CN112529804A

CN112529804A - Image denoising method, device and equipment and computer readable storage medium

Info

Publication number: CN112529804A
Application number: CN202011467107.9A
Authority: CN
Inventors: 黄文政; 唐艳红; 刘旭江; 颜应华; 解家月
Original assignee: Sonoscape Medical Corp
Current assignee: Sonoscape Medical Corp
Priority date: 2020-12-14
Filing date: 2020-12-14
Publication date: 2021-03-19

Abstract

The application discloses an image denoising method, which comprises the steps of carrying out four-dimensional contrast time imaging on a four-dimensional ultrasonic image sequence to be processed to obtain a four-dimensional contrast image sequence, wherein each four-dimensional contrast image in the four-dimensional contrast image sequence is a color-based time contrast image; determining a target four-dimensional contrast image where noise is located from the four-dimensional contrast image sequence, and determining a noise color in the target four-dimensional contrast image and a noise clipping region where the noise color is located; constructing a color cutting template by using the noise colors and the noise cutting area; denoising the four-dimensional contrast image sequence by using the color cutting template; the image denoising method can realize accurate segmentation of noise information and organization information in image data, and effectively improves the accuracy of image denoising. The application also discloses an image denoising device, equipment and a computer readable storage medium, which have the beneficial effects.

Description

Image denoising method, device and equipment and computer readable storage medium

Technical Field

The present application relates to the field of image processing technologies, and in particular, to an image denoising method, and further, to an image denoising apparatus, device, and computer-readable storage medium.

Background

With the rapid development of scientific technology, image processing has been widely applied in many fields such as traffic monitoring, medical diagnosis, and the like. The image denoising is an important component of an image processing technology, and aims to remove noise information in an image so as to make the image clearer.

In the related art, an image denoising method based on spatial information is mostly adopted, but it is difficult to rapidly and accurately remove noise information by editing an image based on spatial information only, for example, when the noise information in the image is adjacent to or intersected with tissue information, since it is difficult to outline the tissue information, part of the tissue information is removed while the noise information is removed, thereby causing inaccuracy of image denoising.

Therefore, how to accurately segment the noise information and the tissue information in the image data and further improve the accuracy of image denoising is a problem to be solved urgently by those skilled in the art.

Disclosure of Invention

The image denoising method can realize accurate segmentation of noise information and tissue information in image data, and effectively improves the accuracy of image denoising; another object of the present application is to provide an image denoising device, an image denoising apparatus, and a computer readable storage medium, all of which have the above advantages.

In a first aspect, the present application provides an image denoising method, including:

carrying out four-dimensional contrast time imaging on a four-dimensional ultrasonic image sequence to be processed to obtain a four-dimensional contrast image sequence, wherein each four-dimensional contrast image in the four-dimensional contrast image sequence is a color-based time contrast image;

determining a target four-dimensional contrast image where noise is located from the four-dimensional contrast image sequence, and determining a noise color in the target four-dimensional contrast image and a noise clipping region where the noise color is located;

constructing a color cutting template by using the noise colors and the noise cutting area;

and denoising the four-dimensional contrast image sequence by using the color cutting template.

Preferably, the constructing a color clipping template by using the noise color and the noise clipping region includes:

generating a three-dimensional cutting area according to an area corresponding to the noise color in the noise cutting area;

calculating an intersection region of the three-dimensional cutting region and the target four-dimensional contrast image;

and generating a color cutting template corresponding to the noise color in the intersection area.

Preferably, the generating a color clipping template corresponding to the noise color in the intersection region includes:

acquiring the RGB value of each pixel in the intersection area;

and obtaining noise pixel points corresponding to the cutting colors in the intersection area according to the RGB value of each pixel, and generating the color cutting template according to the noise pixel points.

acquiring the RGB value of each pixel in the intersection area;

converting each of the RGB values to HSV values;

and obtaining noise pixel points corresponding to the cutting colors in the intersection area according to the H channel numerical values in the HSV values, and generating the color cutting template according to the noise pixel points.

Preferably, after the generating the color clipping template corresponding to the noise color in the intersection region, the method further includes:

accumulating the color cutting templates corresponding to the noise colors in the target four-dimensional radiography image to obtain the color cutting templates corresponding to the target four-dimensional radiography image;

and accumulating the color cutting template corresponding to the target four-dimensional radiography image and the color cutting templates corresponding to the four-dimensional radiography images before the target four-dimensional radiography image in the four-dimensional radiography image sequence to obtain the color cutting template.

Preferably, the denoising processing of the four-dimensional contrast image sequence by using the color clipping template includes:

calculating the intersection area of the color cutting template and each four-dimensional contrast image;

and setting the gray value of each pixel in each intersection region to zero.

Preferably, the determining the noise clipping region where the noise color is located includes:

setting the cutting area in the target four-dimensional contrast image by using a preset cutting tool; wherein the preset cutting tool is a rectangular frame or a tracing tool or an eraser tool.

Preferably, the image denoising method further includes:

determining a target to-be-processed four-dimensional ultrasonic image before the appearance of a tissue region from the to-be-processed four-dimensional ultrasonic image sequence according to the time information;

determining a target cutting area where noise in the target four-dimensional ultrasonic image to be processed is located;

constructing a time cutting template by using the target cutting area;

and denoising the to-be-processed four-dimensional ultrasonic image sequence by utilizing the time cutting template.

Preferably, the constructing a time clipping template by using the target clipping region includes:

generating a target three-dimensional cutting area according to the target cutting area;

taking the intersection region of the target three-dimensional cutting region and the target four-dimensional ultrasonic image to be processed as a target three-dimensional cutting template corresponding to the target four-dimensional ultrasonic image to be processed;

and accumulating the target three-dimensional cutting templates corresponding to the target four-dimensional ultrasonic images to be processed to obtain the time cutting templates.

Preferably, the determining a target clipping region where noise is located in the target four-dimensional ultrasound image to be processed further includes:

determining a tissue region in the four-dimensional ultrasonic image sequence to be processed;

and taking a noise area which is not intersected with the tissue area in each target four-dimensional ultrasonic image to be processed as the target cutting area.

