Image denoising method and device, computer device and storage medium

Abstract

The application relates to an image denoising method and device, computer equipment and a storage medium, wherein the image denoising method comprises the following steps: acquiring a sequence image of a plurality of frames of X-rays, updating a background image of two adjacent frames of the sequence images to obtain a background iteration image corresponding to a current frame of the sequence images; determining a first similarity of the two adjacent frames of the sequence images; determining a second similarity of the current frame of the sequence images and the background iteration image; and updating the current frame of the sequence images according to the first similarity and the second similarity to obtain a target frame image. Through the application, the problem that image background information loss caused by structure motion cannot be effectively avoided is solved, and the image denoising effect is improved.

Description

Technical Field

[0001] This application relates to the field of image processing, and in particular to image noise reduction methods, apparatus, computer equipment, and storage media. Background Technology

[0002] In digital X-ray fluoroscopy, X-rays are used to obtain a sequence of images by passing through the human body for disease diagnosis. However, excessive X-rays can be harmful to both patients and doctors. Therefore, while meeting the needs of use and diagnosis, it is necessary to minimize the X-ray dose. However, reducing the X-ray dose will introduce noise into the images, significantly degrading image quality. Therefore, after acquiring the sequence of images, image denoising algorithms are needed to process them to suppress the image noise and other effects caused by the reduced X-ray dose.

[0003] Current image denoising methods weight the pixels of the current frame image based on the similarity of adjacent frames to obtain an updated current frame image, and then iteratively process subsequent frames based on the temporal sequence of the images. However, in these methods, when the same structure in adjacent frames moves, the similarity between adjacent frames decreases, making the weighting indicators used for weighting inaccurate. This fails to effectively prevent the loss of image background information caused by structural motion, resulting in unsatisfactory image denoising effects.

[0004] There is currently no effective solution to the problem that related technologies cannot effectively avoid the loss of image background information caused by structural motion. Summary of the Invention

[0005] This embodiment provides an image noise reduction method, apparatus, computer device, and storage medium to address problems in related technologies.

[0006] Firstly, this embodiment provides an image denoising method, the method comprising:

[0007] Acquire a sequence of X-ray images of several frames, update the background image of the sequence of images of two adjacent frames, and obtain the background iterative image corresponding to the sequence of images of the current frame;

[0008] Determine the first similarity between two adjacent frames of the sequence of images;

[0009] Determine the second similarity between the sequence image of the current frame and the background iterative image;

[0010] Based on the first similarity and the second similarity, the sequence image of the current frame is updated to obtain the target frame image.

[0011] In some embodiments, before acquiring the sequence of X-ray images, updating the background image of two adjacent frames of the sequence of images, and obtaining the background iterative image corresponding to the sequence of images of the current frame, the method further includes:

[0012] Each frame of the image sequence is preprocessed, including image regularization and / or image mean filtering.

[0013] In some embodiments, updating the background image of the sequence of two adjacent frames to obtain an iterative background image corresponding to the sequence of images of the current frame includes:

[0014] Compare the grayscale values ​​of background pixels in two adjacent frames of the sequence image;

[0015] Based on the comparison results, the grayscale values ​​of the background pixels in the sequence image of the current frame are updated to obtain the background iterative image corresponding to the sequence image of the current frame.

[0016] In some embodiments, comparing the grayscale values ​​of background pixels in two adjacent frames of the image sequence includes:

[0017] By comparing the grayscale values ​​of background pixels at the same position in two adjacent frames of the sequence image, the larger grayscale value of each background pixel at each of the same positions is determined.

[0018] In some embodiments, comparing the grayscale values ​​of background pixels in two adjacent frames of the image sequence includes:

[0019] Based on the neighborhood pixel information of the background pixel in the sequence of images of two adjacent frames, the gray value of each background pixel is weighted to obtain the weighted gray value of the background pixel.

[0020] By comparing the weighted gray values ​​of the background pixels in two adjacent frames of the sequence image, the larger weighted gray value of each background pixel is determined.

[0021] In some embodiments, updating the sequence image of the current frame based on the first similarity and the second similarity to obtain the target frame image includes:

[0022] Based on the first similarity, a first similarity weight map corresponding to the sequence image of the current frame is generated;

[0023] Based on the second similarity, a second similarity weight map corresponding to the sequence image of the current frame is generated;

[0024] An iterative formula is constructed based on preset iteration coefficients, the first similarity weight map, and the second similarity weight map. The sequence image of the current frame is iteratively updated using the iterative formula to obtain the target frame image.

