A method, medium, and apparatus for vesicle identification and analysis

Abstract

This invention proposes a method, medium, and device for identifying and analyzing bubble-like molecules, belonging to the technical field of molecular cloud structure identification in the Milky Way. First, the data cube is segmented and its velocity integrated at multiple signal-to-noise ratio levels and multiple velocity integration scales. Based on the topological features of the closed cavity, the internal region of the bubble is detected and mapped back to the PPV space for weighted accumulation, constructing a bubble weight field. Subsequently, weighted clumps with PPV coherence are segmented and merged in the bubble weight field as candidate bubble cavities, and the inner boundary reference ellipse between the weighted centroid and the fixed center is calculated. A constrained double Gaussian model is used to fit the bubble radius, shell thickness, and symmetry score, and a turbulent normalized expansion significance is constructed. Finally, the shell strength skeleton is extracted and back-projected to fit the skeleton geometry to characterize shell continuity and deformation. This invention has the advantages of high automation, consistent parameter definitions, strong physical interpretability, and applicability to large-scale survey data.

Description

Technical Field

[0001] This invention belongs to the field of molecular cloud structure identification technology in the Milky Way, specifically relating to a method, medium and equipment for identifying and analyzing bubble-like molecules. Background Technology

[0002] Star formation and feedback from massive stars generate expanding cavities, shells, and large-scale bubble structures in molecular clouds. Molecular bubbles typically appear as ring-like enhancements and central defects in two-dimensional velocity integral plots, but the physical reality of bubbles is inherently three-dimensional: spatially, they should have relatively closed shells and internal cavities; in the velocity dimension, they should maintain morphological consistency within a finite velocity range; and in position-velocity (PV) diagnostics, they should exhibit organized velocity structures (e.g., expanding V-shaped / separated red-blue side components, or velocity shift patterns with consistent signs).

[0003] In existing technologies, molecular bubble identification often relies on manual screening or semi-automatic algorithms based on two-dimensional morphology (velocity integral maps, channel maps); while kinematic consistency (PV features) is often manually judged after identification. Due to the crowded, heavily superimposed, and strong velocity gradients of molecular cloud structures in the Milky Way plane, two-dimensional ring-shaped morphologies may be caused by projections of filamentary structures, superposition of unrelated velocity components, or large-scale velocity gradients. This leads to: difficulty in automatically and objectively establishing a bubble catalog in large-sample sky survey data; difficulty in simultaneously satisfying the unified judgment of "morphological closure" and "velocity consistency"; and a fragmented identification and analysis process, making it difficult to output comparable and uniformly defined geometric and kinematic indices.

[0004] Therefore, there is an urgent need for a molecular bubble identification and analysis method that can work natively in the PPV space, encode multi-scale morphological evidence and kinematic coherence in a unified manner, and output reproducible morphological-kinematic indices. Summary of the Invention

[0005] This invention addresses the problems in existing molecular bubble identification technologies, such as the susceptibility of two-dimensional morphology to artifacts, the lack of three-dimensional continuity, the difficulty in automation and scalability, and the inconsistency of certification indicators. It provides a method, medium, and equipment for identifying and analyzing bubble-like molecules.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] In a first aspect, the present invention provides a method for identifying and analyzing bubble-like molecules, comprising the following steps: Step 1: Based on the position-position-velocity PPV raw data obtained from molecular spectral line observations, the FacetClumps algorithm is used to detect the effective signal region and molecular gas clumps; Step 2: Based on a single signal region in the effective signal region, the velocity axis is divided into multiple centralized integration slices at multiple velocity integration scales. A two-dimensional velocity integration map is generated for each slice, and cavity topology detection is performed on the two-dimensional velocity integration map at multiple signal-to-noise ratio levels to obtain the set of voxels inside the cavity at each level, scale, and slice. Step 3: Perform weighted accumulation of the voxel set inside the cavity across signal-to-noise ratio levels, velocity integral scales, and segment locations to construct a bubble weight field; Step 4: Extract the candidate cavity region cluster set in the bubble weight field, and merge the repeated or overlapping parts to obtain the weighted clusters used to represent the candidate molecular bubbles. The weighted clusters and molecular gas clusters are overlapped and matched in the PPV space. The matched molecular gas clusters constitute the cavity gas. Step 5: Calculate the weighted centroid for each weighted clique and fit a reference ellipse; Step 6: Based on the reference ellipse, calculate the bubble radius, shell thickness, and outer radius of the shell, and simultaneously calculate and output the symmetry fractions of the near and far sides of the shell. Step 7: Construct PV slices along the same multi-directional direction as the radial strength profile, decompose the slices, calculate the expansion velocity amplitude, and combine the outer radius of the shell to calculate and output the significance of turbulent normalized expansion; Step 8: Unfold the velocity integral intensity map of the cavity gas in the coordinate system of orientation and bubble radius, extract the shell intensity skeleton from the unfolded map and perform back projection. The unfolded width is the shell thickness. Then, perform ellipse fitting on the shell intensity skeleton points and output the geometric parameters of each candidate molecular bubble.

