality of boundary interval distances of a plurality of internal image points corresponding to the target boundary point position in the plurality of to-be-processed images, wherein the internal image points represent image points located inside an initial region, and the boundary interval distance is a shortest distance from a corresponding internal image point to a region boundary of the initial region. calculating a sum value of the plurality of boundary interval distances corresponding to the target boundary point, to obtain a distance index corresponding to the target boundary point; determining whether the target boundary point is the final boundary point according to the quantity index and the distance index corresponding to the target boundary point.

[0011] In one embodiment, the determining whether the target boundary point is the final boundary point according to the quantity index and the distance index corresponding to the target boundary point comprises: calculating a ratio value of the quantity index and the distance index corresponding to the target boundary point, to obtain a boundary feature index of the target boundary point; in a case where the boundary feature index of the target boundary point is greater than or equal to a first threshold value, determining the target boundary point as the final boundary point; in a case where the boundary feature index of the target boundary point is less than the first threshold value, not determining the target boundary point as the final boundary point.

[0012] In one embodiment, the analyzing the pixel features of each target region position in each to-be-processed image to obtain a plurality of fusion weights corresponding to each target region position comprises: performing similarity analysis on the image region indicated by each target region position in each to-be-processed image and the corresponding initial region, to obtain a plurality of region similarity indexes corresponding to each target region position, wherein the plurality of region similarity indexes correspond to the plurality of to-be-processed images one by one; performing gray scale analysis on a plurality of image points corresponding to each target region position in each to-be-processed image, to obtain a plurality of gray scale significant indexes corresponding to each target region position, wherein the plurality of gray scale significant indexes correspond to the plurality of to-be-processed images one by one; obtaining a plurality of fusion weights corresponding to each target region position according to the plurality of region similarity indexes and the plurality of gray scale significant indexes corresponding to each target region position.

[0013] In one embodiment, the performing similarity analysis on each target region position and the corresponding initial region to obtain a plurality of region similarity indexes corresponding to each target region position comprises: analyzing the region overlap degree between the image region indicated by each target region position in each to-be-processed image and the corresponding initial region, to obtain a plurality of first similarity indexes of each target region position, wherein the plurality of first similarity indexes correspond to the plurality of to-be-processed images one by one; analyzing a boundary overlap degree between an image region indicated by each target region position in each to-be-processed image and a corresponding initial region to obtain a plurality of second similarity indexes of each target region position, wherein the plurality of second similarity indexes and the plurality of to-be-processed images correspond one-to-one; calculating a product of each first similarity index of each target region position and a corresponding second similarity index to obtain a plurality of region similarity indexes corresponding to each target region position.

[0014] In an embodiment, the grayscale analysis on the plurality of image points corresponding to each target region position in each to-be-processed image to obtain a plurality of grayscale significant indexes corresponding to each target region position comprises: analyzing a discrete degree of a plurality of grayscale values of the plurality of image points corresponding to each target region position in each to-be-processed image to obtain a plurality of internal grayscale uniformity indexes corresponding to each target region position, wherein the plurality of internal grayscale uniformity indexes corresponding to each target region position correspond one-to-one to the plurality of to-be-processed images; analyzing a difference between a grayscale distribution of each target region position in each to-be-processed image and a grayscale distribution of a neighboring region of the target region position in the corresponding to-be-processed image to obtain a plurality of external grayscale difference indexes corresponding to each target region position, wherein the plurality of external grayscale difference indexes corresponding to each target region position correspond one-to-one to the plurality of to-be-processed images; obtaining the plurality of grayscale significant indexes corresponding to each target region position according to the plurality of internal grayscale uniformity indexes and the plurality of external grayscale difference indexes corresponding to each target region position.

[0015] In an embodiment, the analyzing the difference between the grayscale distribution of each target region position in each to-be-processed image and the grayscale distribution of the neighboring region of the target region position in the corresponding to-be-processed image to obtain the plurality of external grayscale difference indexes corresponding to each target region position comprises: respectively calculating distances between a corresponding centroid point of each target region position in each to-be-processed image and a plurality of centroid points of a plurality of neighboring regions of the target region position in the corresponding to-be-processed image to obtain a plurality of centroid distances corresponding to each target region position in each to-be-processed image; respectively calculating difference values between a corresponding grayscale mean value of each target region position in each to-be-processed image and a plurality of grayscale mean values of a plurality of neighboring regions of the target region position in the corresponding to-be-processed image to obtain a plurality of grayscale mean value differences corresponding to each target region position in each to-be-processed image; obtaining an external grayscale difference index corresponding to each target region position in each to-be-processed image according to the plurality of centroid distances and the plurality of grayscale mean value differences corresponding to each target region position in each to-be-processed image.

[0016] In one embodiment, the obtaining of the plurality of fusion weights corresponding to each target region position according to the plurality of region similarity indexes and the plurality of gray scale saliency indexes corresponding to each target region position comprises: calculating the product of each region similarity index corresponding to each target region position and the corresponding gray scale saliency index to obtain the plurality of fusion weights corresponding to each target region position.

[0017] In one embodiment, the fusing of the plurality of to-be-processed images according to the plurality of fusion weights corresponding to each target region position to obtain the fusion image comprises: calculating the sum of the plurality of fusion weights corresponding to each target region position to obtain a total weight value corresponding to each target region position; calculating the ratio of the fusion weight corresponding to each target region position in each to-be-processed image to the total weight value corresponding thereto to obtain a corrected weight corresponding to each target region position in each to-be-processed image; calculating the product of the gray scale value of each image point in each to-be-processed image and the corrected weight corresponding to the target region position where the image point is located to obtain a corrected gray scale value of each image point in each to-be-processed image; fusing the plurality of to-be-processed images based on the corrected gray scale value of each image point in each to-be-processed image to obtain the fusion image.