In a second aspect, the present application further discloses an image denoising device, including:

the four-dimensional radiography time imaging module is used for carrying out four-dimensional radiography time imaging on a four-dimensional ultrasonic image sequence to be processed to obtain a four-dimensional radiography image sequence, wherein each four-dimensional radiography image in the four-dimensional radiography image sequence is a color-based time radiography image;

the color cutting information determining module is used for determining a target four-dimensional contrast image in which noise is located from the four-dimensional contrast image sequence, and determining the noise color in the target four-dimensional contrast image and a noise cutting area in which the noise color is located;

the color cutting template building module is used for building a color cutting template by utilizing the noise color and the noise cutting area;

and the four-dimensional contrast image denoising module is used for denoising the four-dimensional contrast image sequence by utilizing the color cutting template.

In a third aspect, the present application further discloses an image denoising device, including:

a memory for storing a computer program;

a processor for implementing the steps of any of the image denoising methods described above when executing the computer program.

In a fourth aspect, the present application further discloses a computer readable storage medium having a computer program stored thereon, which when executed by a processor, implements the steps of any of the image denoising methods described above.

The image denoising method comprises the steps of carrying out four-dimensional contrast time imaging on a four-dimensional ultrasonic image sequence to be processed to obtain a four-dimensional contrast image sequence, wherein each four-dimensional contrast image in the four-dimensional contrast image sequence is a color-based time contrast image; determining a target four-dimensional contrast image where noise is located from the four-dimensional contrast image sequence, and determining a noise color in the target four-dimensional contrast image and a noise clipping region where the noise color is located; constructing a color cutting template by using the noise colors and the noise cutting area; and denoising the four-dimensional contrast image sequence by using the color cutting template.

Therefore, when the noise information is adjacent to or intersected with the tissue information, the noise area and the tissue area can be accurately distinguished according to the color information, the noise information can be accurately removed, the mistaken removal of the tissue information is avoided, and the accuracy of image denoising is effectively improved; in addition, the noise removal of the image sequence is realized by constructing the color cutting template, and the image denoising efficiency is effectively improved.

The image denoising device, the image denoising device and the computer readable storage medium provided by the application all have the beneficial effects, and are not described herein again.

Drawings

In order to more clearly illustrate the technical solutions in the prior art and the embodiments of the present application, the drawings that are needed to be used in the description of the prior art and the embodiments of the present application will be briefly described below. Of course, the following description of the drawings related to the embodiments of the present application is only a part of the embodiments of the present application, and it will be obvious to those skilled in the art that other drawings can be obtained from the provided drawings without any creative effort, and the obtained other drawings also belong to the protection scope of the present application.

Fig. 1 is a schematic flow chart of an image denoising method provided in the present application;

FIG. 2 is a four-dimensional contrast image with noise information provided herein;

FIG. 3 is another four-dimensional angiographic image with noise information as provided herein;

FIG. 4 is a schematic diagram of a sequence of four-dimensional ultrasound images to be processed according to the present application;

FIG. 5 is a four-dimensional ultrasound image to be processed with noise information according to the present application;

FIG. 6 is a schematic diagram of a cropping area in a target four-dimensional ultrasound image to be processed according to the present application;

FIG. 7 is a schematic diagram of generating a temporal cropping template based on a total cropping area as provided herein;

FIG. 8 is a schematic diagram of a temporal cropping template generated based on a partial cropping area as provided herein;

fig. 9 is a schematic diagram illustrating image denoising based on a time clipping template corresponding to all clipping regions according to the present application;

fig. 10 is a schematic diagram illustrating image denoising based on a temporal clipping template corresponding to a partial clipping region according to the present application;

FIG. 11 is a schematic diagram of another pending four-dimensional ultrasound image sequence provided by the present application;

FIG. 12 is a schematic diagram of a four-dimensional contrast image sequence provided by the present application;

FIG. 13 is a schematic diagram of a cropping area in a target four-dimensional ultrasound image according to the present application;

FIG. 14 is a schematic diagram illustrating a color cropping template generated based on color information as provided herein;

FIG. 15 is a schematic diagram illustrating image denoising based on a color clipping template according to the present disclosure;

fig. 16 is a schematic structural diagram of an image denoising apparatus provided in the present application;

fig. 17 is a schematic structural diagram of an image denoising apparatus provided in the present application.

Detailed Description

The core of the application is to provide an image denoising method, which can realize accurate segmentation of noise information and tissue information in image data and effectively improve the accuracy of image denoising; another core of the present application is to provide an image denoising device, an image denoising apparatus, and a computer readable storage medium, which also have the above beneficial effects.

In order to more clearly and completely describe the technical solutions in the embodiments of the present application, the technical solutions in the embodiments of the present application will be described below with reference to the drawings in the embodiments of the present application. It is to be understood that the embodiments described are only a few embodiments of the present application and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.

At present, image denoising is mostly realized based on spatial information, however, it is difficult to rapidly and accurately remove noise information by editing an image based on spatial information, for example, when the noise information in the image is adjacent to or intersected with tissue information, since it is difficult to outline the tissue information, part of the tissue information is removed while the noise information is removed, thereby causing inaccuracy of image denoising. Therefore, in order to solve the above technical problem, an embodiment of the present application provides an image denoising method, which can implement accurate segmentation of noise information and tissue information in image data, and effectively improve the accuracy of image denoising.

Referring to fig. 1, fig. 1 is a schematic flowchart illustrating an image denoising method provided in the present application, where the image denoising method may include:

s101: and performing four-dimensional contrast time imaging on the four-dimensional ultrasonic image sequence to be processed to obtain a four-dimensional contrast image sequence, wherein each four-dimensional contrast image in the four-dimensional contrast image sequence is a color-based time contrast image.

The step aims to realize a four-dimensional radiography time imaging function, and carries out four-dimensional radiography time imaging on the acquired to-be-processed four-dimensional ultrasonic image sequence, namely, carrying out four-dimensional radiography time imaging on each frame of to-be-processed four-dimensional ultrasonic image in the to-be-processed four-dimensional ultrasonic image sequence to obtain a corresponding four-dimensional radiography image, wherein the four-dimensional radiography image corresponding to each frame of to-be-processed four-dimensional ultrasonic image forms the four-dimensional radiography image sequence. The sequence of the four-dimensional ultrasound images to be processed is a continuous image frame which needs to be subjected to noise removal and can be acquired by image acquisition equipment (such as an ultrasound probe).

