Digital materialization method and system for die casting fossil

Through CT scanning and deep learning technology, the problem of digitization and repair of molded fossils is solved, and efficient digitization and classification of decomposed fossils is realized, providing detailed three-dimensional model and distribution information for geological exploration, and supporting references for oil and gas exploration.

CN118918271AActive Publication Date: 2024-11-08INST OF VERTEBRATE PALEONTOLOGY & PALEOANTHROPOLOGY CHINESE ACAD OF SCI
View PDF 7 Cites 0 Cited by

Patent Information

Application Number
CN202411405675.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-10
Publication Date
2024-11-08
Estimated Expiration
2044-10-10

AI Technical Summary

Technical Problem

The existing technology is difficult to effectively digitize and repair molded fossils preserved in surrounding rocks, especially those whose original bones have been partially or completely dissolved, and cannot be repaired through traditional mechanical methods, making it difficult to obtain their original distribution and affecting geologists. Reference basis in oil and gas exploration.

Method used

CT scanning technology is used to obtain CT images of molded fossils, and the CT images are denoised, semantic segmentation and three-dimensional modeling are used to extract features, and the features are classified and repaired through 3D printing technology. Digital entity.

Benefits of technology

It realizes efficient and objective digitization of molded fossils, can obtain its internal structure and distribution without loss, provide detailed three-dimensional models and classification information, and provides geologists with an important reference basis for oil and gas exploration.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118918271A_ABST
    Figure CN118918271A_ABST
Patent Text Reader

Abstract

The invention provides a digital materialization method and system for die casting fossil, and belongs to the technical field of CT imaging. The method comprises the following steps: acquiring a CT image of a die casting fossil; generating a three-dimensional model of the die-cast fossil based on the CT image of the die-cast fossil; extracting local features of the three-dimensional model of the die-cast fossil, performing feature combination on the local features, and classifying the die-cast fossil according to the combined overall features; wherein the local features of the three-dimensional model of the die-cast fossil comprise a tumor point structure feature, a radiation ridge structure feature and a medullary cavity hole structure feature; and based on the three-dimensional model of the die-cast fossil and the classification result of the die-cast fossil, generating the digitally materialized die-cast fossil. According to the method, CT scanning, image denoising, semantic segmentation, three-dimensional modeling, fossil classification and 3D printing materialization can be efficiently and objectively carried out on the die-cast fossil.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the field of CT imaging technology, and in particular to a digital materialization method and system for molded fossils. Background Art

[0002] Paleontology is a discipline that studies the history of life on Earth based on fossils. Fossils refer to the remains, relics or traces of life of organisms from the geological history period preserved in rocks. Fossils can be divided into four categories according to their preservation characteristics: solid fossils, cast fossils, trace fossils and chemical fossils. Solid fossils are fossils formed by the preservation of all or part of the remains of ancient organisms themselves (especially the hard parts). Cast fossils are the impressions and casts of the remains of ancient organisms left in rock formations or surrounding rocks. Trace fossils refer to the traces of activities and relics of ancient organisms during their life preserved in rock formations. Chemical fossils refer to the organic substances such as amino acids and fatty acids formed after the decomposition of the organic components that make up the organisms under certain specific conditions, which are not preserved in the rock formation.

[0003] One type of cast fossils has its original skeleton partially or completely decomposed during the burial process, leaving only cavities preserved in the surrounding rock. These cavities preserve the original distribution of cast fossils in the surrounding rock, which is important information in paleontological research. In addition, statistics on the original distribution of cast fossils can dig out a lot of rich geological information, especially in oil and gas exploration. Geologists can provide important basis for finding and determining the potential direction of oil-producing rock formations by statistically analyzing the distribution of cast fossils in strata.

[0004] However, these cast fossils are usually hollow structures preserved in the surrounding rock, and traditional mechanical methods cannot be used to repair the fossils, so it is difficult to obtain the original distribution of the cast fossils in the surrounding rock. In recent years, thanks to the widespread use of micro-CT technology in non-destructive testing of fossils, paleontologists can obtain more comprehensive three-dimensional fossil information through micro-CT scanning, including three-dimensional microstructures from the inside to the surface of the fossil. At the same time, 3D printing technology is entering all walks of life and playing an increasingly important role, making it possible for paleontologists to combine CT scanning and 3D printing technology to replicate the original distribution of cast fossils, thereby providing a reference for geologists to conduct oil and gas exploration. Summary of the invention

[0005] In order to realize the digital entity of cast fossils whose original bones have been partially or completely decomposed, the present invention provides a digital entity method and system for cast fossils, which can efficiently and objectively perform CT scanning, image denoising, semantic segmentation, three-dimensional modeling, fossil classification and 3D printing entity on cast fossils.