[0008] Optionally, in step 2, during the process of dividing the velocity axis into multiple centralized integral segments, the center position of the segment is... Determine by the following formula:

[0009] in, This is the offset. Indicates the first j A velocity integral scale K j This represents the maximum number of fragments that can be partitioned. Multiple velocity integral scales The value range is from the original speed channel width of the raw data. Starting from, expand to the preset maximum integration scale. .

[0010] Optionally, in step 3, the bubble weight field The definition of is:

[0011] in, Indicates the first i Each signal-to-noise ratio level Indicates the first j A velocity integral scale As an indicator function, when voxels The value is 1 if it belongs to the voxel set inside the cavity, and 0 otherwise. k Indicates the location of the slice.

[0012] Optionally, in step 4, the merging of the candidate cavity region clusters is achieved through... The projection footprint of the plane is performed, and the projection footprint is... Defined as:

[0013] And calculate the candidate cavity region and Asymmetric overlap rate :

[0014] when Greater than the preset threshold and exist When a plane maintains a single connected component, the candidate region is... Merge into We obtain the weighted clusters.

[0015] Optionally, the specific process of step 5 is as follows: Calculate the weighted centroid for each weighted clique. :

[0016] in, This represents the set of voxels corresponding to the weighted clique. Representing voxels; and in Fixed center in plane Fit the inner boundary reference ellipse.

[0017] Optionally, in step 6, the bubble radius, shell thickness, and outer radius are calculated as follows: Based on the reference ellipse, a set of radial intensity profiles is constructed along multiple directions on the velocity integral diagram of the cavity gas to obtain the average intensity profile. The model was fitted using the following constrained double Gaussian model:

[0018] in, Substitution , , and These are the positions, widths, and amplitudes of the two peaks, respectively. For local baseline terms; Bubble radius obtained from fitting parameters Shell thickness With outer radius :

[0019]

[0020]

[0021] in, , ; The symmetry fractions of the shell's proximal and distal sides are calculated as follows: The symmetry of the shell's proximal and distal sides is calculated using the following symmetry scoring function:

[0022] in, Representing the symmetry scoring function, respectively... Substituting the amplitude, peak position, and peak width of the double Gaussian, we obtain three symmetry scores and take the average as the symmetry score of the near and far sides of the shell.

[0023] Optionally, in step 7, the expansion velocity amplitude is calculated as follows: The slice is decomposed into redshift and blueshift components relative to the system velocity, and the expansion velocity amplitude is calculated. :

[0024] in, x The offset relative to the system velocity, the system velocity The average velocity of the gas in the cavity. The outer radius of the shell, and These represent the redshift and blueshift components, respectively. The significance of the turbulent normalized expansion is calculated as follows:

[0025] in, For the significance of turbulent normalized expansion, With center velocity, For the speed of the enclosure environment, For environmental velocity diffusion, In order to be in Use the average velocity curve fitting formula within the range The coefficients of the quadratic term obtained from the fitting.