[0018] In a second aspect, another embodiment of the present application provides a CT feature extraction device of a digital image system of a CT machine, the device comprising: an image acquisition module configured to acquire a plurality of to-be-processed images, wherein the plurality of to-be-processed images are a plurality of images obtained by image rotation of a plurality of initial images according to a standard angle, and the plurality of initial images are a plurality of CT digital images obtained by scanning a target human body part from different angles; an image segmentation module configured to perform image segmentation on the plurality of to-be-processed images respectively to obtain a plurality of initial regions included in each to-be-processed image; a boundary analysis module configured to perform boundary feature analysis on all initial regions included in the plurality of to-be-processed images to obtain a plurality of target region positions, wherein the plurality of target region positions represent a plurality of human body tissues in the target human body part; a pixel analysis module configured to analyze the pixel features of each target region position in each to-be-processed image to obtain a plurality of fusion weights corresponding to each target region position, wherein the plurality of fusion weights corresponding to each target region position are in one-to-one correspondence with the plurality of to-be-processed images; An image fusion module is configured to fuse the multiple to-be-processed images according to the multiple fusion weights corresponding to each target region position, and obtain a fused image.

[0019] In a third aspect, a further embodiment of the present application provides an electronic device, including a processor, a memory, and a computer program stored in the memory and executable on the processor, and when the computer program is executed by the processor, the steps of the method in the first aspect are implemented.

[0020] In a fourth aspect, a further embodiment of the present application provides a computer-readable storage medium, and the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method in the first aspect are implemented.

[0021] The present application has the following advantages: After obtaining multiple CT digital images of a target human body part scanned from different angles, the present application first rotates the multiple CT digital images based on a standard angle to eliminate the angle difference between different images, and obtains multiple to-be-processed images with unified angles; then performs image segmentation on each to-be-processed image to segment multiple initial regions included in each to-be-processed image, and analyzes the boundary features of each initial region to suppress the segmentation error caused by the difference in image scanning angles, and obtain more accurate multiple target region positions for representing multiple human tissues in the target human body part; further analyzes the pixel features of each to-be-processed image in each target region position to determine the fusion weight of each target region position in each to-be-processed image, and finally performs fusion operation on the multiple to-be-processed images, which can adaptively adjust the gray value of each image point based on the pixel features of the region corresponding to different human tissues in each to-be-processed image, and assign higher weight to the region with more prominent image features, and assign lower weight to the region with relatively weak image features, so as to finally ensure that the fused image has high accuracy and clarity, that is, to improve the fusion effect of multiple-angle CT digital images of the same human body part. BRIEF DESCRIPTION OF DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, and the advantages thereof, the drawings needed in the embodiments or the prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can be obtained without creative labor based on these drawings.

[0023] Figure 1A schematic flow chart of a CT feature extraction method of a digital image system for a CT machine according to an embodiment of the present application is shown in FIG. 1. Figure 2 A schematic diagram of an image fusion example according to an embodiment of the present application is shown in FIG. 4. Figure 3 A structural schematic diagram of a CT feature extraction device of a digital image system for a CT machine according to an embodiment of the present application is shown in FIG. 5. Figure 4 A structural schematic diagram of an electronic device according to an embodiment of the present application is shown in FIG. 6. DETAILED DESCRIPTION

[0024] In order to further clarify the technical means and effects taken by the present application to achieve the predetermined purposes, the following describes in detail the specific implementation, structure, features and effects of a CT feature extraction method of a digital image system for a CT machine according to the present application, with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0025] 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 the present application belongs.

[0026] The following specifically describes a specific scheme of a CT feature extraction method of a digital image system for a CT machine according to the present application.

[0027] The present application proposes a CT feature extraction method of a digital image system for a CT machine, which is shown in FIG. 1. Figure 1 The method includes the following steps: Step S1, obtaining a plurality of to-be-processed images.

[0028] The plurality of to-be-processed images are a plurality of images obtained by image rotation on a plurality of initial images according to a standard angle, and the plurality of initial images are a plurality of CT digital images obtained by scanning a target human body part from different angles.

[0029] In the present application, the CT digital image is obtained based on the scanning of a light quantum digital CT machine. The light quantum digital CT machine is composed of two independent parts, i.e., a scanning main body (such as a scanning head, a scanning moving part, etc.) and a scanning bed. The X-ray source and the detector array are located on the same side, serving as a scanning head part. The X-ray source and the detector array are arranged in the scanning head at a specific range of angles. The CT machine uses an X-ray source of a small power, a fixed anode target form, and a large-angle plane fan beam or a cone beam. During the process of shooting the CT digital image, the two-dimensional array detector receives the X-rays penetrating through the human tissue to obtain a two-dimensional array. The two-dimensional array is directly arranged to obtain the CT scanning two-dimensional image. Due to the high degree of freedom and multi-axis intelligent motion of the scanning moving part, the scanning moving part can accurately and quickly move to various scanning positions.

[0030] When the light quantum digital CT machine is used, the patient needs to lie on the scanning bed under the guidance of the doctor. The doctor controls the scanning main body of the CT machine to perform multi-angle CT digital image scanning on the target human body part of the patient. The target human body part can be understood as the part to be detected of the patient, such as the trunk part.

[0031] The standard angle can be adaptively set according to actual needs. For example, the scanning angle corresponding to the initial image with the earliest scanning time in the plurality of initial images can be determined as the standard angle.