[0006] The technical content of the present invention includes: A method for digital materialization of a molded fossil, the method comprising: Obtain CT images of the molded fossils; Generating a three-dimensional model of the mold-cast fossil based on the CT image of the mold-cast fossil; Extracting local features of the three-dimensional model of the mold-cast fossil, combining the local features, and classifying the mold-cast fossil according to the combined overall features; wherein the local features of the three-dimensional model of the mold-cast fossil include: tumor point structure features, radial ridge structure features, and medullary cavity foramen structure features; Based on the three-dimensional model of the cast fossil and the classification results of the cast fossil, a digitally materialized cast fossil is generated.

[0007] Further, generating a CT image of the molded fossil comprises: Cut the rock sample according to the set size to obtain the molded fossil; The X-ray detector is used to image the molded fossil at different angles in a 360° step-by-step manner; The images at each angle were reconstructed, and the filtered back-projection algorithm was used to obtain the CT images of the molded fossil.

[0008] Further, based on the CT image of the molded fossil, a three-dimensional model of the molded fossil is generated, including: De-noising the CT images of the molded fossils to obtain high signal-to-noise ratio CT images; Perform semantic segmentation on high signal-to-noise ratio CT images to obtain semantic segmentation results of molded fossils; The semantic segmentation results are used for isosurface extraction and surface volume rendering to generate a three-dimensional model of the cast fossil.

[0009] Furthermore, the denoising of the CT image of the die-cast fossil to obtain a high signal-to-noise ratio CT image includes: A filter is used to denoise the CT image of the molded fossil to obtain a CT image with a high signal-to-noise ratio; wherein the filter includes: a median filter or a mean filter; or, A CT image denoising network based on deep learning is used to denoise the CT image of the cast fossil to obtain a high signal-to-noise ratio CT image; wherein the CT image denoising network based on deep learning includes: a Noise2Noise network model or a Noise2Void network model.

[0010] Furthermore, the semantic segmentation of the high signal-to-noise ratio CT image to obtain the semantic segmentation result of the molded fossil includes: Get the grayscale mean of the molded fossil , grayscale mean of surrounding rock and the overall grayscale value of high signal-to-noise ratio CT images ; By calculating the optimal threshold The between-class variance , get the optimal threshold ;in, is the probability of a molded fossil, is the probability of surrounding rock; Applying the best threshold Performing binary image segmentation on the high signal-to-noise ratio CT image to obtain a binary mold-cast fossil CT image; After removing the external small particles and internal small gaps in the binarized cast fossil CT image, the seed points of the cast fossil are selected to merge adjacent pixels with similar attributes and combined with manual circle selection to group pixels with similar properties into semantic regions, thus obtaining the semantic segmentation result of the cast fossil.

[0011] Furthermore, the semantic segmentation results are subjected to isosurface extraction and surface volume rendering to generate a three-dimensional model of the molded fossil, including: Obtain the three-dimensional volume data corresponding to each semantic segmentation result; Find the voxels that intersect the isosurface in the three-dimensional volume data and obtain the voxel edges; Applying linear interpolation to solve the intersection of the isosurface and the voxel edge; All intersection points are connected to extract isosurfaces to generate a three-dimensional model of the molded fossil.

[0012] Furthermore, the extracting of features of the three-dimensional model of the cast fossil and classifying the cast fossil based on the features includes: Preprocessing the three-dimensional model of the molded fossil; The VGG16 network model of transfer learning is used to extract the overall features of the pre-processed three-dimensional model of the molded fossil; wherein the features of the pre-processed three-dimensional model of the molded fossil extracted by the VGG16 network model of transfer learning include: Using the front part of the network in the VGG16 network model, a low-level feature map of the three-dimensional model of the molded fossil is obtained, wherein the low-level feature map includes edge structure information of the three-dimensional model of the molded fossil; The low-level feature maps are combined using the middle part network in the VGG16 network model to obtain local features of the three-dimensional model of the molded fossil; The local features are combined by using the latter part of the network in the VGG16 network model to obtain the overall features of the three-dimensional model of the molded fossil; The long short-term memory architecture in the recurrent neural network model is used to fuse the overall features of multiple perspectives; The SVM multi-classifier is selected to classify the fused features to obtain the classification result of the mold cast fossil.

[0013] Further, the three-dimensional model of the molded fossil is preprocessed, including: The PCA method is used to normalize the posture of the three-dimensional model of the molded fossil; and, Using the mean curvature map to enhance the local geometric features of the three-dimensional model of the molded fossil; and, A multi-view rendering tool is used to collect multi-view views of the cast fossil three-dimensional model.

[0014] Further, based on the three-dimensional model of the cast fossil and the classification result of the cast fossil, a digitally materialized cast fossil is generated, which also includes: Repairing the three-dimensional model of the molded fossil, the repairing comprising: inverting reverse triangular facets, stitching gaps, filling holes, deleting bad edges, crossing and overlapping triangular facets, and interfering with shells; The repaired three-dimensional model of the cast fossil is imported into the 3D printing slicing software, and the surrounding rock is given a transparent material, and the cast fossil is given a color material based on the classification result of the cast fossil, so as to obtain a solid three-dimensional model of the cast fossil.

