WO2016165209A1

WO2016165209A1 - 3d image cropping method

Info

Publication number: WO2016165209A1
Application number: PCT/CN2015/081450
Authority: WO
Inventors: 刘静静; 陈永健; 田广野
Original assignee: 青岛海信医疗设备股份有限公司
Priority date: 2015-04-17
Filing date: 2015-06-15
Publication date: 2016-10-20
Also published as: CN104794758A; CN104794758B

Abstract

A 3D image cropping method comprises: rendering an image surface according to a surface element of an object, acquiring a 3D display image of the object in a screen 2D coordinate system, acquiring a 2D coordinate in the screen 2D coordinate system corresponding to each of voxels of the object; alternatively, performing volume rendering according to voxels of the object, acquiring the 3D display image of the object in the screen 2D coordinate system and 2D coordinates corresponding to each of the voxels of the object in the screen 2D coordinate system (100); determining a region to be cropped in the 3D display image (110); discarding a voxel corresponding to a 2D coordinate of a pixel point of the 3D display image within the region to be cropped, and the surface element corresponding to the voxel (120); and re-rendering the image surface according to the remaining surface elements, or re-rendering the image volume according to the remaining voxels (130).

Description

Three-dimensional image cutting method

Technical field

The present invention relates to the field of image processing technologies, and in particular, to a method for cropping a three-dimensional image.

Background technique

Whether you want to obtain a perfect three-dimensional (3D) image of an object or to simulate the operation of the object on a 3D image of the object, you need to crop the 3D image.

Taking medical scenes as an example, complex and precise surgical procedures are increasingly inseparable from 3D medical image analysis and computer-assisted surgery systems. It is difficult to generate perfect 3D images at a time in 3D medical image analysis systems, and it is often necessary to crop them from 3D images to form perfect 3D images. In computer-assisted surgery systems, if you need to perform surgical planning and surgical simulation directly on the 3D image, you also need to crop the 3D image.

Therefore, how to realize the cropping of 3D images is a problem that needs to be solved at present.

Summary of the invention

It is an object of the present invention to provide a method of cropping a three-dimensional image to effect cropping of a 3D image.

The object of the invention is achieved by the following technical solutions:

A method for cropping a three-dimensional image, comprising:

Obtaining an image surface according to a surface element of the object, obtaining a three-dimensional display image of the object in a two-dimensional coordinate system of the screen, and acquiring a corresponding two-dimensional coordinate of each voxel of the object in the two-dimensional coordinate system of the screen; or And performing image volume rendering according to the voxel of the object, obtaining a three-dimensional display image of the object in a two-dimensional coordinate system of the screen, and corresponding two-dimensional corresponding to each voxel of the object in the two-dimensional coordinate system of the screen. coordinate;

Determining a region to be cropped on the three-dimensional display image;

Discarding the two-dimensional coordinates of the pixel of the three-dimensional display image located in the area to be cropped The voxel, and its corresponding surface element;

Re-image the surface based on the remaining voxels or re-image the image based on the retained voxels.

The method provided by the embodiment of the invention implements cropping of the three-dimensional image. Moreover, since the corresponding two-dimensional coordinates of each voxel of the object in the two-dimensional coordinate system of the screen are acquired, the pixel of the three-dimensional display image located in the area to be cropped can be discarded according to the corresponding relationship between the voxel and the two-dimensional coordinates obtained above. The voxel corresponding to the two-dimensional coordinates and its corresponding surface element. Since the positional relationship of the pixels is compared in the two-dimensional coordinate system, the amount of calculation is small, thereby improving the cutting speed and realizing rapid three-dimensional image cropping in real time.

DRAWINGS

FIG. 1 is a flowchart of a method according to an embodiment of the present invention;

2 is a flowchart of a method for performing image rendering according to an embodiment of the present invention;

FIG. 3 is a flowchart of a method for performing a projection transformation on a voxel according to an embodiment of the present invention;

4 is a flowchart of a method for determining whether a pixel of a three-dimensional display image is in a closed curve according to an embodiment of the present invention;

FIG. 5 is a flowchart of another method according to an embodiment of the present invention.

detailed description

Before describing the embodiments of the present invention in detail, first, some of the nouns involved in the embodiments of the present invention are explained.

An object is a three-dimensional object displayed by a three-dimensional image. In the medical field, for example, it refers to a certain organ, such as the liver, heart, and the like.