[0025] In some embodiments, the method further includes:

[0026] Perform multi-scale decomposition on the sequence of images in the current frame to generate an image pyramid corresponding to the sequence of images in the current frame;

[0027] In the image pyramid, background iteration processing is performed on each layer of the image;

[0028] Reconstruct the images of each layer after iteration to obtain the target frame image corresponding to the sequence image of the current frame.

[0029] Secondly, this embodiment provides an image noise reduction device, the device comprising:

[0030] The first update module acquires a sequence of X-ray images, updates the background image of the sequence images of two adjacent frames, and obtains a background iterative image corresponding to the sequence image of the current frame.

[0031] The first matching module determines the first similarity between two adjacent frames of the sequence images;

[0032] The second matching module determines the second similarity between the sequence image of the current frame and the background iterative image;

[0033] The second update module updates the sequence image of the current frame based on the first similarity and the second similarity to obtain the target frame image.

[0034] Thirdly, this embodiment provides a computer device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the image noise reduction method described in the first aspect above.

[0035] Fourthly, this embodiment provides a storage medium storing a computer program that, when executed by a processor, implements the image noise reduction method described in the first aspect above.

[0036] Compared with related technologies, the image denoising method, apparatus, computer equipment, and storage medium provided in this embodiment acquire a sequence of X-ray images of several frames, update the background image of the sequence images of two adjacent frames to obtain a background iterative image corresponding to the sequence image of the current frame; determine the first similarity of the sequence images of two adjacent frames; determine the second similarity between the sequence image of the current frame and the background iterative image; and update the sequence image of the current frame according to the first and second similarities to obtain the target frame image. This solves the problem of the inability to effectively avoid the loss of image background information caused by structural motion and achieves improved image denoising effect.

[0037] Details of one or more embodiments of this application are set forth in the following drawings and description to make other features, objects and advantages of this application more readily apparent. Attached Figure Description

[0038] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0039] Figure 1 This is a hardware structure block diagram of a terminal device for an image noise reduction method provided in an embodiment of this application;

[0040] Figure 2 This is a flowchart of an image noise reduction method provided in an embodiment of this application;

[0041] Figure 3 This is a flowchart of an image noise reduction method provided in a preferred embodiment of this application;

[0042] Figure 4 This is a structural block diagram of an image noise reduction device provided in an embodiment of this application.

[0043] In the diagram: 102, processor; 104, memory; 106, transmission device; 108, input / output device; 10, first update module; 20, first matching module; 30, second matching module; 40, second update module. Detailed Implementation

[0044] To better understand the purpose, technical solution, and advantages of this application, the application is described and illustrated below in conjunction with the accompanying drawings and embodiments.

[0045] Unless otherwise defined, the technical or scientific terms used in this application shall have the general meaning as understood by one of ordinary skill in the art to which this application pertains. Words such as “a,” “an,” “an,” “the,” “the,” and “these,” used in this application, do not indicate quantitative limitation and may be singular or plural. The terms “comprising,” “including,” “having,” and any variations thereof used in this application are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or device that comprises a series of steps or modules (units) is not limited to the listed steps or modules (units) but may include steps or modules (units) not listed, or may include other steps or modules (units) inherent to such processes, methods, products, or devices. The terms “connected,” “linked,” and “coupled,” used in this application, are not limited to physical or mechanical connections but may include electrical connections, whether direct or indirect. The term “multiple” used in this application refers to two or more. The "and / or" operator describes the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: A alone, A and B simultaneously, and B alone. Typically, the character " / " indicates that the objects before and after it are in an "or" relationship. The terms "first," "second," and "third," etc., used in this application are merely for distinguishing similar objects and do not represent a specific ordering of the objects.

[0046] The method embodiments provided in this example can be executed on a terminal, computer, or similar computing device. For example, it can run on a terminal. Figure 1 This is a hardware structure block diagram of the terminal for the image noise reduction method in this embodiment. For example... Figure 1 As shown, a terminal may include one or more ( Figure 1 Only one is shown in the diagram. A processor 102 and a memory 104 for storing data are also included. The processor 102 may be, but is not limited to, a microprocessor (MCU) or a programmable logic device (FPGA). The terminal may also include a transmission device 106 for communication functions and an input / output device 108. Those skilled in the art will understand that… Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the terminal described above. For example, the terminal may also include components that are larger than... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown are illustrated.