It can be understood that each frame of the four-dimensional contrast image in the four-dimensional contrast image sequence is a color-based time contrast image, and after the four-dimensional contrast image is obtained by performing four-dimensional contrast time imaging on the four-dimensional ultrasonic image, the colors of rendered images of a noise part and an organization part in the four-dimensional contrast image are inconsistent, so that the noise area and the organization area can be accurately distinguished according to color information in the images, thereby realizing noise removal based on the color information and achieving the purpose of accurately removing noise.

S102: and determining a target four-dimensional contrast image where the noise is located from the four-dimensional contrast image sequence, and determining the noise color in the target four-dimensional contrast image and a noise cutting area where the noise color is located.

The step aims to determine cutting information, wherein the cutting information refers to relevant information of noise needing to be cut, specifically to noise color and a region where the noise color is located, namely the noise cutting region. Firstly, a target four-dimensional contrast image is determined from a four-dimensional contrast image sequence, the target four-dimensional contrast image is an image frame with noise, the traversal can be started from a first frame of four-dimensional contrast image in the four-dimensional contrast image sequence until the four-dimensional contrast image with noise is traversed, and the target four-dimensional contrast image is taken as the target four-dimensional contrast image.

Further, determining a noise color in the target four-dimensional contrast image and a noise clipping region where the noise color is located may be specifically realized based on technical experience of a worker, that is, by observing the obtained four-dimensional contrast image, and more specifically, when rendering colors of the noise information and the tissue information in the target four-dimensional contrast image are different, the noise color and the region where the noise color is located may be directly determined according to color distribution, for example, in the four-dimensional contrast image of the uterine tube, the noise information is generally red, at this time, the noise color in the target four-dimensional contrast image may be set to red, and the region where the red is located in the target four-dimensional contrast image is taken as the noise clipping region; when the rendering colors of the noise information and the tissue information in the target four-dimensional contrast image are the same, the determination may be performed according to the display shape, for example, when information that does not conform to the uterine tube shape exists in the target four-dimensional contrast image, the display color of the region where the information is located may be used as the noise color, and the region where the information is located may be used as the noise clipping region.

It should be noted that, because of the difference in noise color and the difference in noise occurrence region, the number of noise colors in the same target four-dimensional contrast image may be multiple, and the number of corresponding noise clipping regions may also be multiple, but the image denoising method provided by the present application is applicable to both a noise clipping region and multiple noise clipping regions, and can achieve rapid and accurate image noise removal.

As a preferred embodiment, the determining the noise clipping region where the noise color is located may include: setting a cutting area in a target four-dimensional contrast image by using a preset cutting tool; wherein the preset cutting tool is a rectangular frame or a tracing tool or an eraser tool.

The preferred embodiment provides a method for determining a noise clipping region, which can be implemented based on a corresponding clipping tool, including but not limited to a rectangular frame, a tracing tool, an eraser tool, etc. It can be understood that the noise clipping region is a region where noise is located, and the region may be a partial region of the target four-dimensional contrast image or a whole region of the target four-dimensional contrast image, and may be set by a technician according to actual requirements. For example, referring to fig. 2, fig. 2 is a four-dimensional contrast image with noise information provided by the present application, wherein

reference numerals

1 and 2 represent different colors, assuming that the information represented by reference numeral 1 is the noise information, and the information represented by reference numeral 2 is the detected target tissue information, at this time, a rectangular frame may be used to set a noise clipping region, such as a region shown in region 1, where region 1 is a partial region of the target four-dimensional contrast image, and the partial region is the noise clipping region; of course, when the tissue information represented by reference numeral 2 is not present in the target four-dimensional contrast image, the entire region of the target four-dimensional contrast image may be selected and set as the noise clipping region. For another example, referring to fig. 3, fig. 3 is another four-dimensional contrast image with noise information provided by the present application, wherein

reference numerals

1, 2 and 3 represent different colors, the information represented by reference numeral 2 includes information in a circular area and information in an irregularly shaped area, according to the display shape, the information in the irregular area is the tissue information, the information in the circular area is the noise information, that is, the information in the circular area represented by the reference numeral 1 and the information in the circular area represented by the reference numeral 2 are both the noise information, the information in the irregular area represented by the reference numeral 2 and the information represented by the reference numeral 3 are both the tissue information, in this case, a noise clipping region may be set by using a rectangular frame and a tracing tool, such as

regions

1 and 2 shown in fig. 3, where the

regions

1 and 2 are the noise clipping regions of the target four-dimensional contrast image.

S103: constructing a color cutting template by using the noise color and the noise cutting area;

the method comprises the steps of constructing a color cutting template, specifically constructing the color cutting template by using noise colors and noise cutting areas, wherein the color cutting template is a template for realizing image denoising based on color information. It can be understood that the color cropping template corresponds to the target four-dimensional contrast image, in other words, when there is one noise color, the color cropping template is constructed based on the noise color and the noise cropping area thereof, that is, the template corresponding to the noise color is the color cropping template corresponding to the target four-dimensional contrast image; when the noise color is multiple, the color clipping template is constructed based on the multiple noise colors and the noise clipping regions corresponding to the multiple noise colors, specifically, the template corresponding to each noise color is constructed first, and then the templates corresponding to the colors are superposed to obtain the color clipping template corresponding to the target four-dimensional image.

S104: and denoising the four-dimensional contrast image sequence by using a color cutting template.

The method comprises the steps of denoising an image of a four-dimensional contrast image sequence, and denoising the four-dimensional contrast image sequence by using a color cutting template, namely denoising each frame of four-dimensional contrast image in the four-dimensional contrast image sequence by using the color cutting template. It should be noted that the color cropping template corresponds to the four-dimensional contrast image sequence, because the color cropping template is constructed and obtained based on the four-dimensional contrast image in the four-dimensional contrast image sequence, for another acquired four-dimensional contrast image sequence, the construction of the color cropping template needs to be performed again, and then the image denoising is completed.