The world coordinate system, the coordinate system defined in the space where the three-dimensional object is located, is a three-dimensional coordinate system. It is assumed that for a certain object, there are N slice images to form an image sequence. In the embodiment of the present invention, the upper left corner of the first slice image is defined as the coordinate origin of the world coordinate system, and the row direction of the slice image is the world coordinate system. The positive direction of the x-axis, the column direction of the slice image is the positive direction of the y-axis of the world coordinate system, the direction of the image sequence is the positive direction of the z-axis, and the unit of length is 1 pixel. In the embodiment of the present invention, the coordinates in the world coordinate system are in one-to-one correspondence with the coordinates in the coordinate system of the three-dimensional display image.

The screen two-dimensional coordinate system defines the upper left corner of the screen display area as the origin of the screen two-dimensional coordinate system, the row direction of the screen display area is the x-axis positive direction of the screen two-dimensional coordinate system, and the column direction of the screen display area is the screen two-dimensional The positive direction of the y-axis of the coordinate system is 1 pixel in length.

A voxel, a three-dimensional image data, can be obtained from a sequence of images of an object.

Facet, another kind of three-dimensional image data, can be obtained from the image sequence of the object, and there is a correspondence between the voxels and the noodle at the same coordinates in the world coordinate system.

3D display image, 3D effect image displayed on the screen.

FIG. 1 is a schematic diagram of a method for cropping a three-dimensional image according to an embodiment of the present invention. The specific implementation manner includes the following operations:

Step 100: Perform image surface rendering according to the surface element of the object, obtain a three-dimensional display image of the object in a two-dimensional coordinate system of the screen, and obtain a corresponding two-dimensionality of each voxel of the object in the two-dimensional coordinate system of the screen. Coordinates; or, the image volume rendering is performed according to the voxels of the object, and the three-dimensional display image of the object in the two-dimensional coordinate system of the screen is obtained, and each of the voxels of the object corresponds to the two-dimensional coordinate system of the screen. The two-dimensional coordinates.

Performing image volume rendering or surface rendering will display the object in three dimensions on the screen.

In addition, the embodiment of the present invention does not limit the projection transformation mode adopted by the drawing process, and may adopt parallel projection or perspective projection.

Step 110: Determine a region to be cropped on the three-dimensional display image.

Step 120: Discard the voxels corresponding to the two-dimensional coordinates of the pixels of the three-dimensional display image located in the to-be-cut region, and the voxels corresponding to the voxels.

Step 130: Perform image surface rendering again according to the retained voxels or re-image the image according to the retained voxels.

This step displays the cropped object of the 3D image on the screen.

The method provided by the embodiment of the invention implements cropping of the three-dimensional image. Moreover, since the corresponding two-dimensional coordinates of each voxel of the object in the two-dimensional coordinate system of the screen are acquired, the pixel of the three-dimensional display image located in the area to be cropped can be discarded according to the corresponding relationship between the voxel and the two-dimensional coordinates obtained above. The voxel corresponding to the two-dimensional coordinates and its corresponding surface element. Since the positional relationship of the pixels is compared in the two-dimensional coordinate system, the amount of calculation is small, thereby improving the cutting speed and realizing rapid three-dimensional image cropping in real time. In addition, since the positional relationship is compared in the two-dimensional coordinate system of the screen, and the corresponding voxels and voxels are discarded according to the corresponding two-dimensional coordinates of each voxel of the object in the two-dimensional coordinate system of the screen, it can be seen that the corresponding region to be cropped is not required to be constructed. The three-dimensional region and the positional judgment of the three-dimensional image data and the three-dimensional region are directly performed in the three-dimensional coordinate system. Wherein, the above three-dimensional area needs to be constructed along the opposite direction of the projection direction of the volume rendering or the surface drawing, which is limited by the system computing capability, and only parallel projection can be adopted when performing volume rendering or surface rendering. Since the solution provided by the embodiment of the present invention does not need to construct a three-dimensional area corresponding to the area to be cropped, it is not necessary to define a projection mode for performing volume rendering or surface rendering, and either a parallel projection or a perspective projection may be adopted. Therefore, the appropriate projection method can be selected according to actual needs. In addition, if the perspective projection is used, the visual features of the human eye "nearly large and small" can be simulated, resulting in a better user experience.

The respective steps of the embodiments of the present invention are described in detail below.