[0047] The memory 104 can be used to store computer programs, such as application software programs and modules, like the computer program corresponding to the image noise reduction method in this embodiment. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, thereby implementing the above-described method. The memory 104 may include high-speed random access memory and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0048] The transmission device 106 is used to receive or send data via a network. This network includes a wireless network provided by the terminal's communication provider. In one example, the transmission device 106 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device 106 can be a Radio Frequency (RF) module used for wireless communication with the Internet.

[0049] This embodiment provides an image noise reduction method. Figure 2 This is a flowchart of the image noise reduction method in this embodiment, as follows: Figure 2 As shown, the process includes the following steps:

[0050] Step S210: Acquire a sequence of X-ray images of several frames, update the background image of the sequence images of two adjacent frames, and obtain the background iterative image corresponding to the sequence image of the current frame.

[0051] Specifically, for a sequence of two adjacent frames, the grayscale value of each background pixel in the previous frame is compared with the grayscale value of the corresponding pixel in the current frame. Based on the comparison result, the grayscale values ​​of the background pixels in the current frame are updated to obtain an iterative background image corresponding to the current frame.

[0052] Step S220: Determine the first similarity between two adjacent frames of sequential images.

[0053] Specifically, similarity calculation is performed on the sequence of images of two adjacent frames, that is, based on the difference in gray values, the first similarity between corresponding pixels in the two frames is calculated.

[0054] Step S230: Determine the second similarity between the sequence image of the current frame and the background iterative image.

[0055] It's important to know that the calculation method for the first similarity and the second similarity should be chosen based on actual needs. These similarity calculation methods include, but are not limited to, structural similarity, histogram similarity, and block matching similarity.

[0056] Step S240: Update the sequence image of the current frame according to the first similarity and the second similarity to obtain the target frame image.

[0057] Specifically, based on the first similarity, a first similarity weight map is generated corresponding to the sequence of images of the current frame, and based on the second similarity, a second similarity weight map is generated corresponding to the sequence of images of the current frame, so as to reflect the similarity weight of different regions in the image.

[0058] Furthermore, an iterative formula is constructed based on preset iteration coefficients, a first similarity weight map, and a second similarity weight map. The sequence image of the current frame is iteratively updated using the iterative formula to obtain the target frame image, thereby achieving adaptive weight iterative update.

[0059] Current image denoising methods weight the pixels of the current frame image based on the similarity of adjacent frames to obtain an updated current frame image, and then iteratively process subsequent frames based on the temporal sequence of the images. However, in these methods, when the same structure in adjacent frames moves, the similarity between adjacent frames decreases, making the weighting indicators used for weighting inaccurate. This fails to effectively prevent the loss of image background information caused by structural motion, resulting in unsatisfactory image denoising effects. This application, building upon existing technologies, iteratively accumulates background grayscale information in the image using background iteration images according to the temporal sequence of the images. It updates the current frame image based on the similarity between adjacent frames and between the current frame image and the background iteration images, thereby avoiding the loss of background information during the denoising process and preventing unsatisfactory denoising effects in some image areas.

[0060] This embodiment acquires a sequence of X-ray images of several frames, updates the background image of two adjacent frames to obtain a background iterative image corresponding to the current frame's sequence image, determines the first similarity between the two adjacent frames' sequence images, determines the second similarity between the current frame's sequence image and the background iterative image, and updates the current frame's sequence image based on the first and second similarities to obtain the target frame image. This solves the problem of not being able to effectively avoid the loss of image background information caused by structural motion, and improves the image noise reduction effect.

[0061] In some embodiments, before acquiring the sequence of X-ray images, updating the background image of two adjacent frames, and obtaining the background iterative image corresponding to the sequence image of the current frame, the following steps are also included:

[0062] Step S201: Preprocess each frame of the sequence image, including image regularization and / or image mean filtering.