As a preferred embodiment, the denoising processing of the four-dimensional contrast image sequence by using the color cropping template may include: calculating the intersection area of the color cutting template and each four-dimensional contrast image; and setting the gray value of each pixel in each intersection area to zero.

Specifically, when the color clipping template is used to denoise the four-dimensional contrast image sequence, the data clipping may be implemented by zeroing the gray value of each pixel in the noise region, so as to complete the noise removal. It can be understood that the noise region is an intersection region of the color cropping template and the four-dimensional contrast image of each frame, because the color cropping template is constructed based on the four-dimensional contrast image sequence, a region of the four-dimensional contrast image that intersects with the color cropping template is necessarily a region where the noise is located.

As described above, the color clipping template may be constructed and obtained based on the noise color and the noise clipping region corresponding to the noise color, and therefore, on the basis of the above embodiment, as a preferred embodiment, the constructing of the color clipping template by using the noise color and the noise clipping region may include: generating a three-dimensional cutting area according to an area corresponding to the noise color in the noise cutting area; calculating an intersection area of the three-dimensional cutting area and the target four-dimensional contrast image; and generating a color cutting template corresponding to the noise color in the intersection area.

The preferred embodiment provides a method for constructing a color cutting template, wherein the color cutting template is a three-dimensional template. Firstly, determining a region corresponding to a noise color in a noise clipping region, and generating a three-dimensional clipping region in the region, for example, as shown in fig. 2, a circular region with a label 1 in a region 1 may be generated into a three-dimensional clipping template, and it can be understood that, here, generating a three-dimensional clipping region from a region corresponding to a noise color, instead of generating a three-dimensional clipping region from a noise clipping region, may effectively ensure the accuracy of a finally constructed color clipping template, and avoid a situation that non-noise information is removed due to the color clipping template including a non-noise region; further, since the three-dimensional cropping template is generated based on the region corresponding to the noise color, and the information in the region is necessarily noise information, the region where the three-dimensional cropping template intersects with the target four-dimensional contrast image is necessarily a noise region, and thus, the color cropping template corresponding to the noise color can be generated in the intersection region.

It can be understood that, in the four-dimensional imaging time, due to different preset CHROMA pseudo-color parameters, the color distribution form in the corresponding generated four-dimensional imaging image may also be different, and may be divided into a gradient color and a non-gradient color, and for a target four-dimensional imaging image with different color distribution forms, in the process of generating a color cropping template based on the intersection region of the three-dimensional cropping template and the target four-dimensional imaging image, different implementation manners may be adopted:

1. the color distribution is in the form of non-gradient colors:

as a preferred embodiment, the generating a color clipping template corresponding to the noise color in the intersection region may include: acquiring the RGB value of each pixel in the intersection area; and obtaining noise pixel points corresponding to the cutting colors in the intersection areas according to the RGB values of the pixels, and generating a color cutting template according to the noise pixel points.

When the color distribution form in the target four-dimensional contrast image is not a gradual color, the color clipping template can be generated by using the pixel RGB value in the process of generating the color clipping template based on the intersection area of the three-dimensional clipping template and the target four-dimensional contrast image. Specifically, when the rendering colors of the noise information and the organization information in the four-dimensional contrast image are different, the RGB values of the pixels in the corresponding noise region are different from the RGB values of the pixels in the non-noise region, so that the RGB values of the pixels in the intersection region of the three-dimensional clipping template and the target four-dimensional contrast image can be obtained, and noise pixel points are obtained from all the pixels in the intersection region by using the RGB values of the pixels corresponding to the noise as a standard, that is, by using the RGB values corresponding to the determined noise colors as a standard, thereby generating the color clipping template based on the noise.

It can be understood that, when extracting the noise pixel point based on the pixel RGB value, the determination may be specifically performed according to R, G, B any channel or channel combination value. For example, referring to table 1, table 1 is a table of RGB values of non-gradient colors provided in the present application:

TABLE 1A RGB value table without color gradient

Based on the table, it can be known that the values of the G channels of the color segments are not consistent, and therefore, the determination of the noise pixel point can be realized by reading the values of the G channels. Assuming that the noise color is red, there are:

if(G_Channel＝＝35)

COLOR＝Red

therefore, the pixels with the G channel value of 35 are used as noise pixel points.

2. The color distribution is in the form of a gradient color:

as a preferred embodiment, the generating a color clipping template corresponding to the noise color in the intersection region may include: acquiring the RGB value of each pixel in the intersection area; converting each RGB value into an HSV value; and obtaining noise pixel points corresponding to the cutting colors in the intersection area according to the H channel (Hue channel, Hue value) value in each HSV value, and generating a color cutting template according to the noise pixel points.

When the color distribution form in the target four-dimensional contrast image is a gradient color, the color cutting template can be generated by using the H channel value in the pixel HSV value in the process of generating the color cutting template based on the intersection area of the three-dimensional cutting template and the target four-dimensional contrast image. It can be understood that, for the case that the color distribution form in the target four-dimensional contrast image is not a gradient color, the color clipping template can be constructed based on the RGB values of the pixels, and for the case that the color distribution form in the target four-dimensional contrast image is a gradient color, since the RGB values of each segment of color are gradient and cannot be directly judged through the RGB values, the color clipping template can be constructed based on the H channel values in the HSV values of the pixels, because the values of the H channels corresponding to the colors are increased after being converted into the HSV color space, and the range is 0 to 200.

Specifically, the RGB values of the pixels in the intersection region of the three-dimensional clipping template and the target four-dimensional contrast image may be obtained first, and then the RGB values are converted into HSV values, at this time, noise pixel points are obtained from all pixels in the intersection region using an H channel value of an HSV value corresponding to a noise color as a standard, that is, an H channel value of an HSV value corresponding to a determined noise color as a standard, so that the color clipping template is generated based on the noise pixel points. For example, please refer to table 2, table 2 is a table of values of H channel in HSV values with gradient colors provided in the present application:

TABLE 2A numerical table of H channels in HSV values of gradual change color

Based on the table, when the noise color is red, then there are:

if(Hue_Channel＜15)

COLOR＝Red

therefore, pixels with the H channel numerical value smaller than 15 are used as noise pixel points.