[0032] Exemplarily, the process of performing image rotation on the plurality of initial images according to the standard angle to obtain the plurality of to-be-processed images can be as follows: obtaining the angle difference (acute angle) between the scanning angle corresponding to each initial image and the standard angle and the angle deviation direction; rotating the corresponding initial image (the rotation center is the image center point of the initial image) by the corresponding angle (i.e., the angle difference of the corresponding initial image) in the reverse direction of the angle deviation direction corresponding to the initial image to obtain the plurality of to-be-processed images.

[0033] For example, if the angle difference between the scanning angle corresponding to an initial image and the standard angle is 25 degrees, the rotation angle of the initial image when rotated in the reverse direction of the corresponding angle deviation direction is also 25 degrees.

[0034] The image rotation operation can eliminate the angle difference between different CT digital images, so as to facilitate the smooth execution of the subsequent image fusion operation.

[0035] Step S2, performing image segmentation on the plurality of to-be-processed images respectively to obtain a plurality of initial regions included in each to-be-processed image.

[0036] Exemplarily, a region growing algorithm can be employed to perform image segmentation on each of the to-be-processed images to obtain a plurality of segmentation regions (i.e., a plurality of initial regions) included in each of the to-be-processed images, and each of the segmentation regions can also be referred to as a connected domain.

[0037] In step S3, boundary feature analysis is performed on all the initial regions included in the plurality of to-be-processed images to obtain a plurality of target region positions, where the plurality of target region positions represent a plurality of human tissues in the target human body part.

[0038] Due to the difference in scanning angles, the same human tissue in the target human body part can exhibit different gray value performances in different to-be-processed images, although the different to-be-processed images are all obtained by scanning the target human body part. The different gray value performances will affect the image segmentation of the plurality of to-be-processed images, and thus different segmentation results are generated. By performing boundary feature analysis on all the initial regions included in the plurality of to-be-processed images, the different segmentation results of the different to-be-processed images can be regularized, so that the plurality of to-be-processed images are regionally divided based on the same segmentation standard, thereby improving the fusion accuracy of the plurality of to-be-processed images in the subsequent step.

[0039] Specifically, the boundary feature analysis on all the initial regions included in the plurality of to-be-processed images to obtain the plurality of target region positions includes: analyzing a plurality of initial boundary points corresponding to the plurality of to-be-processed images to determine a plurality of final boundary points, where the initial boundary points are image points located on the region boundaries of the initial regions in the to-be-processed images; determining the plurality of target region positions based on the plurality of final boundary points.

[0040] For each of the to-be-processed images, after the image segmentation is performed to obtain the corresponding plurality of initial regions, the plurality of image points in the to-be-processed image can be divided into two categories, i.e., initial boundary points and internal image points, where the initial boundary points are image points located on the region boundaries (also referred to as region contours) of the initial regions, and the internal image points are image points located in the interiors of the initial regions.

[0041] Compared with the manner of analyzing the region contours of the plurality of initial regions corresponding to the plurality of to-be-processed images, the analysis on the plurality of initial boundary points corresponding to the plurality of to-be-processed images can achieve fine-grained processing of the region segmentation trends exhibited by the plurality of to-be-processed images, so that the determined target region positions are more accurate and reliable.

[0042] The plurality of target region positions constitute a unified segmentation standard for the plurality of to-be-processed images, each target region position corresponds to a plurality of image regions, and the plurality of image regions corresponding to each target region position correspond to the plurality of to-be-processed images in a one-to-one manner. An image region corresponding to a target region position in a to-be-processed image can be understood as a region indicated by the target region position in the to-be-processed image.

[0043] Further, the plurality of initial boundary points on the region boundary of each initial region are analyzed to determine a plurality of final boundary point positions, including: calculating a ratio of a first quantity and a second quantity of the target boundary point position to obtain a quantity index corresponding to the target boundary point position, wherein the target boundary point position represents a position of any one of all initial boundary points included in the plurality of to-be-processed images, the first quantity represents a quantity of initial boundary points corresponding to the target boundary point position in the plurality of to-be-processed images, and the second quantity represents a quantity of all image points corresponding to the target boundary point position in the plurality of to-be-processed images; obtaining a plurality of boundary interval distances of a plurality of internal image points corresponding to the target boundary point position in the plurality of to-be-processed images, wherein the internal image points represent image points located inside the initial region, and the boundary interval distance is a shortest distance from the corresponding internal image point to the region boundary of the initial region; calculating a sum value of the plurality of boundary interval distances corresponding to the target boundary point position to obtain a distance index corresponding to the target boundary point position; determining whether the target boundary point position is the final boundary point position according to the quantity index and the distance index corresponding to the target boundary point position.

[0044] In the above process, the quantity index is calculated to statistically obtain a proportion of initial boundary points corresponding to the target boundary point position in the plurality of to-be-processed images. The greater the quantity index, the higher the frequency of the target boundary point position indicating the initial boundary point in the plurality of to-be-processed images, that is, the higher the probability of the target boundary point position being a segmentation region boundary point position (i.e., a final boundary point position) in the unified segmentation standard.

[0045] The distance index is calculated to fully utilize data of internal image points corresponding to the target boundary point position in the plurality of to-be-processed images to inversely demonstrate the probability of the target boundary point position being a segmentation region boundary point position in the unified segmentation standard. The greater the distance index, the greater the magnitude of the target boundary point position deviating from the region boundary when indicating the internal image point in the plurality of to-be-processed images, that is, the lower the probability of the target boundary point position being a segmentation region boundary point position in the unified segmentation standard.