[0026] Optionally, in step 8, the extraction of the shell strength skeleton includes constructing a continuous ridge tracing based on graph search in the azimuth-bubble radius unfolded map, where the edge weights between pixel nodes are determined by the following formula:

[0027] in, For pixels i , j The distance between, I The strength is obtained by locally regularizing the unfolded graph; the strength skeleton is obtained through the minimum cost path. And project back to The plane forms the shell geometry description defined by the emission.

[0028] In a second aspect, the present invention provides a computer-readable storage medium storing a computer program that causes a computer to perform the bubble molecule identification and analysis method as described in the first aspect.

[0029] Thirdly, the present invention provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, it implements the bubble molecule identification and analysis method as described in the first aspect.

[0030] The beneficial effects of this invention are: (1) PPV native: Bubble evidence does not rely on a single two-dimensional projection, but directly encodes the cavity evidence into a three-dimensional weighted field, naturally constraining speed coherence; (2) Multi-scale accumulation: Enhance the real cavity and suppress accidental projection pseudo-loops by using multiple signal-to-noise ratio levels, multiple velocity integral widths and multiple piecewise accumulation; (3) Divisible and quantifiable: The weighted blocks provide a set of PPV cavity voxels that can be directly divided; subsequent outputs include unified indicators such as radius, thickness, symmetry, and skeleton geometry; (4) High degree of automation: It is suitable for large-scale sky survey cubes and can generate reproducible bubble catalogs and parameter tables; (5) Strong interpretability: Each step has a clear physical meaning (cavity evidence, shell photometry, PV organization, intensity skeleton continuity), which facilitates astrophysical interpretation and sample comparison. Attached Figure Description

[0031] Figure 1 This is a flowchart illustrating a method for identifying and analyzing vesicular molecules according to the present invention.

[0032] Figure 2 This is a MWISP three-color composite image and a schematic diagram of the "eye-shaped cavity".

[0033] Figure 3 It is a spatial distribution map of the weight field and weight clusters.

[0034] Figure 4 It is a triorthogonal projection of a single weighted clique.

[0035] Figure 5 It is a graph of gas mass correlation and radial / PV sampling geometry.

[0036] Figure 6 This is a schematic diagram of the channel / slice integral image sequence and the superposition of reference ellipses.

[0037] Figure 7 This is a schematic diagram of the radial intensity profile set and the double Gaussian fitting.

[0038] Figure 8 It is a statistical output graph of PV slices and velocity profiles, as well as the significance of expansion.

[0039] Figure 9 It is an unfolded ring, an unfolded skeleton extraction, and a back-projection fitting diagram.

[0040] Figure 10 This is a comparison diagram of sample distribution and infrared / radio emission. Detailed Implementation

[0041] The technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings.

[0042] In one embodiment, the present invention proposes a method for identifying and analyzing bubble-like molecules, the process of which is as follows: Figure 1 As shown, the input is a cube of molecular spectral line PPV data, and the output is a list of candidate bubbles and their morphological-kinematic parameters. The overall process includes the following steps: Step 1: Data preparation and determination of effective signal area (combined with...) Figure 2 ).

[0043] like Figure 2 As shown, the input PPV data cube has the following coordinates: ,in For longitude, For latitude, For speed channels. Estimate the noise per channel (e.g., RMS statistics through no-signal regions) to obtain the noise baseline required for subsequent SNR stratification. Determine the effective signal region (e.g., in several...) (Threshold-generated signal mask), and can divide the full cube into several local sub-cubes to reduce computation and reduce pollution from line-of-sight-independent emission.

[0044] Figure 2 The study revealed a typical complex region: a significant eye-shaped cavity (EoC) exists in the two-dimensional projection, but it may contain multiple PPV structures, which need to be identified and decomposed in subsequent steps.

[0045] Step 2: Multi-scale two-dimensional piecewise integration and topological cavity detection (combined) Figure 1 , Figure 3 , Figure 6 ).