[0063] It's important to understand that X-ray dose directly affects the grayscale values ​​of sequential images. During X-ray imaging, some X-rays penetrate the body and are received by a detector. The detector converts these rays into electrical signals, which are then transmitted to a computer for processing to obtain the final X-ray image. Therefore, if the X-ray dose increases, the signal received by the detector will be stronger, and the grayscale values ​​of the sequential image will also increase accordingly.

[0064] Specifically, since different X-ray doses cause differences in the grayscale values ​​of the sequence images, regularization processing is required to correct the grayscale values ​​of the sequence images. This regularization processing includes, but is not limited to, grayscale stretching, histogram equalization, logarithmic transformation, and gamma correction.

[0065] Furthermore, to remove some noise from the images beforehand, mean filtering or other filtering processes can be applied to the image sequence. For example, median filtering and Gaussian filtering.

[0066] In this embodiment, before obtaining the background iterative image corresponding to the sequence image of the current frame, the sequence image is preprocessed to correct the grayscale value of the sequence image and remove some image noise, thereby improving the accuracy of subsequent image processing results.

[0067] In some embodiments, background image updates are performed on a sequence of images between two adjacent frames to obtain an iterative background image corresponding to the sequence of images in the current frame, including the following steps:

[0068] Step S211: Compare the gray values ​​of background pixels in two adjacent frames of the sequence image;

[0069] Step S212: Based on the comparison result, update the grayscale value of the background pixels in the sequence image of the current frame to obtain the background iterative image corresponding to the sequence image of the current frame.

[0070] Specifically, a sequence of images from two adjacent frames is acquired, and the background region of the previous frame is determined. The grayscale value of each pixel in the background region is compared with the grayscale value of the corresponding pixel in the current frame's sequence image. Based on the comparison result, the grayscale values ​​of the background pixels in the current frame's sequence image are updated to obtain the corresponding iterative background image.

[0071] It is important to know that when comparing the gray values ​​of background pixels in two adjacent frames of sequential images, the gray value of each pixel can be weighted based on the neighboring pixel information of each pixel in the background region, and the weighted pixel gray values ​​in the two frames of sequential images can be compared accordingly.

[0072] In this embodiment, the gray values ​​of background pixels in the sequence images of two adjacent frames are compared, and the gray values ​​of background pixels in the sequence image of the current frame are updated based on the comparison results, so as to obtain the background iterative image corresponding to the sequence image of the current frame, thereby accumulating the background gray value information in the sequence image and effectively avoiding the loss of image background information caused by structural motion.

[0073] In some embodiments, comparing the grayscale values ​​of background pixels in two adjacent frames of a sequence of images includes the following steps:

[0074] Compare the gray values ​​of background pixels at the same position in two adjacent frames of the image sequence, and determine the larger gray value of each background pixel at each same position.

[0075] Specifically, the gray value of each pixel in the background region of the previous frame sequence image is compared with the gray value of the corresponding pixel in the current frame sequence image. The larger gray value is then used as the gray value of the background pixel in the current frame sequence image.

[0076] It's important to understand that in X-ray image sequences, high-attenuation regions include information about human skeletons, interventional devices such as guidewires or catheters, and the grayscale values ​​of pixels in these regions are higher than those in low-attenuation regions. Therefore, when the main observation area of ​​an image is a high-attenuation region, the low-attenuation region essentially becomes the background. In this case, a larger grayscale value needs to be selected to update the grayscale values ​​of the background region in the current frame sequence to ensure effective accumulation of background grayscale information.

[0077] In this embodiment, the gray values ​​of background pixels at the same position in two adjacent frames of sequential images are compared to determine the larger gray value of each background pixel at each same position. Thus, the background region of the current frame sequence image can be updated based on the larger gray value obtained from the comparison, thereby realizing the accumulation of background gray value information.

[0078] In some embodiments, comparing the grayscale values ​​of background pixels in two adjacent frames of a sequence of images includes the following steps:

[0079] Based on the neighborhood pixel information of background pixels in two adjacent frames of sequential images, the gray value of each background pixel is weighted to obtain the weighted gray value of the background pixel.

[0080] Compare the weighted gray values ​​of background pixels in two adjacent frames of the image sequence to determine the larger weighted gray value for each background pixel.

[0081] Specifically, based on the neighboring pixel information of each pixel, the gray value of the pixel is weighted and calculated. The weighted gray values ​​of the pixels in the two frame sequence images are compared to obtain the larger weighted gray value, which is then used as the gray value of the corresponding pixel in the current frame image.