[0046] Based on the above setting, the data of the target boundary point in the corresponding multiple image points in the multiple to-be-processed images can be fully utilized, and whether the target boundary point can be determined as the final boundary point can be comprehensively and accurately evaluated.

[0047] Further, the determining whether the target boundary point is the final boundary point according to the quantity index and the distance index corresponding to the target boundary point comprises: calculating a ratio of the quantity index and the distance index corresponding to the target boundary point to obtain a boundary feature index of the target boundary point; in a case where the boundary feature index of the target boundary point is greater than or equal to a first threshold value, determining the target boundary point as the final boundary point; in a case where the boundary feature index of the target boundary point is less than the first threshold value, not determining the target boundary point as the final boundary point.

[0048] Exemplarily, the boundary feature index of the jthinitial boundary point among all initial boundary points corresponding to all initial boundary points included in the multiple to-be-processed images can be represented as: wherein, denotes the number of initial boundary points corresponding to the jthinitial boundary point in the multiple to-be-processed images, denotes the total number of the multiple to-be-processed images, denotes the boundary interval distance of the kthinner image point corresponding to the jthinitial boundary point in the multiple to-be-processed images, denotes the number of inner image points corresponding to the jthinitial boundary point in the multiple to-be-processed images, denotes the quantity index of the jthinitial boundary point, denotes the distance index of the jthinitial boundary point.

[0049] wherein, the above first threshold value can be adaptively set based on actual needs, for example, the first threshold value is set to 0.35.

[0050] Step S4, analyzing the pixel features of each target region position in each to-be-processed image to obtain multiple fusion weights corresponding to each target region position.

[0051] wherein, the multiple fusion weights corresponding to each target region position and the multiple to-be-processed images are in one-to-one correspondence.

[0052] ​​After the plurality of target region positions are determined, by further analyzing the pixel features of each target region position in each to-be-processed image, the clarity of the image region corresponding to each target region position in each to-be-processed image can be identified, and accordingly the fusion weight of the image region corresponding to each target region position in each to-be-processed image can be determined, so as to adaptively adjust the gray value of each image point based on the pixel features of the image region corresponding to different human tissues in each to-be-processed image, and higher fusion weight is assigned to the image region with higher clarity, and lower weight is assigned to the image region with relatively lower clarity, so as to ensure that the fused image has higher accuracy and clarity, and thus the fusion effect of the multi-angle CT digital images of the same human body part is improved.

[0053] It should be noted that the clarity in the present application should be understood as the degree of clarity of the image region representing the corresponding human tissue.

[0054] Specifically, the analysis of the pixel features of each target region position in each to-be-processed image obtains a plurality of fusion weights corresponding to each target region position, including: Similar analysis is performed on the image region indicated by each target region position in each to-be-processed image and the corresponding initial region, to obtain a plurality of region similarity indexes corresponding to each target region position, wherein the plurality of region similarity indexes correspond one-to-one to the plurality of to-be-processed images; Gray analysis is performed on the plurality of image points corresponding to each target region position in each to-be-processed image, to obtain a plurality of gray significant indexes corresponding to each target region position, wherein the plurality of gray significant indexes correspond one-to-one to the plurality of to-be-processed images; According to the plurality of region similarity indexes and the plurality of gray significant indexes corresponding to each target region position, a plurality of fusion weights corresponding to each target region position are obtained.

[0055] Specifically, the plurality of fusion weights corresponding to each target region position are obtained according to the plurality of region similarity indexes and the plurality of gray significant indexes corresponding to each target region position, including: The product of each region similarity index corresponding to each target region position and the corresponding gray significant index is calculated to obtain a plurality of fusion weights corresponding to each target region position.

[0056] The initial region corresponding to the image region indicated by the target region position in the to-be-processed image should be understood as the initial region with the largest area of overlap between the plurality of initial regions included in the corresponding to-be-processed image and the image region indicated by the corresponding target region position in the to-be-processed image.

[0057] In one example, the region similarity index corresponding to each target region position can be obtained by calculating the intersection-over-union between the image region indicated by each target region position in each to-be-processed image and the corresponding initial region.

[0058] In the above arrangement, the calculation of the region similarity index can evaluate the definition of the corresponding human tissue in the corresponding to-be-processed image from the dimension of region segmentation accuracy. The higher the definition of the corresponding human tissue in the corresponding to-be-processed image, the easier the corresponding human tissue is to be accurately segmented in the image segmentation process, and the higher the similarity between the segmented initial region and the image region indicated by the target region position. In other words, the higher the region similarity index, the higher the weight that should be given to the image region indicated by the target region position in the corresponding to-be-processed image.

[0059] The calculation of the gray-scale saliency index can evaluate the definition of the corresponding human tissue in the corresponding to-be-processed image from the dimension of image gray-scale distribution. The higher the definition of the corresponding human tissue in the corresponding to-be-processed image, the more regular the gray-scale distribution of the corresponding human tissue in the corresponding image region, and the higher the corresponding gray-scale saliency index. In other words, the higher the gray-scale saliency index, the higher the weight that should be given to the image region indicated by the target region position in the corresponding to-be-processed image.

[0060] Further, the similarity analysis of each target region position and the corresponding initial region to obtain the plurality of region similarity indexes corresponding to each target region position comprises: analyzing the region overlap degree between the image region indicated by each target region position in each to-be-processed image and the corresponding initial region to obtain a plurality of first similarity indexes of each target region position, wherein the plurality of first similarity indexes and the plurality of to-be-processed images correspond one-to-one; analyzing the boundary overlap degree between the image region indicated by each target region position in each to-be-processed image and the corresponding initial region to obtain a plurality of second similarity indexes of each target region position, wherein the plurality of second similarity indexes and the plurality of to-be-processed images correspond one-to-one; calculating the product of each first similarity index and the corresponding second similarity index of each target region position to obtain the plurality of region similarity indexes corresponding to each target region position.