The specific implementation of step 100:

There are various implementations of the image surface rendering according to the surface of the object or the image volume rendering according to the voxel of the object. The present invention does not limit this. For example, the process shown in FIG. 2 is taken as an example. As shown in FIG. 2, step 100 specifically includes the following operations:

Step 200, constructing a world coordinate system;

Step 210: Perform model transformation on the three-dimensional image data (voxel or voxel), and record the model transformation matrix used.

Specifically, it is rotated, translated, and scaled, similar to where a three-dimensional object is placed in the scene.

Step 220: Perform projective transformation on the model-converted three-dimensional image data, and record The projection transformation matrix used.

In this step, both parallel projection and perspective projection can be performed.

Step 230: Display the data obtained by the model transformation and the projection transformation on a specified area of the screen window, and record the viewport transformation matrix.

It should be noted that the above process is applicable to both volume rendering and surface rendering.

Wherein, if the volume is drawn, the model transformation, the projection transformation and the viewport transformation are sequentially performed according to the voxels; if the surface rendering is performed, the model transformation, the projection transformation and the viewport transformation are sequentially performed according to the surface elements.

The above steps 200 to 230 may be, but are not limited to, implemented on an existing image rendering platform (eg, OpenGl platform).

If surface rendering is performed, correspondingly, there are various ways to obtain the corresponding two-dimensional coordinates of the voxels in the two-dimensional coordinate system of the screen.

For example, a voxel of the above object is subjected to projection transformation to obtain a corresponding two-dimensional coordinate of each voxel in a two-dimensional coordinate system of the screen. That is, by mapping the process of projection transformation, the mapping relationship between the world coordinate system and the screen two-dimensional coordinate system is established. This implementation can be used regardless of volume rendering or surface rendering. Especially when using surface rendering, the implementation can be adopted because the corresponding two-dimensional coordinates of the voxels in the two-dimensional coordinate system of the screen cannot be obtained during the surface rendering process. The implementation manner is as shown in FIG. 3, and specifically includes the following operations:

Step 300: Obtain three-dimensional coordinates of each voxel of the object in the world coordinate system.

Step 310: Multiply each of the three-dimensional coordinates by a model transformation matrix, a projection transformation matrix, and a viewport transformation matrix, to obtain two-dimensional coordinates corresponding to each voxel of the object in the two-dimensional coordinate system of the screen.

Among them, the model transformation matrix, the projection transformation matrix and the viewport transformation matrix are projection parameters.

It should be noted that the model transformation matrix, the projection transformation matrix, and the viewport transformation matrix as projection parameters are merely examples and are not limiting.

The manner in which the projection parameters are acquired in the embodiment of the present invention is not limited. For example, the projection parameters used in the image 100 surface rendering or volume rendering process of step 100 can be utilized. Existing image rendering technology, can be right It is displayed in a complete and accurate manner as a three-dimensional display image, thanks to the projection parameters used in the process. Therefore, by using the projection parameters used in the image drawing process to simulate projection, each voxel can be accurately mapped to the two-dimensional coordinate system of the screen, ensuring that the coordinates of the voxels and the visual three-dimensional objects are always one-to-one, ensuring The accuracy of the subsequent cutting process.

If image volume rendering is performed, the corresponding two-dimensional coordinates of each voxel of the object in the two-dimensional coordinate system of the above-mentioned screen can be obtained by using the projection transformation in the volume rendering process. The existing image rendering technology can display the object completely and accurately as a three-dimensional display image, thanks to the projection parameters used in the process. Therefore, the accuracy of the subsequent cropping process is ensured by using the mapping relationship of each voxel obtained in the image drawing process accurately in the two-dimensional coordinate system of the screen.

In order to further improve the processing speed, in the above step 100, if the image surface rendering is performed, the CUDA acceleration method may be used to acquire the corresponding two-dimensional coordinates of each voxel of the object in the two-dimensional coordinate system of the screen. For example, it is implemented by GPU acceleration.

The specific implementation of step 110:

There are several ways to determine the area to be cropped. For example, the closed curve of the area to be cropped selected by the user on the three-dimensional display image is obtained according to the mouse movement trajectory, and correspondingly, the area to be cropped refers to the area within the closed curve. For another example, the closed curve selected by the user on the three-dimensional display image is obtained according to the mouse movement trajectory, and then the inverse selection operation is performed. Correspondingly, the area to be cropped is the area outside the closed curve. and many more.