As a preferred embodiment, after generating the color cropping template corresponding to the noise color in the intersection region, the method may further include: accumulating the color cutting templates corresponding to the noise colors in the target four-dimensional radiography image to obtain the color cutting templates corresponding to the target four-dimensional radiography image; and accumulating the color cutting template corresponding to the target four-dimensional radiography image and the color cutting templates corresponding to the four-dimensional radiography images in front of the target four-dimensional radiography image in the four-dimensional radiography image sequence to obtain the color cutting template.

It is to be understood that the above construction of the color clipping template is performed in the case where the noise color is one, then, when the noise color is plural, a template corresponding to each noise color may be constructed first, and then the templates corresponding to each color may be superimposed to obtain a color clipping template corresponding to the target four-dimensional radiographic image, and finally, accumulating the color cutting template corresponding to the target four-dimensional contrast image and the color cutting templates corresponding to the four-dimensional contrast images before the target four-dimensional contrast image to obtain a color cutting template, for example, assume that the target four-dimensional contrast image is the fourth frame image in the four-dimensional contrast image sequence, the color clipping template corresponding to the target four-dimensional radiography image and the color clipping templates corresponding to the first three frames of four-dimensional radiography images can be accumulated to obtain the final color clipping template which can be used for realizing the noise removal of the four-dimensional radiography image sequence.

Therefore, the above embodiment provides a specific construction process of the color clipping template, and the noise removal of multi-frame images can be realized by constructing the color clipping template to perform image denoising, so that the image denoising efficiency is effectively improved.

As described above, the image denoising method provided by the above embodiment is implemented based on color information, and effectively overcomes the problem that the noise information and the tissue information cannot be accurately distinguished due to adjacency or intersection of the two information. It can be understood that the above image denoising method based on color information is applicable to both the case where the noise information is adjacent to or intersects with the tissue information, and the case where the noise information and the tissue information are not intersected. Furthermore, another image denoising method is provided for the condition that the noise information and the tissue information are not intersected, namely, the image denoising based on the time information is provided, the image denoising method does not need to carry out four-dimensional contrast time imaging processing on a four-dimensional ultrasonic image sequence to be processed, and the image denoising efficiency can be effectively ensured.

Therefore, on the basis of the foregoing embodiments, as a preferred embodiment, the image denoising method may further include: determining a target to-be-processed four-dimensional ultrasonic image before the appearance of the tissue region from the to-be-processed four-dimensional ultrasonic image sequence according to the time information; determining a target cutting area where noise in a target four-dimensional ultrasonic image to be processed is located; constructing a time cutting template by using the target cutting area; and denoising the four-dimensional ultrasonic image sequence to be processed by utilizing a time cutting template.

Firstly, for an acquired sequence of four-dimensional ultrasound images to be processed, a four-dimensional ultrasound image to be processed before a tissue region appears may be determined therefrom and is taken as a target four-dimensional ultrasound image to be processed, it may be understood that the number of the target four-dimensional ultrasound images to be processed is not unique, and may be multiple frames, or may be one frame, and specifically may be determined based on time information, and the time information may be an image frame number or an image frame time value, for example, when the image frame number is set to be "second frame", then the second frame and the four-dimensional ultrasound image to be processed before the second frame (i.e., first frame) may be taken as the target four-dimensional ultrasound image to be processed; when the image frame time value is set to "0.8 s", the four-dimensional ultrasound image to be processed before 0.8s (including 0.8s) may be used as the target four-dimensional ultrasound image to be processed. Specifically, similarly to the determination method of the target four-dimensional contrast image, the sequence of four-dimensional ultrasound images to be processed may also be traversed until each of the four-dimensional ultrasound images to be processed before the occurrence of the tissue information is obtained as the target four-dimensional ultrasound image to be processed. For example, referring to fig. 4, fig. 4 is a schematic diagram of a to-be-processed four-dimensional ultrasound image sequence provided in the present application, and it is assumed that an irregular shaded portion in the image sequence is an organization region, and the remaining regular shaded portions (including a circular region and a rectangular region) are noise regions, it can be known that noise information already appears in a first frame of to-be-processed four-dimensional ultrasound image, and organization information begins to appear in a third frame of to-be-processed four-dimensional ultrasound image, at this time, the first frame of to-be-processed four-dimensional ultrasound image and the second frame of to-be-processed four-dimensional ultrasound image may be used as the target to-be-processed four-dimensional ultrasound.

Furthermore, after the target four-dimensional ultrasonic image to be processed is determined, the region where the noise is located can be determined from the target four-dimensional ultrasonic image, the region is used as a target cutting region, and a corresponding time cutting template is constructed by using the target cutting region, so that the four-dimensional ultrasonic image sequence to be processed can be denoised by using the time cutting template. Similarly, the number of the target clipping areas is not only one, but may be only one, and there may also be a plurality of target clipping areas, so when there is one target clipping area, it is only necessary to construct a corresponding time clipping template directly based on the target clipping area; when a plurality of target clipping areas are provided, the templates can be created for each target clipping area, and then the templates are accumulated to obtain the final time clipping template. In addition, the setting of the target cutting area can also be realized based on a preset cutting tool, and the target cutting area can be a partial area of the four-dimensional ultrasonic image to be processed or a whole area of the four-dimensional ultrasonic image to be processed, and the setting can be performed by a technician according to actual requirements, which is not limited in the present application.

And finally, denoising the image of the four-dimensional ultrasonic image sequence to be processed by using the time cutting template, and sequentially denoising each frame of four-dimensional ultrasonic image to be processed in the four-dimensional ultrasonic image sequence to be processed. Similarly, image denoising can be realized by adopting a method of setting the gray value of each pixel in the noise area of the four-dimensional ultrasonic image to be processed to zero, that is, the intersection area of the time clipping template and each frame of the four-dimensional ultrasonic image to be processed is calculated, and the gray value of each pixel in each intersection area is set to zero.