[0082] It's important to understand that the grayscale value of each pixel is typically influenced by its surrounding pixels. By combining information from neighboring pixels to calculate the grayscale value of background pixels, we can more accurately evaluate the grayscale value of each pixel using information from its surrounding pixels. Furthermore, weighted calculations using neighboring pixels can reduce noise or discontinuities in the image, achieving a smoothing effect.

[0083] In this embodiment, based on the neighborhood pixel information of background pixels in two adjacent frames of sequential images, the gray value of each background pixel is weighted to obtain the weighted gray value of the background pixel. The weighted gray values ​​of background pixels in two adjacent frames of sequential images are compared to determine the larger weighted gray value of each background pixel. Thus, the background region of the current frame of sequential images can be updated based on the larger gray value obtained by comparison, effectively accumulating background information.

[0084] In some embodiments, the sequence image of the current frame is updated based on a first similarity and a second similarity to obtain the target frame image, including the following steps:

[0085] Step S241: Based on the first similarity, generate a first similarity weight map corresponding to the sequence images of the current frame;

[0086] Step S242: Based on the second similarity, generate a second similarity weight map corresponding to the sequence images of the current frame;

[0087] Step S243: Construct an iterative formula based on preset iteration coefficients, a first similarity weight map, and a second similarity weight map, and iteratively update the sequence image of the current frame using the iterative formula to obtain the target frame image.

[0088] Specifically, the similarity of two adjacent frames of images is calculated to obtain the first similarity between the gray values ​​of corresponding pixels in the two frames. Based on the first similarity, a first similarity weight map corresponding to the current frame of images is generated. In addition, the similarity of the current frame of images and the background iterative image is calculated to obtain the second similarity between the gray values ​​of corresponding pixels in the two frames. A corresponding second similarity weight map is then generated.

[0089] Furthermore, an iterative formula is constructed based on preset iteration coefficients, a first similarity weight map, and a second similarity weight map. This iterative formula is then used to iteratively update the sequence of images in the current frame to obtain the target frame image. In this embodiment, the specific iterative formula is as follows:

[0090] N1'=k×(N1×(1-W1 α )+P1×W1 α )+(1-k)×(N1×(1-W2 β )+B1×W2 β );

[0091] Where N1′ represents the updated current frame sequence image, which serves as the previous frame sequence image in the next iteration; N1 represents the sequence image of the current frame in two adjacent frames; P1 represents the sequence image of the previous frame in two adjacent frames; B1 represents the background iteration image corresponding to the current frame sequence image; W1 represents the first similarity weight map; W2 represents the second similarity weight map; α, β, and k are all preset iteration coefficients, and the nonlinear coefficients α and β are both greater than 1, while the linear coefficient k ranges from (0,1), allowing for flexible parameter tuning.

[0092] It's important to know that the calculation method for the first and second similarities should be selected according to the actual needs. Similarity calculation methods include, but are not limited to, structural similarity, histogram similarity, and block matching similarity. For example, calculating block matching similarity for a sequence of two adjacent frames involves matching the similarity of each pixel in the previous frame with the corresponding pixel in the current frame. This includes: in both frames, using the pixel at the same position as the center pixel of a 3x3 rectangle, and dividing the neighborhood pixels corresponding to the center pixel; within the two rectangles, calculating the absolute difference in grayscale values ​​between corresponding pixels; and then taking the average of these absolute differences as the first similarity for that pixel position.

[0093] In this embodiment, after determining the first similarity between two adjacent image frames, the calculated similarities are mapped to the corresponding pixel positions and recorded to obtain a similarity template. The similarity template is then normalized to obtain the final first similarity weight map, which reflects the similarity weights of different regions in the image. Alternatively, after determining the second similarity between the current frame's sequence image and the background iterative image, the above method can also be used to generate a second similarity weight map.

[0094] In this embodiment, a first similarity weight map corresponding to the sequence image of the current frame is generated based on the first similarity; a second similarity weight map corresponding to the sequence image of the current frame is generated based on the second similarity; thereby, an iterative formula can be constructed according to the preset iteration coefficients, the first similarity weight map and the second similarity weight map, and the sequence image of the current frame can be iteratively updated through the iterative formula to obtain the target frame image, avoiding the loss of image background information caused by structural motion, eliminating motion artifacts and improving the image noise reduction effect.