Since the mouse movement trajectory has a discontinuous point, that is, a non-closed curve, cubic spline interpolation is required. Even if the manually selected area is not closed, the interpolation will be automatically performed to form a closed curve, which will not affect the subsequent judgment. E.g:

Suppose there are n+1 data nodes (x ₀ , y ₀ ), (x ₁ , y ₁ ), (x ₂ , y ₂ )...(x _n , y _n )

Calculating the step size h _i =x _i+1 -x _i (i=0,1,...,n-1);

Substituting the data node and the specified first endpoint condition into the matrix equation;

Solving the matrix equation and obtaining a quadratic differential value m _i , the matrix being a three-diagonal matrix;

Calculate the coefficients of the spline curve:

a _i =y _i

Create an equation in each subinterval x _i ≤ x ≤ x _i+1 :

g _i (x)=a _i +b _i (xx _i )+c _i (xx _i ) ² +d _i (xx _i ) ³

The specific implementation manner of forming a closed curve by cubic spline interpolation can refer to the existing implementation manner, and the present invention will not be described again.

The specific implementation of the above step 120:

Before the step 120 is performed, the method provided by the embodiment of the present invention further includes the step of determining. That is, it is determined whether the pixel of the three-dimensional display image at each of the above two-dimensional coordinates is within the area to be cropped.

Wherein, if the step 110 is specifically to obtain a closed curve of the area to be cropped selected by the user on the three-dimensional display image, or to reverse the closed curve selected by the user, the area to be cropped is determined. Then, in the step of determining, specifically determining whether the pixel of the three-dimensional display image at each of the two-dimensional coordinates is within the closed curve. Taking the example shown in FIG. 4 as an example, a specific implementation manner of the determining step will be described. As shown in FIG. 4, the following operations are specifically included:

Step 400: Construct a rectangular image of the same size as the display area of the three-dimensional display image in the two-dimensional coordinate system of the screen.

The pixel value of all rectangular image pixel points of the rectangular image is a first pixel value.

For example, the pixel value of all matrix image pixel points of the matrix image is set to zero.

Since the matrix image is established in the two-dimensional coordinate system of the screen, the matrix image pixel points of the matrix image are in one-to-one correspondence with the pixels of the three-dimensional display image.

For example, the size of the display area of the three-dimensional display image is 100 pixels*100 pixels, and then the size of the matrix image is also 100 pixels*100 pixels.

Step 410: The two-dimensional corresponding to the closed curve (ie, the area to be cropped) in the rectangular image. The pixel value of the rectangular image pixel at the coordinate is modified to the second pixel value to obtain a contour curve.

The value of the first pixel value is different from the value of the second pixel value.

Specifically, the contour of the closed curve is searched for by the contour detection, and mapped to a corresponding area of the rectangular image, and the pixel value of the corresponding area is changed to a second pixel value, for example, 255, to obtain a contour curve.

Step 420: Determine whether each rectangular image pixel in the rectangular image is within the contour curve.

Specifically, the comparison is judged by the ray method. If the number of intersections of the ray and the contour curve drawn by the pixel of one rectangular image is an even number, it means that the pixel of the rectangular image is outside the contour curve. If the number of intersections of the ray and the contour curve drawn by the pixel of one rectangular image is an odd number, then Indicates that the rectangular image pixel is inside the contour curve.

In order to further improve the processing speed, the above determining step may be specifically implemented by using a CUDA acceleration method. For example, it is implemented by GPU acceleration.

The specific implementation of step 130:

The projection transformation method used in the drawing process is not limited, and either a parallel projection or a perspective projection can be used.

It should be noted that the specific implementations of the various steps described above can be coordinated.

Based on any of the foregoing method embodiments, before performing image rendering on the three-dimensional image data of the object, the method further includes:

The image sequence including the object is segmented to obtain voxels and voxels of the three-dimensional image of the object.

Specifically, the image sequence including the object is segmented to obtain a target image sequence, and the voxel and the voxel of the object are obtained according to the target image sequence.

By performing the segmentation, since the three-dimensional coordinates of the plurality of objects do not interfere with each other, the plurality of objects displayed in the image sequence can be segmented into different object image sequences, and then for different segmentation purposes and application scenarios, simultaneously for a single or Multiple objects are image-cut in parallel to increase processing speed. Similar to the simulated surgical procedure, the scalpel can remove the lesion location of an organ tissue, and at the same time Remove the location of adjacent organ tissue lesions. The cutting object can be carried out according to actual needs, which is a very important step in the simulation surgery, and it is also in line with the actual situation.