[0061] In one example, the ratio of the area of the overlapping region between the image region indicated by each target region position in each to-be-processed image and the corresponding initial region to the sum of the areas of the two regions can be calculated as a quantitative representation of the degree of region overlap between the two regions. In addition, the proportion of the number of boundary points overlapping between the image region indicated by each target region position in each to-be-processed image and the corresponding initial region can be calculated as a quantitative representation of the degree of boundary overlap between the two regions. Exemplarily, the region similarity index corresponding to the xth target region position in the yth to-be-processed image may be represented as: ; wherein represents the area of the overlapping region between the image region corresponding to the xth target region position in the yth to-be-processed image and the corresponding initial region, the area of the initial region corresponding to the xth target region position in the yth to-be-processed image, represents the area of the image region corresponding to the xth target region position in the yth to-be-processed image, represents the number of boundary points overlapping between the image region corresponding to the xth target region position in the yth to-be-processed image and the corresponding initial region, represents the total number of boundary points of the image region corresponding to the xth target region position in the yth to-be-processed image, represents the first similarity index corresponding to the xth target region position in the yth to-be-processed image, represents the second similarity index corresponding to the xth target region position in the yth to-be-processed image.

[0062] Based on the above settings, the similarity between different regions is evaluated from two aspects of the region overlap area and the proportion of the number of overlapping boundary points, which can avoid errors that may be generated in single-dimensional evaluation and improve the accuracy of the finally obtained region similarity index.

[0063] Further, the gray scale analysis is performed on the plurality of image points corresponding to each target region position in each to-be-processed image to obtain a plurality of gray scale significant indices corresponding to each target region position, including: the dispersion degree of a plurality of gray scale values of the plurality of image points corresponding to each target region position in each to-be-processed image is analyzed to obtain a plurality of internal gray scale uniformity indices corresponding to each target region position, and the plurality of internal gray scale uniformity indices corresponding to each target region position correspond one-to-one to the plurality of to-be-processed images; analyze a difference between a gray scale distribution of each target region position in each to-be-processed image and a gray scale distribution of a neighboring region of the corresponding to-be-processed image, to obtain a plurality of external gray scale difference indexes corresponding to each target region position, the plurality of external gray scale difference indexes corresponding to each target region position corresponding to the plurality of to-be-processed images one by one; obtain a plurality of gray scale saliency indexes corresponding to each target region position according to the plurality of internal gray scale uniformity indexes and the plurality of external gray scale difference indexes corresponding to each target region position.

[0064] It should be noted that the higher the definition of the human tissue corresponding to the image region in the to-be-processed image, the higher the consistency of the gray scale values of each image point in the image region corresponding to the human tissue, and the greater the gray scale difference between the image region corresponding to the human tissue and other surrounding image regions.

[0065] Based on this, by obtaining the internal gray scale uniformity index and the external gray scale difference index, the gray scale saliency of each image region is accurately analyzed from the internal consistency and external difference of the image region, that is, the definition of the corresponding human tissue is accurately analyzed.

[0066] The external gray scale difference index can be the reciprocal of the gray scale value range of the plurality of internal image points included in the corresponding image region of the corresponding to-be-processed image of the target region position.

[0067] Exemplarily, the process of obtaining the plurality of gray scale saliency indexes corresponding to each target region position according to the plurality of internal gray scale uniformity indexes and the plurality of external gray scale difference indexes corresponding to each target region position can be: calculate the product of each internal gray scale uniformity index and the corresponding external gray scale difference index corresponding to each target region position, to obtain each gray scale saliency index corresponding to each target region position.

[0068] Further, the analysis of the difference between the gray scale distribution of each target region position in each to-be-processed image and the gray scale distribution of the neighboring region in the corresponding to-be-processed image, to obtain a plurality of external gray scale difference indexes corresponding to each target region position, includes: respectively calculate a distance between a corresponding centroid point of each target region position in each to-be-processed image and a plurality of centroid points of a plurality of neighboring regions in the corresponding to-be-processed image, to obtain a plurality of centroid distances corresponding to each target region position in each to-be-processed image; calculate the difference between the corresponding gray mean value of each target region position in each to-be-processed image and the gray mean values of a plurality of adjacent regions of the corresponding to-be-processed image, to obtain a plurality of gray mean value differences of each target region position in each to-be-processed image corresponding to the target region position; According to the plurality of centroid distances and the plurality of gray mean value differences of each target region position in each to-be-processed image, obtain the external gray difference index of each target region position in each to-be-processed image corresponding to the target region position.

[0069] The above centroid point should be understood as: the geometric center of the corresponding image region, also known as the centroid.

[0070] In the above setting, by analyzing the distance difference and the gray difference between the corresponding image region and its adjacent region, the determined external gray difference index can be more accurate and reliable.

[0071] It should be noted that the greater the gray difference between the corresponding image region and its adjacent region (i.e., the greater the gray mean value difference), the more obvious the pixel feature difference between the corresponding image region and its adjacent region, i.e., the higher the clarity of the corresponding image region representing the corresponding human tissue. Correspondingly, the smaller the distance difference between the corresponding image region and its adjacent region (i.e., the greater the centroid distance), the closer the corresponding image region and its adjacent region, and under the condition that the gray difference is unchanged, it can be explained that the gray difference between the corresponding image region and the adjacent region is more significant, and further that the clarity of the corresponding image region representing the corresponding human tissue is higher.