[0046] The core of this step is: with different speed integral widths Layering with different signal-to-noise ratios The three-dimensional problem is transformed into multiple two-dimensional "hole (cavity) finding" problems, and then mapped back to PPV.

[0047] 2.1 Speed-based fragment construction Integral width for each velocity Define the fragment center as follows:

[0048] Each slice covers The speed range. Through multiple (From native channel width) To the maximum scale It achieves compatibility with both "narrow velocity coherent cavity" and "wide velocity coherent cavity".

[0049] 2.2 Generation of Two-Dimensional Piecewise Integral Maps and SNR Layering For each segment, the two-dimensional intensity map is obtained by integrating along the velocity. Then set up multiple levels. Threshold, will Convert to a binary / segmentation image and identify "closed cavities within ring structures" at each level.

[0050] 2.3 Topological Cavity Detection In each The next step is to find closed cavity regions that satisfy the condition of being "surrounded by a shell and having a relatively weak interior". In practice, this can be achieved using connected component and hole detection (e.g., through topological features such as closed contours or Eulerian characteristic numbers) to obtain the set of internal pixels of the cavity in a 2D image, and then mapping this set back to a set of 3D voxels within the corresponding piecewise velocity range. .

[0051] Figure 6 The sequence of channel / segment integral plots shows that the real shell often maintains a coherent pattern of "loops" and "internal losses" in multiple adjacent segments. This coherence will be enhanced in the next step through weight accumulation.

[0052] Step 3: Construct a 3D bubble weight field (combined) Figure 3 , Figure 4 ).

[0053] Step 2 in different ,different Different fragment positions The accumulated set of cavity voxels is used to construct a weight field:

[0054] in, This represents the cumulative strength of evidence that each voxel in the PPV belongs to the cavity. This represents the signal-to-noise ratio (SNR) layer value. A higher value indicates that the cavity is still visible under a more stringent threshold, and thus has higher credibility. It represents the velocity integral width. The smaller the width, the more the cavity can be established within a narrower velocity range, which is more in line with the physical intuition of tight velocity continuity. This is an indicator function used to accumulate the detected cavity voxels into the weight field. Figure 3 (a) is the molecular emission integral diagram of the main signal region, and 3(b) shows that the weighted integral diagram can highlight the cavity; Figure 4 This demonstrates that individual weighted clusters exhibit consistent PPV coherence across the three orthogonal directions, which forms the basis for subsequent segmentation and analysis.

[0055] Step 4: Weighted block segmentation and candidate merging (combined) Figure 3 , Figure 4 , Figure 5 ).

[0056] 4.1 Weighted Block Segmentation exist Extracting local maxima yields a set of candidate regions. Since multi-scale accumulation may result in multiple partially overlapping candidates for the same physical cavity, they need to be merged.

[0057] 4.2 Candidate Merging Rules (Spatial Projection Overlap and Connectivity Constraints) Define candidates in Projected footprint on the plane:

[0058] Calculate the asymmetric overlap rate:

[0059] when and exist When a plane maintains a single connected component, Merge into , to obtain weighted cliques ( Figure 3 (c)).

[0060] 4.3 Matching of Molecular Clusters with Weighted Clusters The weighted gas clusters are overlapped and matched with the molecular gas clusters detected by the FacetClumps algorithm in the PPV space. All matched molecular gas clusters constitute the cavity gas, such as... Figure 5 As shown in (a) and (b), calculate the average velocity of the gas in the complete cavity and treat it as the system velocity.

[0061] Step 5: Geometric reference ellipse (combined with...) Figure 5 ).

[0062] weighted clique voxel set Calculate the weighted centroid:

[0063] Extract the inner boundary of the clumping on a 2D projection and fit a reference ellipse, with the center of the ellipse fixed at [value missing]. Its equation is:

[0064] in They are the major and minor semi-axles, The rotation angle; The coordinates are the coordinates after rotation around the center.