The method provided by the embodiment of the present invention is described in detail below by taking a specific application scenario as an example. As shown in FIG. 5, the following operations are specifically included:

In step 500, the image sequence of the chest cavity obtained by CT scanning is segmented to obtain a liver image sequence, and the liver voxels and surface elements are obtained according to the liver image sequence.

Among them, the image sequence of the chest can be, but is not limited to, obtained from DICOM.

It should be noted that, in step 500, the segmentation can not only obtain the liver image sequence, but also obtain image sequences of other organs, and the image sequences of different organs are separately stored. This embodiment is exemplified only by the cropping of the liver image sequence.

If the image sequence of multiple organs is segmented and these organs require three-dimensional image cropping, three-dimensional image cropping of these organs can be simultaneously achieved.

Step 510: Construct a world coordinate system.

Suppose there are N liver image sequences, with the upper left corner of the first liver image as the coordinate origin, the row direction as the positive X-axis direction, the column direction as the positive direction of the Y-axis, and the direction of the image sequence as the positive direction of the Z-axis to construct the world coordinate system. .

Step 520: Perform model transformation on the voxel or noodle of the liver, and record the model transformation matrix used.

The so-called model transformation, that is, the rotation, translation and scaling of a three-dimensional object model (constructed by voxels or voxels), similar to the appropriate position of a three-dimensional object placed in the scene. The model transformation matrix used for model transformation is called modeMatrix.

Step 530: Perform projective transformation on the model transformed data, and record the projection transformation matrix used.

The so-called projection transformation, taking parallel projection as an example, the projected view body is a rectangular flat pipe, no matter how far the object is, the size and size of the object after projection are unchanged. The projection transformation matrix used for projection transformation is called projMatrix.

The above is exemplified by the parallel projection. The embodiment can also adopt a perspective projection, and the specific implementation manner thereof will not be described again.

Step 540: Display the three-dimensional display image of the liver obtained by the model transformation and the projection transformation on a specified area of the screen window, and record the viewport transformation matrix (also referred to as a viewport vector).

Specifically, after the model transformation and the projection transformation, it is also necessary to undergo a cropping transformation, that is, framing. The three-dimensional display image of the liver after the framing is displayed in a designated area of the screen window. Suppose the coordinates of each voxel or voxel are objcoor[objx, objy, objz, 1.0], multiply the model view matrix modelMatrix, the projection transformation matrix projMatrix, and finally use the viewport transformation matrix to represent the three-dimensional display image pixels in homogeneous order. Limited to the current viewport.

It should be noted that steps 510 to 540 are processes for image rendering, and data format conversion (conversion into STL format) is required before image rendering.

Step 550: Obtain three-dimensional coordinates of each voxel of the liver in a world coordinate system.

Step 560: Multiply the three-dimensional coordinates of each voxel of the liver by the model transformation matrix, the projection transformation matrix, and the viewport transformation matrix to obtain corresponding two-dimensional coordinates of each voxel of the liver in the above-mentioned screen two-dimensional coordinate system.

Step 570: Acquire a closed curve of the to-be-cut region selected by the user on the three-dimensional display image of the liver according to the mouse movement track.

Step 580: Construct a rectangular image of the same size as the display area of the liver three-dimensional display image in the two-dimensional coordinate system of the screen.

Step 590: Modify a pixel value of a rectangular image pixel point at a two-dimensional coordinate corresponding to the closed curve in the rectangular image to a second pixel value to obtain a contour curve.

Step 5100: Determine whether each rectangular image pixel in the rectangular image is within the contour curve.

Step 5110: Discard the voxels corresponding to the two-dimensional coordinates of the three-dimensional display image pixel located in the closed curve, and the voxels corresponding to the voxels.

Step 5120: Re-image the image according to the retained liver voxel or facial layer.

Those skilled in the art will appreciate that embodiments of the present invention can be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or a combination of software and hardware. Moreover, the invention can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) including computer usable program code.

The present invention has been described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (system), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or FIG. These computer program instructions can be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing device to produce a machine for the execution of instructions for execution by a processor of a computer or other programmable data processing device. Means for implementing the functions specified in one or more of the flow or in a block or blocks of the flow chart.

The computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing device to operate in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture comprising the instruction device. The apparatus implements the functions specified in one or more blocks of a flow or a flow and/or block diagram of the flowchart.

These computer program instructions can also be loaded onto a computer or other programmable data processing device such that a series of operational steps are performed on a computer or other programmable device to produce computer-implemented processing for execution on a computer or other programmable device. The instructions provide steps for implementing the functions specified in one or more of the flow or in a block or blocks of a flow diagram.