As a preferred embodiment, the constructing the temporal clipping template by using the target clipping region may include: generating a target three-dimensional cutting area according to the target cutting area; taking the intersection area of the target three-dimensional cutting area and the target four-dimensional ultrasonic image to be processed as a target three-dimensional cutting template corresponding to the target four-dimensional ultrasonic image to be processed; and accumulating the target three-dimensional cutting templates corresponding to the target four-dimensional ultrasonic images to be processed to obtain a time cutting template.

The preferred embodiment provides a method for constructing a time clipping template, which is similar to the above-mentioned color clipping template construction process, and the time clipping template is also a three-dimensional template. Specifically, each target cutting area is generated into a corresponding target three-dimensional cutting area, and since the target cutting area is an area where noise is located, an intersection area of each target three-dimensional cutting area and a target four-dimensional ultrasonic image to be processed is inevitably a noise area, so that a target three-dimensional cutting template corresponding to the target four-dimensional ultrasonic image to be processed can be generated in the intersection area; and finally, accumulating the target three-dimensional cutting templates corresponding to the target four-dimensional ultrasonic images to be processed to obtain a time cutting template, for example, if the target four-dimensional ultrasonic images to be processed are the first four frames of images in the sequence of the four-dimensional ultrasonic images to be processed, accumulating the target three-dimensional cutting templates corresponding to the four frames of target four-dimensional ultrasonic images to be processed to obtain the final time cutting template which can be used for removing the noise of the sequence of the four-dimensional ultrasonic images to be processed.

As a preferred embodiment, the determining a target clipping region where noise is located in the target four-dimensional ultrasound image to be processed may include: determining a tissue region in a four-dimensional ultrasonic image sequence to be processed; and taking a noise area which is not intersected with the tissue area in each target four-dimensional ultrasonic image to be processed as the target cutting area.

In the whole denoising process, a time cutting template can be used for denoising the four-dimensional ultrasonic image, then the four-dimensional ultrasonic image obtained after denoising is subjected to four-dimensional contrast time imaging, and then the four-dimensional contrast image is denoised according to the color cutting template to obtain a final four-dimensional contrast image. For example, referring to fig. 5, fig. 5 is a to-be-processed four-dimensional ultrasound image with noise information provided by the present application, wherein it is assumed that the circular shadow areas in the area 1 and the area 2 are both noise areas, and the remaining irregular-shaped shadow areas are both tissue areas, if an image denoising method based on time information is adopted, since the circular shadow area in the area 2 intersects with the tissue area and the circular shadow area in the area 1 does not intersect with the tissue area, the area 1 can be used as a target clipping area of the target to-be-processed four-dimensional ultrasound image. For the noise information in the circular shadow region in the region 2, an image denoising method based on color information can be adopted to remove the noise.

Therefore, the above embodiment provides an image denoising method based on time information, and combines the image denoising method based on color information with the image denoising method based on time information, so that the image denoising problem when a noise region intersects or does not intersect with a tissue region can be effectively solved, and the accuracy of image denoising and the image denoising efficiency can be ensured.

Based on the above embodiments, the embodiments of the present application provide an image denoising method for a specific application scene, which is implemented as follows:

1. the image denoising method based on the time information comprises the following steps:

step one, reading a four-dimensional ultrasonic image sequence to be processed, and assuming that the four-dimensional ultrasonic image sequence to be processed is as shown in fig. 4.

And step two, entering image editing, setting a time clipping frame number or a clipping time value, and further determining a target to-be-processed four-dimensional ultrasonic image, wherein for the to-be-processed four-dimensional ultrasonic image sequence shown in fig. 4, it can be determined that organization information appears in a third frame of to-be-processed four-dimensional ultrasonic image, and therefore, the first frame of to-be-processed four-dimensional ultrasonic image and the second frame of to-be-processed four-dimensional ultrasonic image can be taken as target to-be-processed four-dimensional ultrasonic images.

Step three, setting a cutting area, where the cutting area supports a full image or a partial area, where the partial area may be implemented by a cutting tool such as a rectangular frame, a tracing tool, and an eraser, and taking the second frame of the four-dimensional ultrasound image to be processed as an example, as shown in fig. 6, fig. 6 is a schematic diagram of the cutting area in the target four-dimensional ultrasound image to be processed provided by the present application.

Step four, generating a time cutting template according to the cutting area:

firstly, generating a three-dimensional cutting area by a cutting area (two-dimensional) set by a user, calculating an intersection area of the three-dimensional cutting area and a target to-be-processed four-dimensional ultrasonic image, further generating a three-dimensional cutting template based on the intersection area, and finally accumulating the three-dimensional cutting template corresponding to a first frame of to-be-processed four-dimensional ultrasonic image and the three-dimensional cutting template corresponding to a second frame of to-be-processed four-dimensional ultrasonic image to obtain a final time cutting template; referring to fig. 7 and 8, fig. 7 is a schematic diagram illustrating a process of creating a time clipping template based on a whole clipping area according to the present application, and fig. 8 is a schematic diagram illustrating a process of creating a time clipping template based on a part of the clipping area according to the present application;

cutting each frame of to-be-processed four-dimensional ultrasonic image in the to-be-processed four-dimensional ultrasonic image sequence by using a time cutting template, namely setting the gray value of each pixel in the intersection area of each to-be-processed four-dimensional ultrasonic image and the time cutting template to zero to obtain a denoised four-dimensional ultrasonic image; for a specific implementation process of image denoising based on the time clipping template, please refer to fig. 9 and fig. 10, fig. 9 is a schematic diagram of image denoising based on the time clipping template corresponding to all clipping regions provided by the present application, and fig. 10 is a schematic diagram of image denoising based on the time clipping template corresponding to part of the clipping regions provided by the present application.

Therefore, the image denoising method based on the time information is realized, the image denoising method based on the time information utilizes the image frame before the tissue information appears to establish the time cutting template, and then utilizes the time cutting template to denoise the image, so that the effect of removing the noise information before the tissue information at one time is achieved. Noise that does not intersect the tissue region may also be removed.

2. The image denoising method based on the color information comprises the following steps:

step one, reading a four-dimensional ultrasound image sequence to be processed, assuming that the four-dimensional ultrasound image sequence to be processed is as shown in fig. 11, and fig. 11 is a schematic diagram of another four-dimensional ultrasound image sequence to be processed provided by the present application.