[0065] Figure 5 (a) The cyan ellipse is the inner boundary reference ellipse, which provides a geometric reference for the unified sampling of the profile and PV. Figure 5 (b) is a hierarchically related extended gas structure.

[0066] Step 6: Radial strength profile and constrained double Gaussian fitting (combined) Figure 5 (c) Figure 7 ).

[0067] 6.1 Profile Construction and Shell Ridge Anchoring Sampling points are taken along the circumferential direction of the reference ellipse on the velocity integral intensity map of the cavity gas, and a radial tangent is constructed from the center outward. Figure 5 (c) Yellow line). To ensure the starting point of the profile stably falls on the brightest ridge of the shell, for each sampling direction, in The pixel with the highest local mean intensity within the neighborhood is selected as the anchor point, thereby reducing the impact of pixelation and slight geometric deviations.

[0068] Interpolate the profiles in each direction to a unified radial coordinate system. and averaged ( Figure 7 (Red line).

[0069] 6.2 Fitting and Parameter Output of the Double Gaussian Model A constrained double Gaussian model is adopted:

[0070] Among them, the two Gaussian components correspond to the bright peaks on both sides of the shell in the radial profile ( Figure 7 (two dashed lines) The location of the two peaks, For the peak width, For amplitude; This is the local baseline term (background).

[0071] The radius and thickness are obtained from the fitting:

[0072]

[0073] in, This represents the average distance from the two shell peaks to the central valley; Indicates the typical width of the shell; As the outer radius boundary of kinematic statistics ( Figure 7 , Figure 8 (The arrow in the image).

[0074] Symmetry scoring is based on:

[0075] Among them, the descriptive quantities for both sides of any shell (Such as amplitude, peak position, peak width), the closer the two sides are, The closer to 1, the greater the difference, and the lower the score. Multiple [items] can be evaluated separately. Calculate and average the results to determine the overall symmetry.

[0076] Step 7: PV slice construction, red-blue component decomposition, and dilation significance (combined with...) Figure 5 (c) Figure 8 ).

[0077] 7.1 PV slices and velocity profiles PV slices were extracted along a direction consistent with the radial profile. Figure 5 (c) Light red line Figure 8 (a) Calculate the intensity-weighted average velocity for each slice. With velocity dispersion ( Figure 8 (b) Black lines and shading). System speed. It can be obtained by intensity-weighted averaging of the surrounding gas associated with the cavity in the PPV. Figure 5 (a)(b) The associated cliques are used to support this estimate, serving as a reference for the red / blue component decomposition.

[0078] 7.2 Expansion Rate Amplitude outer radius Define the expansion velocity range within the range:

[0079] in, , For relative The average velocity of the redshift / blueshift components ( Figure 8 (b) Red / Blue line), take the maximum offset as the characteristic velocity amplitude of the bubble.

[0080] 7.3 Significance of Turbulent Normalized Expansion

[0081] Define two normalized diagnostic metrics:

[0082] in, Indicates the center position ( The average speed in the vicinity; This represents the average velocity within the shell's environmental region; It indicates the dispersion of environmental velocity and reflects the background of local turbulence; The intensity of the "depression / deviation" of the center velocity relative to the environment is represented by turbulence normalization; Is Internal The quadratic term obtained from the fitting describes whether the velocity profile exhibits a V-shaped / concave organized curvature; It is a measure of curvature, in Magnify and Normalization; final By combining amplitude comparison and shape consistency, the overall significance of the expansion structure is obtained. Figure 8 (c) Box output).

[0083] Step 8: Orientation-radius development and strength framework (combined) Figure 9 ).

[0084] To characterize the continuity and deformation of the shell in terms of orientation, this invention further expands the shell strength and extracts the skeleton.

[0085] 8.1 Definition of Expanding the Circumferential Zone by Set the expansion reference radius around the center. And correlate the annular width with the shell thickness:

[0086] Bilinear interpolation sampling of the two-dimensional intensity map within the annular zone yields the unfolded map. ( Figure 9 (a)(b)).