[0072] In one example, according to the plurality of centroid distances and the plurality of gray mean value differences of each target region position in each to-be-processed image, the process of obtaining the external gray difference index of each target region position in each to-be-processed image corresponding to the target region position can be: Calculate the ratio of each gray mean value difference and the corresponding centroid distance of each target region position in each to-be-processed image, to obtain a plurality of adjacent gray difference values of each target region position in each to-be-processed image corresponding to the target region position. Calculate the sum of a plurality of adjacent gray difference values of each target region position in each to-be-processed image corresponding to the target region position, to obtain the external gray difference index of each target region position in each to-be-processed image corresponding to the target region position.

[0073] Wherein, the adjacent region can be understood as other image regions adjacent to the corresponding image region.

[0074] Exemplarily, the e-th external gray difference index of the x-th target region position associated with the y-th to-be-processed image Can be expressed as: ; wherein, represents a mean value of a plurality of gray scale values of a plurality of internal image points of the image region corresponding to the xth target region position in the yth to-be-processed image, represents a mean value of a plurality of gray scale values of a plurality of internal image points of the e th adjacent region of the image region corresponding to the xth target region position in the yth to-be-processed image, represents a centroid distance between the image region corresponding to the xth target region position in the yth to-be-processed image and the e th adjacent region thereof.

[0075] Exemplarily, the gray scale significant index of the xth target region position in association with the yth to-be-processed image can be represented as: ; wherein, represents a range of gray scale values of a plurality of internal image points of the image region corresponding to the xth target region position in the yth to-be-processed image, represents the e th external gray scale difference index of the xth target region position in association with the yth to-be-processed image, represents a number of adjacent regions of the image region corresponding to the xth target region position in the yth to-be-processed image.

[0076] Step S5, fusing the plurality of to-be-processed images according to the plurality of fusion weights corresponding to each target region position to obtain a fused image.

[0077] Specifically, the fusing the plurality of to-be-processed images according to the plurality of fusion weights corresponding to each target region position to obtain a fused image comprises: calculating a sum value of the plurality of fusion weights corresponding to each target region position to obtain a total weight value corresponding to each target region position; calculating a ratio of the fusion weight corresponding to each target region position in each to-be-processed image to the total weight value corresponding thereto to obtain a corrected weight corresponding to each target region position in each to-be-processed image; calculating a product of the gray scale value of each image point in each to-be-processed image and the corrected weight corresponding to the target region position in which the image point is located to obtain a corrected gray scale value of each image point in each to-be-processed image; fusing the plurality of to-be-processed images based on the corrected gray scale value of each image point in each to-be-processed image to obtain the fused image.

[0078] ​Wherein, the number of image points included in the fusion image is same as the number of image points included in any to-be-processed image, and the gray value of each image point in the fusion image is obtained by summing up the corresponding multiple corrected gray values of the image point in multiple to-be-processed images.

[0079] It is to be explained that, for the same boundary line commonly corresponding to different target region positions, the corrected weight corresponding to the image point on the boundary line should be understood as the mean value of the different corrected weights of the different target region positions associated with the boundary line.

[0080] For example, the process of fusing the multiple to-be-processed images to obtain the fusion image can be as shown in Figure 2 , Figure 2 The four images on the left side in the above figure can be understood as the aforementioned multiple to-be-processed images, Figure 2 The image on the right side in the above figure is the fusion image obtained by fusing the four images on the left side based on the above method, and it is to be explained that, Figure 3 The images shown in the above figure are examples of images after blur processing, and do not involve privacy leakage of related users.

[0081] In general, after obtaining multiple CT digital images scanned from different angles of a target human body part, the present application first performs image rotation on the multiple CT digital images based on a standard angle, so as to eliminate the angle difference between different images, and obtain multiple to-be-processed images with unified angles; then performs image segmentation on each to-be-processed image, so as to segment multiple initial regions included in each to-be-processed image, and analyze the boundary features of each initial region, so as to comprehensively consider the image segmentation of different to-be-processed images, suppress the segmentation error caused by the angle difference of image scanning, and obtain multiple target region positions more accurately representing multiple human tissues in the target human body part; then further analyze the pixel features of different to-be-processed images in each target region position, so as to determine the fusion weight of each target region position in each to-be-processed image; finally, perform fusion operation on the multiple to-be-processed images according to the fusion weight, which can adaptively adjust the gray value of each image point based on the pixel features of the region corresponding to different human tissues in each to-be-processed image, assign higher weight to the region with more obvious image features, and assign lower weight to the region with relatively weak image features, so as to finally ensure that the image obtained by fusion has high accuracy and clarity, that is, to improve the fusion effect of multiple-angle CT digital images of the same human body part.

[0082] The present application provides a CT feature extraction device of a digital image system of a CT machine, please refer to Figure 4 , which shows a structure schematic diagram of a CT feature extraction device 300 of a digital image system of a CT machine according to an embodiment of the present application, and the device comprises: The image acquisition module 301 is configured to acquire a plurality of to-be-processed images, wherein the plurality of to-be-processed images are a plurality of images obtained by performing image rotation on a plurality of initial images according to standard angles, and the plurality of initial images are a plurality of CT digital images obtained by scanning a target human body part from different angles. The image segmentation module 302 is configured to perform image segmentation on each of the plurality of to-be-processed images, to obtain a plurality of initial regions included in each to-be-processed image. The boundary analysis module 303 is configured to perform boundary feature analysis on all initial regions included in the plurality of to-be-processed images, to obtain a plurality of target region positions, wherein the plurality of target region positions represent a plurality of human body tissues in the target human body part. The pixel analysis module 304 is configured to analyze pixel features of each target region position in each to-be-processed image, to obtain a plurality of fusion weights corresponding to each target region position, wherein the plurality of fusion weights corresponding to each target region position correspond to the plurality of to-be-processed images one by one. The image fusion module 305 is configured to perform fusion on the plurality of to-be-processed images according to the plurality of fusion weights corresponding to each target region position, to obtain a fusion image.