While the preferred embodiment of the invention has been described, it will be understood that Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and the modifications and

It is apparent that those skilled in the art can make various modifications and variations to the invention without departing from the spirit and scope of the invention. Thus, it is intended that the present invention cover the modifications and modifications of the invention

Claims

A method for cropping a three-dimensional image, comprising:

Obtaining an image surface according to a surface element of the object, obtaining a three-dimensional display image of the object in a two-dimensional coordinate system of the screen, and acquiring a corresponding two-dimensional coordinate of each voxel of the object in the two-dimensional coordinate system of the screen; or And performing image volume rendering according to the voxel of the object, obtaining a three-dimensional display image of the object in a two-dimensional coordinate system of the screen, and corresponding two-dimensional corresponding to each voxel of the object in the two-dimensional coordinate system of the screen. coordinate;

Determining a region to be cropped on the three-dimensional display image;

Discarding the voxel corresponding to the two-dimensional coordinate of the pixel of the three-dimensional display image in the area to be cropped, and its corresponding surface element;

Re-image the surface based on the remaining voxels or re-image the image based on the retained voxels.
The method according to claim 1, wherein if the image surface rendering is performed according to a surface element of the object, a three-dimensional display image of the object in a two-dimensional coordinate system of the screen is obtained, and the acquiring each of the voxels is The corresponding two-dimensional coordinates of the two-dimensional coordinate system of the screen include:

Performing a projection transformation on the voxels of the object, and acquiring corresponding two-dimensional coordinates of each voxel of the object in the two-dimensional coordinate system of the screen.
The method according to claim 2, wherein the voxel of the object is subjected to projection transformation, and corresponding two-dimensional coordinates of each voxel of the object in the two-dimensional coordinate system of the screen are obtained, including :

Obtaining three-dimensional coordinates of each of the voxels in a world coordinate system;

Each of the three-dimensional coordinates is sequentially multiplied by a model transformation matrix, a projection transformation matrix, and a viewport transformation matrix to obtain two-dimensional coordinates corresponding to each voxel in the two-dimensional coordinate system of the screen.
The method according to claim 3, wherein the voxel of the object is subjected to projection transformation to obtain a corresponding two-dimensional coordinate of each voxel in the two-dimensional coordinate system of the screen, The method also includes:

Obtaining a model transformation matrix, a projection transformation matrix, and a viewport transformation matrix used for the image surface rendering process according to the facet of the object.
The method according to any one of claims 1 to 4, wherein the voxel corresponding to the two-dimensional coordinates of the three-dimensional display image pixel located in the area to be cropped is discarded, and the corresponding facet is The method also includes:

Determining whether the three-dimensional display image pixel point at each of the two-dimensional coordinates is within the to-be-cut region.
The method according to claim 5, wherein the determining whether the pixel of the three-dimensional display image at each of the two-dimensional coordinates is within the area to be cropped comprises:

Constructing a rectangular image of the same size as the display area of the three-dimensional display image in the two-dimensional coordinate system of the screen, wherein the pixel values of all rectangular image pixel points of the rectangular image are the first pixel value;

Modifying a pixel value of a rectangular image pixel point at a two-dimensional coordinate corresponding to the to-be-cut region in the rectangular image to a second pixel value, to obtain a contour curve;

Determining whether each rectangular image pixel in the rectangular image is within the contour curve.
The method according to claim 2, wherein the voxel of the object is subjected to projection transformation, and corresponding two-dimensional coordinates of each voxel of the object in the two-dimensional coordinate system of the screen are obtained, including :

The voxels of the object are subjected to projection transformation using a CUDA acceleration method to obtain corresponding two-dimensional coordinates of each voxel of the object in the two-dimensional coordinate system of the screen.
The method of claim 5, wherein determining whether the pixel of the three-dimensional display image at each of the two-dimensional coordinates is within the area to be cropped comprises:

The CUDA acceleration mode is used to determine whether the pixel of the three-dimensional display image at each of the two-dimensional coordinates is within the area to be cropped.
The method according to any one of claims 1 to 4, wherein before the image surface rendering according to the facet of the object or before the image volume rendering according to the voxel of the object, the method further comprises:

A sequence of images containing the object is segmented to obtain voxels and venopes of the object.
The method according to any one of claims 1 to 4, wherein the determining the area to be cropped on the three-dimensional display image comprises:

A closed curve of the region to be cropped selected by the user on the three-dimensional display image is determined.