And step two, performing four-dimensional contrast time imaging on the four-dimensional ultrasonic image sequence to be processed to obtain a four-dimensional contrast image sequence, where the four-dimensional contrast image sequence is shown in fig. 12, and fig. 12 is a schematic diagram of the four-dimensional contrast image sequence provided by the present application, where different shades represent different colors.

And step three, entering image editing, setting a cutting frame number, and setting a fourth frame of four-dimensional contrast image as a target four-dimensional contrast image for the four-dimensional contrast image sequence shown in the figure 12.

Step four, starting color clipping, and setting clipping colors (i.e. noise colors, such as red, yellow, etc.) and a clipping region, where the clipping region supports a full image or a partial region, where the partial region may be implemented by a clipping tool, such as a rectangular frame, a tracing tool, and an eraser, as shown in fig. 13, fig. 13 is a schematic diagram of the clipping region in the target four-dimensional ultrasound image provided by the present application, and the clipping region is a region shown by the rectangular frame and a region shown by the tracing frame.

When the noise color and the tissue color are the same, as long as the areas where the noise color and the tissue color appear are different, the area frame where the noise appears can be selected through a cutting tool, the part corresponding to the noise color in the frame selection area is used as a color cutting template, the noise can be flexibly removed in the tissue area, and the cutting of an effective tissue area is avoided.

And step five, generating a color cutting template according to the cutting color and the cutting area.

Firstly, generating a three-dimensional cutting area in a cutting area (two-dimensional) set by a user, calculating an intersection area of the three-dimensional cutting area and a fourth frame of four-dimensional contrast image, judging colors in the intersection area to generate a cutting template corresponding to the cutting colors in the intersection area, further accumulating the cutting templates corresponding to the same or different cutting colors in all the intersection areas to obtain a three-dimensional cutting template corresponding to the fourth frame of four-dimensional contrast image, and finally accumulating the three-dimensional cutting template corresponding to the fourth frame of four-dimensional contrast image and the three-dimensional cutting templates corresponding to the first three frames of four-dimensional contrast image to obtain a final color cutting template; referring to fig. 14, fig. 14 is a schematic diagram illustrating a process of constructing a color clipping template according to color information provided by the present application.

Cutting each frame of four-dimensional contrast image in the four-dimensional contrast image sequence by using a color cutting template, namely setting the gray value of each pixel in the intersection area of each four-dimensional contrast image and the color cutting template to zero to obtain a denoised four-dimensional contrast image; fig. 15 is a schematic diagram of a process for implementing image denoising based on a color clipping template, where fig. 15 is a schematic diagram of image denoising based on a color clipping template provided in the present application.

Therefore, image denoising based on color information is realized, the color cutting template is constructed based on the color information by the image denoising method based on the color information, image denoising is carried out by using the color cutting template, the tissue information and the noise information can be accurately distinguished under the condition that the tissue information and the noise information are intersected, then the noise information is accurately removed, and the image denoising accuracy is effectively ensured.

As described above, the image denoising function is realized based on the time information and the color information, in the actual ultrasonic four-dimensional imaging process, the ultrasonic video stream can be denoised in a mode of combining the time information and the color information, before the color rendering, the image denoising based on the time information is adopted, and after the color rendering is completed, the image denoising based on the color information is adopted, so that the image denoising efficiency and the image denoising accuracy can be effectively ensured.

Therefore, the image denoising method provided by the embodiment of the application realizes the image denoising function based on the color information, and can divide the noise cutting area according to the noise color and construct the color cutting template after the four-dimensional ultrasonic image is imaged in the four-dimensional imaging time to obtain the four-dimensional contrast image, so that the image denoising is realized by using the color cutting template, and therefore, when the noise information is adjacent to or intersected with the tissue information, the noise area can be accurately distinguished from the tissue area according to the color information, the noise information can be accurately removed, the mistaken removal of the tissue information is avoided, and the image denoising accuracy is effectively improved; in addition, the noise removal of the image sequence is realized by constructing the color cutting template, and the image denoising efficiency is effectively improved.

To solve the above technical problem, the present application further provides an image denoising apparatus, please refer to fig. 16, where fig. 16 is a schematic structural diagram of the image denoising apparatus provided in the present application, and the image denoising apparatus may include:

the four-dimensional radiography time imaging module 1 is used for performing four-dimensional radiography time imaging on a four-dimensional ultrasonic image sequence to be processed to obtain a four-dimensional radiography image sequence, wherein each four-dimensional radiography image in the four-dimensional radiography image sequence is a color-based time radiography image;

the color cutting information determining module 2 is used for determining a target four-dimensional contrast image in which noise is located from the four-dimensional contrast image sequence, and determining the noise color in the target four-dimensional contrast image and a noise cutting area in which the noise color is located;

the color cutting template building module 3 is used for building a color cutting template by using noise colors and noise cutting areas;

and the four-dimensional contrast image denoising module 4 is used for denoising the four-dimensional contrast image sequence by using the color cutting template.

Therefore, the image denoising device provided by the embodiment of the application realizes the image denoising function based on the color information, and can divide the noise cutting area according to the noise color and construct the color cutting template after the four-dimensional ultrasonic image is imaged in the four-dimensional imaging time to obtain the four-dimensional contrast image, so that the image denoising is realized by using the color cutting template, and therefore, when the noise information is adjacent to or intersected with the tissue information, the noise area can be accurately distinguished from the tissue area according to the color information, the noise information can be accurately removed, the mistaken removal of the tissue information is avoided, and the image denoising accuracy is effectively improved; in addition, the noise removal of the image sequence is realized by constructing the color cutting template, and the image denoising efficiency is effectively improved.

As a preferred embodiment, the color cropping template building module 3 may include:

the three-dimensional region generating unit is used for generating a three-dimensional cutting region according to a region corresponding to the noise color in the noise cutting region;

the intersection region calculation unit is used for calculating the intersection region of the three-dimensional cutting region and the target four-dimensional contrast image;

and the cutting template generating unit is used for generating a color cutting template corresponding to the noise color in the intersection area.