[0083] It should be noted that the apparatus provided in the above embodiments is only used as an example for the division of the above functional modules. In actual applications, the above functions can be completed by different functional modules according to needs, that is, the internal structure of the computer device is divided into different functional modules to complete all or part of the functions described above. In addition, the CT feature extraction apparatus for a digital image system of a CT machine and the CT feature extraction method for a digital image system of a CT machine provided in the above embodiments belong to the same concept, and the specific implementation process is described in the method embodiments, which will not be repeated here.

[0084] The embodiment of the present application further provides an electronic device. Please refer to Figure 1 The electronic device can include a processor 401, a memory 402, and a program 4021 stored in the memory 402 and executable on the processor 401.

[0085] When the program 4021 is executed by the processor 401, it can implement Figure 1 Any step in the corresponding method embodiment and achieve the same beneficial effects, which will not be repeated here.

[0086] Those skilled in the art can understand that all or part of the steps of the above-mentioned embodiment methods can be completed by program instructions related to hardware, and the program can be stored in a readable medium.

[0087] The embodiment of the present application further provides a readable storage medium, wherein the readable storage medium stores a computer program, and the computer program can realize the above-mentioned method when executed by a processor. ​ Any step in the corresponding method embodiment can be achieved, and the same technical effects can be achieved, to avoid repetition, which will not be described here.

[0088] The computer readable storage medium of the embodiment of the present application can adopt any combination of one or more computer readable media. The computer readable medium can be a computer readable signal medium or a computer readable storage medium. The computer readable storage medium may, for example, be an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or apparatus, or any combination thereof. More specific examples (non-exhaustive list) of the computer readable storage medium include: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In this document, the computer readable storage medium can be any tangible medium containing or storing a program that can be used by or in connection with an instruction execution system, device or apparatus.

[0089] The computer readable signal medium can include a data signal propagating in a baseband or as part of a carrier wave propagating through a transmission medium, in which the computer readable program code is embodied. Such a propagating data signal can take many forms, including but not limited to electro-magnetic, optical, or any suitable combination thereof. The computer readable signal medium can also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate or transport program for use by or in connection with an instruction execution system, apparatus or device.

[0090] The program code contained on the storage medium can be transmitted in any suitable medium, including but not limited to wireless, wire, optical cable, RF, etc., or any suitable combination thereof.

[0091] Computer program code for carrying out operations of the present application can be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C++ or the like, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider).

[0092] The embodiment of the present application further provides a computer program product, which, when running on a computer, enables the computer to execute the above related steps to realize the CT feature extraction method of the CT machine digital image system provided by the above embodiment.

[0093] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description, and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multi-task processing and parallel processing are also possible or can be advantageous.

[0094] Each of the embodiments in the specification is described in a progressive manner, and the same and similar parts between the embodiments can be referred to each other. Each embodiment focuses on the difference from other embodiments.

Claims

1. A method for CT feature extraction from a digital imaging system for CT scanners, characterized in that, The method includes: Acquire multiple images to be processed, wherein the multiple images to be processed are multiple images obtained by rotating multiple initial images according to a standard angle, and the multiple initial images are multiple CT digital images obtained by scanning target human body parts from different angles; The multiple images to be processed are segmented to obtain multiple initial regions included in each image to be processed; Boundary feature analysis is performed on all initial regions included in multiple images to be processed to obtain multiple target region locations, wherein the multiple target region locations represent the locations of multiple human tissues within the target human body part in the images to be processed; The pixel features of each target region location in each image to be processed are analyzed to obtain multiple fusion weights corresponding to each target region location. The multiple fusion weights corresponding to each target region location correspond one-to-one with the multiple images to be processed. Based on multiple fusion weights corresponding to the location of each target region, the multiple images to be processed are fused to obtain a fused image.

2. The CT feature extraction method for a digital imaging system for CT scanners according to claim 1, characterized in that, The boundary feature analysis of all initial regions included in the multiple images to be processed yields the locations of multiple target regions, including: The multiple initial boundary points corresponding to the multiple images to be processed are analyzed to determine the multiple final boundary point positions, wherein the initial boundary points are image points in the images to be processed located on the boundary of the initial region; Based on the multiple final boundary points, the locations of the multiple target areas are determined.

3. The CT feature extraction method for a digital imaging system for CT scanners according to claim 2, characterized in that, The analysis of multiple initial boundary points on the boundary of each initial region to determine multiple final boundary point locations includes: Calculate the ratio of the first quantity to the second quantity of the target boundary points to obtain the quantity index corresponding to the target boundary points, wherein the target boundary point represents the position of any one of the initial boundary points among all the initial boundary points included in the plurality of images to be processed, the first quantity represents the number of initial boundary points corresponding to the target boundary point in the plurality of images to be processed, and the second quantity represents the number of all image points corresponding to the target boundary point in the plurality of images to be processed. Obtain multiple boundary interval distances of multiple internal image points corresponding to the target boundary point in multiple images to be processed, wherein the internal image point represents an image point located inside the initial region, and the boundary interval distance is the shortest distance from the corresponding internal image point to the region boundary of the initial region; Calculate the sum of multiple boundary interval distances corresponding to the target boundary point to obtain the distance index corresponding to the target boundary point; Based on the quantity index and distance index corresponding to the target boundary point, determine whether the target boundary point is the final boundary point.