[0087] 8.2 Extracting Intensity Ridges (Intensity Skeleton) via Graph Search Treating pixels in the unfolded graph as nodes and connecting their neighborhoods, the edge weights are defined as follows:

[0088] in, For pixels i , j The distance between; I The local regularization strength is denoted by . Smaller weights indicate closer distances and higher strength. Given the weights of the minimum spanning tree, the optimal path is the one with the largest sum of its inverse weights, thus favoring the path along high-strength continuous ridges. This yields ridges that are continuous with orientation. ( Figure 9 (b) Red dot).

[0089] 8.3 Back projection and skeleton ellipse fitting Project the skeleton points back to The brightest ridge of the shell is obtained from the plane. Figure 9 (c) Red line), and fit the skeleton ellipse to the skeleton points ( Figure 9 (c) Cyan ellipse). The comparison between the skeletal ellipse and the cavity boundary ellipse can quantify non-ideal symmetries such as shell displacement, ellipticity variation, and orientation torsion.

[0090] Step 9: Sample Output and Instance Validation (combined) Figure 10 ).

[0091] The output includes the PPV range for each bubble candidate. clumping range), weighted peak value and voxel volume; geometric parameters ( , ); Symmetry index ( ); kinematic indicators ( , (and symbol consistency statistics); strength skeleton and skeleton ellipse parameters.

[0092] In another embodiment, the present invention provides a computer-readable storage medium storing a computer program that causes a computer to execute the bubble molecule identification and analysis method of the foregoing embodiments.

[0093] In another embodiment, the present invention provides an electronic device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, it implements the bubble molecule identification and analysis method of the foregoing embodiments.

[0094] In the embodiments disclosed in this application, a computer storage medium may be a tangible medium that may contain or store programs for use by or in conjunction with an instruction execution system, apparatus, or device. The computer storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of computer storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CDROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0095] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed in this application can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0096] The above are merely preferred embodiments of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should be considered within the scope of protection of the present invention.

Claims

1. A method for the identification and analysis of vesicular molecules, characterized in that, Includes the following steps: Step 1: Based on the position-position-velocity PPV raw data obtained from molecular spectral line observations, the FacetClumps algorithm is used to detect the effective signal region and molecular gas clumps; Step 2: Based on a single signal region in the effective signal region, the velocity axis is divided into multiple centralized integration slices at multiple velocity integration scales. A two-dimensional velocity integration map is generated for each slice, and cavity topology detection is performed on the two-dimensional velocity integration map at multiple signal-to-noise ratio levels to obtain the set of voxels inside the cavity at each level, scale, and slice. Step 3: Perform weighted accumulation of the voxel set inside the cavity across signal-to-noise ratio levels, velocity integral scales, and segment locations to construct a bubble weight field; Step 4: Extract the candidate cavity region cluster set in the bubble weight field, and merge the repeated or overlapping parts to obtain the weighted clusters used to represent the candidate molecular bubbles. The weighted clusters and molecular gas clusters are overlapped and matched in the PPV space. The matched molecular gas clusters constitute the cavity gas. Step 5: Calculate the weighted centroid for each weighted clique and fit a reference ellipse; Step 6: Based on the reference ellipse, calculate the bubble radius, shell thickness, and outer radius of the shell, and simultaneously calculate and output the symmetry fractions of the near and far sides of the shell. Step 7: Construct PV slices along the same multi-directional direction as the radial strength profile, decompose the slices, calculate the expansion velocity amplitude, and combine the outer radius of the shell to calculate and output the significance of turbulent normalized expansion; Step 8: Unfold the velocity integral intensity map of the cavity gas in the coordinate system of orientation and bubble radius, extract the shell intensity skeleton from the unfolded map and perform back projection. The unfolded width is the shell thickness. Then, perform ellipse fitting on the shell intensity skeleton points and output the geometric parameters of each candidate molecular bubble.