As a preferred embodiment, the cropping template generating unit may be specifically configured to obtain RGB values of pixels in the intersection region; and obtaining noise pixel points corresponding to the cutting colors in the intersection areas according to the RGB values of the pixels, and generating a color cutting template according to the noise pixel points.

As a preferred embodiment, the cropping template generating unit may be specifically configured to obtain RGB values of pixels in the intersection region; converting each RGB value into an HSV value; and obtaining noise pixel points corresponding to the cutting colors in the intersection area according to the H channel numerical value in each HSV value, and generating a color cutting template according to the noise pixel points.

As a preferred embodiment, the color cropping template building module 3 may further include a template accumulating unit, configured to accumulate color cropping templates corresponding to noise colors in the target four-dimensional contrast image to obtain a color cropping template corresponding to the target four-dimensional contrast image; and accumulating the color cutting template corresponding to the target four-dimensional radiography image and the color cutting templates corresponding to the four-dimensional radiography images in front of the target four-dimensional radiography image in the four-dimensional radiography image sequence to obtain the color cutting template.

As a preferred embodiment, the four-dimensional contrast image denoising module 4 may be specifically configured to calculate an intersection region between the color cropping template and each four-dimensional contrast image; and setting the gray value of each pixel in each intersection area to zero.

As a preferred embodiment, the color cropping information determining module 2 may be specifically configured to set a cropping area in the target four-dimensional contrast image by using a preset cropping tool; wherein the preset cutting tool is a rectangular frame or a tracing tool or an eraser tool.

As a preferred embodiment, the image denoising apparatus may further include:

the target image determining module is used for determining a target to-be-processed four-dimensional ultrasonic image before the appearance of the tissue region from the to-be-processed four-dimensional ultrasonic image sequence according to the time information;

the cutting area determining module is used for determining a target cutting area where noise in the target four-dimensional ultrasonic image to be processed is located;

the time cutting template building module is used for building a time cutting template by utilizing the target cutting area;

and the four-dimensional ultrasonic image denoising module is used for denoising the four-dimensional ultrasonic image sequence to be processed by utilizing the time cutting template.

As a preferred embodiment, the time clipping template building module may be specifically configured to generate a target three-dimensional clipping region according to the target clipping region; taking the intersection area of the target three-dimensional cutting area and the target four-dimensional ultrasonic image to be processed as a target three-dimensional cutting template corresponding to the target four-dimensional ultrasonic image to be processed; and accumulating the target three-dimensional cutting templates corresponding to the target four-dimensional ultrasonic images to be processed to obtain a time cutting template.

As a preferred embodiment, the clipping region determining module may be specifically configured to determine a tissue region in a four-dimensional ultrasound image sequence to be processed; and taking a noise area which is not intersected with the tissue area in the four-dimensional ultrasonic image to be processed of each target as a target cutting area.

For the introduction of the apparatus provided in the present application, please refer to the above method embodiments, which are not described herein again.

To solve the above technical problem, the present application further provides an image denoising apparatus, please refer to fig. 17, where fig. 17 is a schematic structural diagram of the image denoising apparatus provided in the present application, and the image denoising apparatus may include:

a memory 10 for storing a computer program;

the processor 20, when executing the computer program, may implement the steps of any of the image denoising methods described above.

For the introduction of the device provided in the present application, please refer to the above method embodiment, which is not described herein again.

To solve the above problem, the present application further provides a computer-readable storage medium, on which a computer program is stored, and the computer program, when executed by a processor, can implement any of the steps of the image denoising method described above.

The computer-readable storage medium may include: various media capable of storing program codes, such as a usb disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.

For the introduction of the computer-readable storage medium provided in the present application, please refer to the above method embodiments, which are not described herein again.

The embodiments are described in a progressive manner in the specification, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other. The device disclosed by the embodiment corresponds to the method disclosed by the embodiment, so that the description is simple, and the relevant points can be referred to the method part for description.

Those of skill would further appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both, and that the various illustrative components and steps have been described above generally in terms of their functionality in order to clearly illustrate this interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.

The steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. A software module may reside in Random Access Memory (RAM), memory, Read Only Memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

The technical solutions provided by the present application are described in detail above. The principles and embodiments of the present application are explained herein using specific examples, which are provided only to help understand the method and the core idea of the present application. It should be noted that, for those skilled in the art, without departing from the principle of the present application, several improvements and modifications can be made to the present application, and these improvements and modifications also fall into the protection scope of the present application.

Claims

1. An image denoising method, comprising:

2. The image denoising method of claim 1, wherein the constructing a color clipping template using the noise color and the noise clipping region comprises:

3. The image denoising method of claim 2, wherein the generating a color clipping template corresponding to the noise color in the intersection region comprises:

acquiring the RGB value of each pixel in the intersection area;

4. The image denoising method of claim 2, wherein the generating a color clipping template corresponding to the noise color in the intersection region comprises:

acquiring the RGB value of each pixel in the intersection area;

converting each of the RGB values to HSV values;

5. The image denoising method of claim 2, wherein after generating the color cropping template corresponding to the noise color at the intersection region, the method further comprises:

6. The image denoising method of claim 1, wherein the denoising the sequence of four-dimensional contrast images using the color cropping template comprises:

and setting the gray value of each pixel in each intersection region to zero.

7. The image denoising method of claim 1, wherein determining a noise clipping region where the noise color is located comprises:

8. The image denoising method according to any one of claims 1 to 7, further comprising:

constructing a time cutting template by using the target cutting area;

9. The image denoising method of claim 8, wherein the constructing a temporal clipping template using the target clipping region comprises:

10. The image denoising method of claim 8, wherein the determining a target clipping region where noise is located in the target four-dimensional ultrasound image to be processed further comprises:

11. An image denoising apparatus, comprising:

12. An image denoising apparatus, comprising:

a memory for storing a computer program;

a processor for implementing the steps of the image denoising method according to any one of claims 1 to 10 when executing the computer program.

13. A computer-readable storage medium, characterized in that a computer program is stored thereon, which, when being executed by a processor, carries out the steps of the image denoising method according to any one of claims 1 to 10.