4. The CT feature extraction method for a digital imaging system for CT scanners according to claim 3, characterized in that, The step of determining whether the target boundary point is the final boundary point based on the quantity index and distance index corresponding to the target boundary point includes: Calculate the ratio of the quantity index to the distance index corresponding to the target boundary point to obtain the boundary feature index of the target boundary point; If the boundary feature index of the target boundary point is greater than or equal to the first threshold, the target boundary point is determined as the final boundary point. If the boundary feature index of the target boundary point is less than the first threshold, the target boundary point will not be determined as the final boundary point.

5. The CT feature extraction method for a digital imaging system for CT scanners according to claim 1, characterized in that, The step involves analyzing the pixel features of each target region location in each image to obtain multiple fusion weights corresponding to each target region location, including: For each target region location, a similarity analysis is performed on the image region indicated by each image to be processed and the corresponding initial region to obtain multiple region similarity indices corresponding to each target region location. The multiple region similarity indices correspond one-to-one with the multiple images to be processed. Gray-scale analysis is performed on multiple image points corresponding to each target region location in each image to be processed to obtain multiple gray-scale significance indices corresponding to each target region location, wherein the multiple gray-scale significance indices correspond one-to-one with the multiple images to be processed; Based on multiple regional similarity indices and multiple grayscale significance indices corresponding to each target region location, multiple fusion weights are obtained for each target region location.

6. The CT feature extraction method for a digital imaging system for CT scanners according to claim 5, characterized in that, The similarity analysis of each target region location with its corresponding initial region yields multiple region similarity indices for each target region location, including: The degree of overlap between the image region indicated by each target region location in each image to be processed and the corresponding initial region is analyzed to obtain multiple first similarity indices for each target region location, wherein the multiple first similarity indices correspond one-to-one with the multiple images to be processed; The degree of boundary overlap between the image region indicated by each target region location in each image to be processed and the corresponding initial region is analyzed to obtain multiple second similarity indices for each target region location, wherein the multiple second similarity indices correspond one-to-one with the multiple images to be processed; Calculate the product of each first similarity index and the corresponding second similarity index for each target region location to obtain multiple region similarity indices for each target region location.

7. The CT feature extraction method for a digital imaging system for CT scanners according to claim 5, characterized in that, The grayscale analysis of multiple image points corresponding to each target region location in each image to be processed is performed to obtain multiple grayscale significance indices corresponding to each target region location, including: The dispersion of multiple gray values ​​of multiple image points corresponding to each target region location in each image to be processed is analyzed to obtain multiple internal gray uniformity indices corresponding to each target region location. Each internal gray uniformity index corresponding to each target region location corresponds one-to-one with the multiple images to be processed. The difference between the gray-level distribution of each target region in each image to be processed and the gray-level distribution of its neighboring regions in the corresponding image to be processed is analyzed to obtain multiple external gray-level difference indices corresponding to each target region. Each of the multiple external gray-level difference indices corresponding to each target region corresponds one-to-one with the multiple images to be processed. Based on the multiple internal gray-scale uniformity indices and multiple external gray-scale difference indices corresponding to each target area location, multiple gray-scale saliency indices corresponding to each target area location are obtained.

8. The CT feature extraction method for a digital imaging system for CT scanners according to claim 7, characterized in that, The analysis compares the grayscale distribution of each target region location in each image to be processed with the grayscale distribution of its neighboring regions in the corresponding image to be processed, resulting in multiple external grayscale difference indices corresponding to each target region location, including: Calculate the distance between the centroid point corresponding to the location of each target region in each image to be processed and the centroid points of multiple adjacent regions in the corresponding image to be processed, so as to obtain the centroid distances corresponding to the location of each target region in each image to be processed. Calculate the difference between the mean gray value of each target region in each image to be processed and the mean gray values ​​of its multiple neighboring regions in the corresponding image to be processed, so as to obtain the difference of the mean gray values ​​of each target region in each image to be processed. Based on the multiple centroid distances and multiple gray-level mean differences corresponding to the location of each target region in each image to be processed, the external gray-level difference index corresponding to the location of each target region in each image to be processed is obtained.

9. The CT feature extraction method for a digital imaging system for CT scanners according to claim 5, characterized in that, The process involves obtaining multiple fusion weights for each target region location based on multiple region similarity indices and multiple grayscale significance indices, including: Calculate the product of the region similarity index and the corresponding gray-scale significance index for each target region location to obtain multiple fusion weights for each target region location.

10. The CT feature extraction method for a digital imaging system for CT scanners according to claim 1, characterized in that, The step of fusing the multiple images to be processed according to multiple fusion weights corresponding to the location of each target region to obtain a fused image includes: Calculate the sum of multiple fusion weights corresponding to each target region location to obtain the total weight value corresponding to each target region location; Calculate the ratio of the fusion weight corresponding to each target region location in each image to be processed to its corresponding total weight value, so as to obtain the correction weight corresponding to each target region location in each image to be processed; Calculate the product of the gray value of each image point in each image to be processed and the correction weight corresponding to its location in the target region, so as to obtain the correction gray value of each image point in each image to be processed. The multiple images to be processed are fused based on the corrected grayscale value of each image point in each image to be processed, to obtain the fused image.

Images