[0095] In some embodiments, the image denoising method further includes the following steps:

[0096] Perform multi-scale decomposition on the sequence of images in the current frame to generate an image pyramid corresponding to the sequence of images in the current frame;

[0097] In the image pyramid, background iterative processing is performed on each layer of the image;

[0098] Reconstruct the images of each layer after iteration to obtain the target frame image corresponding to the sequence image of the current frame.

[0099] Specifically, an image pyramid is a technique for representing and processing images at different scales. It decomposes the sequence of images of the current frame into multiple scales to generate a corresponding image pyramid containing multiple image levels with different resolutions. Using the same noise reduction method as when processing two adjacent frames, it iteratively processes the background of each image level in turn. Then, it reconstructs the iteratively updated images of each level to obtain the target frame image corresponding to the sequence of images of the current frame.

[0100] It is important to know that, depending on the actual application, different noise reduction parameters and convolution kernels of different sizes can be selected for different image layers to match the scale of structural information in different image layers.

[0101] In this embodiment, the sequence of images of the current frame is decomposed into multiple scales to generate an image pyramid corresponding to the sequence of images of the current frame. In the image pyramid, background iterative processing is performed on each layer of the image, and the iterated images of each layer are reconstructed to obtain the target frame image corresponding to the sequence of images of the current frame. In this way, the multi-scale representation of the image pyramid is used to perform noise reduction processing on the image at different scales, which more comprehensively reduces image noise and preserves the details and structural information of the image.

[0102] The present embodiment will now be described and illustrated through preferred embodiments.

[0103] Figure 3 This is a flowchart of the image noise reduction method of this preferred embodiment, as shown below. Figure 3 As shown, the image denoising method includes the following steps:

[0104] Step S310: Acquire a sequence of X-ray images of several frames;

[0105] Step S320: Compare the gray values ​​of background pixels in two adjacent frames of the sequence image;

[0106] Step S330: Based on the comparison result, update the grayscale value of the background pixels in the sequence image of the current frame to obtain the background iterative image corresponding to the sequence image of the current frame;

[0107] Step S340: Determine the first similarity between the sequence images of two adjacent frames, and generate the first similarity weight map corresponding to the sequence image of the current frame based on the first similarity.

[0108] Step S350: Determine the second similarity between the sequence image of the current frame and the background iterative image, and generate a second similarity weight map corresponding to the sequence image of the current frame based on the second similarity.

[0109] Step S360: Construct an iterative formula based on preset iteration coefficients, a first similarity weight map, and a second similarity weight map, and iteratively update the sequence image of the current frame using the iterative formula to obtain the target frame image.

[0110] This embodiment acquires a sequence of X-ray images of several frames, compares the grayscale values ​​of background pixels in adjacent frames, and updates the grayscale values ​​of background pixels in the current frame based on the comparison results, thereby obtaining an iterative background image corresponding to the current frame and accumulating background grayscale information. It determines the first similarity between adjacent frames and generates a first similarity weight map corresponding to the current frame. It also determines the second similarity between the current frame and the iterative background image and generates a second similarity weight map. Furthermore, it constructs an iterative formula based on preset iteration coefficients, the first similarity weight map, and the second similarity weight map, and iteratively updates the current frame's image using this formula to obtain the target frame image. This solves the problem of image background information loss due to structural motion and improves image noise reduction.

[0111] It should be noted that the steps shown in the above process or in the flowchart of the accompanying figures can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0112] This embodiment also provides an image noise reduction device for implementing the above embodiments and preferred embodiments; details already described will not be repeated. The terms "module," "unit," "subunit," etc., used below refer to combinations of software and / or hardware that perform predetermined functions. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0113] Figure 4 This is a structural block diagram of the image noise reduction device in this embodiment, as shown below. Figure 4 As shown, the device includes: a first update module 10, a first matching module 20, a second matching module 30, and a second update module 40;

[0114] The first update module 10 acquires the sequence of X-ray images, updates the background image of the sequence images of two adjacent frames, and obtains the background iterative image corresponding to the sequence image of the current frame.