2. The method for identifying and analyzing vesicular molecules as described in claim 1, characterized in that: In step 2, the center position of the division in the process of dividing the velocity axis into a plurality of centering integration patches is determined as follows: wherein, is an offset, denotes the j th velocity integration scale, M j is the maximum number of fragments. Multiple velocity binning scales ranging from the native velocity channel width of the raw data up to a preset maximum binning scale .

3. The method for identifying and analyzing vesicular molecules as described in claim 1, characterized in that: In step 3, the bubble weight field is defined as: wherein, represents the i th signal-to-noise ratio level, represents the j th velocity integration scale, is an indicator function that takes the value 1 when the voxel belongs to the set of cavity interior voxels and 0 otherwise, k represents the position of the slice.

4. The method for identifying and analyzing vesicular molecules as described in claim 3, characterized in that: In step 4, the merging of the candidate cavity region blob set is performed by projecting the planar footprint of each blob onto the 3D model and determining the intersection of the projected footprint with the 3D model. The planar footprint of a blob is defined as: The planar footprint of a blob is defined as: and calculating a candidate cavity region and asymmetry overlap ratio : When greater than a preset threshold and In When the plane keeps a single connected component, the candidate region merge into , to get the weight block.

5. A method for the identification and analysis of vesicular molecules as claimed in claim 1, wherein: The specific process of step 5 is as follows: calculating a weight-weighted centroid for each weight tile : wherein, represents a set of voxels corresponding to a weight patch, represents a voxel; and in fixing the center in the plane fitting an inner boundary reference ellipse.

6. A method of vesicular molecular authentication and analysis according to claim 1, wherein: In step 6, the bubble radius, shell thickness, and outer radius are calculated as follows: Based on the reference ellipse, a set of radial intensity profiles along multi-azimuth directions is constructed on the velocity integral map of the cavity gas, and an average intensity profile is obtained and the following constrained bi-Gaussian model is used for fitting: wherein substituting , , and are the two peak positions, the peak width and the amplitude, respectively, is the local baseline term; Bubble radius from fit parameters Shell thickness Outer radius : wherein , ; The symmetry fractions of the shell's proximal and distal sides are calculated as follows: The symmetry of the shell's proximal and distal sides is calculated using the following symmetry scoring function: wherein, represents the symmetry score function, respectively, and The three symmetry scores are obtained by substituting the amplitude, peak position and peak width of the double Gaussian, and the average of the three symmetry scores is taken as the symmetry score of the near and far sides of the shell layer.

7. The method for identifying and analyzing vesicular molecules as described in claim 1, characterized in that: In step 7, the expansion velocity amplitude is calculated as follows: Decompose the slices into redshift and blueshift components relative to the systemic velocity, and calculate the expansion velocity amplitudes : wherein x is the offset relative to the system velocity, the system velocity is the average velocity of the cavity gas, is the outer radius of the shell, and denote the red- and blue-shift components, respectively; The significance of the turbulent normalized expansion is calculated as follows: in, For the significance of turbulent normalized expansion, With center velocity, For the speed of the enclosure environment, For environmental velocity diffusion, In order to be in Use the average velocity curve fitting formula within the range The coefficients of the quadratic term obtained from the fitting.

8. The method for identifying and analyzing vesicular molecules as described in claim 1, characterized in that: In step 8, the extraction of the shell strength skeleton includes constructing a continuous ridge tracing based on graph search in the azimuth-bubble radius unfolded map, and the edge weights between pixel nodes are determined by the following formula: in, For pixels i , j The distance between, I The strength is obtained by locally regularizing the unfolded graph; the strength skeleton is obtained through the minimum cost path. And project back to The plane forms the shell geometry description defined by the emission.

9. A computer-readable storage medium storing a computer program, characterized in that, The computer program causes the computer to perform the bubble molecule identification and analysis method as described in any one of claims 1-8.

10. An electronic device, characterized in that, include: The memory, the processor, and the computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the bubble molecule identification and analysis method as described in any one of claims 1-8.

Images