[0115] The first matching module 20 determines the first similarity between two adjacent frames of sequential images;

[0116] The second matching module 30 determines the second similarity between the sequence image of the current frame and the background iterative image;

[0117] The second update module 40 updates the sequence image of the current frame based on the first similarity and the second similarity to obtain the target frame image.

[0118] The apparatus provided in this embodiment acquires a sequence of X-ray images of several frames, updates the background image of two adjacent frames to obtain a background iterative image corresponding to the current frame's sequence image, determines the first similarity between the two adjacent frames' sequence images, determines the second similarity between the current frame's sequence image and the background iterative image, and updates the current frame's sequence image based on the first and second similarities to obtain the target frame image. This solves the problem of not being able to effectively avoid the loss of image background information caused by structural motion and improves the image noise reduction effect.

[0119] In some of these embodiments, in Figure 4 Based on this, the device also includes a preprocessing module for preprocessing each frame of the image sequence, including image regularization and / or image mean filtering.

[0120] In some of these embodiments, in Figure 4 Based on this, the device also includes a first comparison module, which is used to compare the gray values ​​of background pixels in the sequence images of two adjacent frames; based on the comparison result, the gray values ​​of background pixels in the sequence image of the current frame are updated to obtain a background iterative image corresponding to the sequence image of the current frame.

[0121] In some of these embodiments, in Figure 4 Based on this, the device also includes a second comparison module, which is used to compare the gray values ​​of background pixels at the same position in two adjacent frames of sequential images, and determine the larger gray value of each background pixel at each same position.

[0122] In some of these embodiments, in Figure 4 Based on this, the device also includes a computing module, which is used to perform weighted calculation on the gray value of each background pixel based on the neighborhood pixel information of the background pixel in the sequence of two adjacent frames of images, to obtain the weighted gray value of the background pixel; and compare the weighted gray values ​​of the background pixels in the sequence of two adjacent frames of images to determine the larger weighted gray value of each background pixel.

[0123] In some of these embodiments, in Figure 4 Based on this, the device also includes an iteration module, which is used to generate a first similarity weight map corresponding to the sequence image of the current frame based on the first similarity; generate a second similarity weight map corresponding to the sequence image of the current frame based on the second similarity; construct an iteration formula according to the preset iteration coefficients, the first similarity weight map and the second similarity weight map, and iteratively update the sequence image of the current frame through the iteration formula to obtain the target frame image.

[0124] In some of these embodiments, in Figure 4 Based on this, the device also includes a reconstruction module, which is used to perform multi-scale decomposition on the sequence images of the current frame to generate an image pyramid corresponding to the sequence images of the current frame; in the image pyramid, background iterative processing is performed on each layer of the image; and the iteratively reconstructed images of each layer are used to obtain the target frame image corresponding to the sequence images of the current frame.

[0125] It should be noted that the above modules can be functional modules or program modules, and can be implemented through software or hardware. For modules implemented through hardware, the above modules can reside in the same processor; or the above modules can be located in different processors in any combination.

[0126] This embodiment also provides a computer device, including a memory and a processor, wherein the memory stores a computer program and the processor is configured to run the computer program to perform the steps in any of the above method embodiments.

[0127] Optionally, the computer device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor and the input / output device is connected to the processor.

[0128] It should be noted that the specific examples in this embodiment can refer to the examples described in the above embodiments and optional implementations, and will not be repeated in this embodiment.

[0129] Furthermore, in conjunction with the image denoising methods provided in the above embodiments, this embodiment can also provide a storage medium for implementation. This storage medium stores a computer program; when executed by a processor, the computer program implements any of the image denoising methods described in the above embodiments.

[0130] It should be understood that the specific embodiments described herein are merely illustrative of the application and not intended to limit it. All other embodiments derived by those skilled in the art based on the embodiments provided in this application without inventive effort are within the scope of protection of this application.

[0131] Obviously, the accompanying drawings are merely some examples or embodiments of this application. Those skilled in the art can apply this application to other similar situations based on these drawings without any creative effort. Furthermore, it is understood that although the work done in this development process may be complex and lengthy, for those skilled in the art, certain design, manufacturing, or production modifications made based on the technical content disclosed in this application are merely conventional technical means and should not be considered as insufficient disclosure of this application.

[0132] The term "embodiment" in this application refers to a specific feature, structure, or characteristic described in connection with an embodiment that may be included in at least one embodiment of this application. The appearance of this phrase in various places in the specification does not necessarily imply the same embodiment, nor does it imply that it is mutually exclusive with or independent of other embodiments. It will be clearly or implicitly understood by those skilled in the art that the embodiments described in this application may be combined with other embodiments without conflict.

[0133] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of patent protection. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the appended claims.

Claims

1. An image denoising method, characterized in that, The method includes: Acquire a sequence of X-ray images of several frames, update the background image of the sequence of images of two adjacent frames, and obtain the background iterative image corresponding to the sequence of images of the current frame; The step of updating the background image of the sequence images of two adjacent frames to obtain a background iterative image corresponding to the sequence image of the current frame includes: comparing the gray values ​​of background pixels in the sequence images of two adjacent frames; and updating the gray values ​​of background pixels in the sequence image of the current frame based on the comparison result to obtain a background iterative image corresponding to the sequence image of the current frame. Determine the first similarity between two adjacent frames of the sequence of images; Determine the second similarity between the sequence image of the current frame and the background iterative image; Based on the first similarity and the second similarity, the sequence image of the current frame is updated to obtain the target frame image; The step of updating the sequence image of the current frame based on the first similarity and the second similarity to obtain the target frame image includes: generating a first similarity weight map corresponding to the sequence image of the current frame based on the first similarity; generating a second similarity weight map corresponding to the sequence image of the current frame based on the second similarity; constructing an iterative formula based on preset iteration coefficients, the first similarity weight map and the second similarity weight map, and iteratively updating the sequence image of the current frame through the iterative formula to obtain the target frame image.

2. The image denoising method according to claim 1, characterized in that, Before acquiring a sequence of X-ray images of several frames, updating the background image of two adjacent frames of the sequence of images, and obtaining an iterative background image corresponding to the sequence of images of the current frame, the method further includes: Each frame of the image sequence is preprocessed, including image regularization and / or image mean filtering.

3. The image denoising method according to claim 1, characterized in that, The step of comparing the grayscale values ​​of background pixels in two adjacent frames of the sequence image includes: By comparing the grayscale values ​​of background pixels at the same position in two adjacent frames of the sequence image, the larger grayscale value of each background pixel at each of the same positions is determined.

4. The image denoising method according to claim 1, characterized in that, The step of comparing the grayscale values ​​of background pixels in two adjacent frames of the sequence image includes: Based on the neighborhood pixel information of the background pixel in the sequence of images of two adjacent frames, the gray value of each background pixel is weighted to obtain the weighted gray value of the background pixel. By comparing the weighted gray values ​​of the background pixels in two adjacent frames of the sequence image, the larger weighted gray value of each background pixel is determined.

5. The image denoising method according to claim 1, characterized in that, The method further includes: Perform multi-scale decomposition on the sequence of images in the current frame to generate an image pyramid corresponding to the sequence of images in the current frame; In the image pyramid, background iteration processing is performed on each layer of the image; Reconstruct the images of each layer after iteration to obtain the target frame image corresponding to the sequence image of the current frame.

6. An image noise reduction device, characterized in that, The device includes: The first update module acquires a sequence of X-ray images, updates the background image of the sequence images of two adjacent frames, and obtains a background iterative image corresponding to the sequence image of the current frame. The first update module is further configured to compare the gray values ​​of background pixels in the sequence images of two adjacent frames; based on the comparison result, update the gray values ​​of the background pixels in the sequence image of the current frame to obtain a background iterative image corresponding to the sequence image of the current frame; The first matching module determines the first similarity between two adjacent frames of the sequence images; The second matching module determines the second similarity between the sequence image of the current frame and the background iterative image; The second update module updates the sequence image of the current frame based on the first similarity and the second similarity to obtain the target frame image; The second update module is used to generate a first similarity weight map corresponding to the sequence image of the current frame based on the first similarity; generate a second similarity weight map corresponding to the sequence image of the current frame based on the second similarity; construct an iterative formula according to preset iteration coefficients, the first similarity weight map and the second similarity weight map, and iteratively update the sequence image of the current frame through the iterative formula to obtain the target frame image.

7. A computer device, comprising a memory and a processor, characterized in that, The memory stores a computer program, and the processor is configured to run the computer program to perform the steps of the image denoising method according to any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the image noise reduction method according to any one of claims 1 to 5.

Images