[0015] A digital materialization system for a molded fossil, the system comprising: An image acquisition module, used for acquiring a CT image of the molded fossil; A model generation module, used for generating a three-dimensional model of the molded fossil based on the CT image of the molded fossil; A fossil classification module is used to extract local features of the three-dimensional model of the die-cast fossil, combine the local features, and classify the die-cast fossil according to the combined overall features; wherein the local features of the three-dimensional model of the die-cast fossil include: tumor point structure features, radial ridge structure features and medullary cavity foramen structure features; The digital materialization module is used to generate digital materialized cast fossils based on the three-dimensional model of the cast fossils and the classification results of the cast fossils.

[0016] A storage medium stores a computer program, which implements any of the above methods when executed by a processor.

[0017] An electronic device comprises a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute any one of the above methods.

[0018] Compared with the prior art, the present invention has the following advantages: The present invention adopts an unsupervised learning method to denoise the CT images of cast fossils and their surrounding rocks to improve the signal-to-noise ratio even if it is impossible to obtain noise-free CT images of cast fossils. Without damaging the cast fossils, the anatomical structure and distribution of the cast fossils inside the surrounding rocks can be obtained, and the types and quantities of the cast fossils can be counted. By combining CT scanning and 3D printing technology, the surrounding rocks can be made transparent, clearly showing the real preservation scene of the cast fossils in the surrounding rocks, reproducing the burial positions of different types of cast fossils and different parts of the same type, and thus providing a reference for geologists in oil and gas exploration. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a flowchart of the method for digital materialization of a molded fossil provided in an embodiment of the present invention.

[0020] Figure 2 It is a flowchart of obtaining the semantic region of each molded fossil provided by an embodiment of the present invention.

[0021] Figure 3 It is a flowchart of obtaining the types and quantities of cast fossils provided by an embodiment of the present invention.

[0022] Figure 4 It is a pie chart of the types and quantities of molded fossils counted in the embodiments of the present invention.

[0023] Figure 5 This is a visual display diagram of the in-situ preservation of the molded fossil before and after transparent treatment in an embodiment of the present invention.

[0024] Figure 6 It is a block diagram of a method and device for digital materialization of a molded fossil provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0025] In order to make the purpose, scheme and advantages of the present invention more clearly understood, the present invention is further described in detail by taking the experiment conducted on the early vertebrate mold fossil as an example. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0026] The method for digitalizing cast fossils of the present invention includes the steps of acquiring cast fossil CT images, denoising cast fossil CT images based on deep learning, rendering the surface of cast fossil CT images based on stepping cubes, counting the number of cast fossil three-dimensional model types based on deep learning, and actualizing cast fossils.

[0027] Specifically, if Figure 1 As shown, the present invention has the following steps: Step 1: Obtain a CT image of the cast fossil.

[0028] The present invention uses micro-CT to obtain CT images of cast fossils according to the CT image acquisition standard of cast fossils, thereby reducing the influence of noise and artifacts of the CT images of cast fossils. The format of the CT images of cast fossils can be a DICOM file or a non-DICOM common image format BMP / TIF file.

[0029] In one example, the molded fossils were collected from vertebrate microfossils of the Early Devonian in Yunnan, China. By selecting some rock samples containing fossil-rich layers, a small cutting machine was used to cut the rock samples into a cuboid with a length of 1 cm, a width of 1 cm and a height of about 5 cm. The size of the cuboid was summarized based on the actual size of the molded fossil after repeated experimental verification. This size can not only ensure the high efficiency of fossil CT scanning, but also ensure that the fossil CT data has sufficient spatial resolution for creating a three-dimensional model of the molded fossil, meeting the needs of identifying the microstructure of the molded fossil. After mechanical processing of the molded fossil, the molded fossil can enter the microscopic CT scanning link, and a 225kV micro-industrial CT is used to obtain the CT image of the molded fossil. Among them, the parameters of the 225kV micro-industrial CT are set as follows: voltage: 100kV; current: 100μA; exposure time: 1000ms; number of superimposed frames: 2 frames; number of projection images: 1440; voxel size: 5.96μm; filter; 1mm thick aluminum.

[0030] Step 2: Generate a three-dimensional model of the cast fossil based on the CT image of the cast fossil.

[0031] Step 2.1: Denoise the CT image to obtain a high signal-to-noise ratio CT image.

[0032] Step 1 uses the GPU-accelerated FDK algorithm to convert the collected 1,440 projection images into 1,536 2048×2048 pixel tomograms. The CT images of the cast fossils are directly input into the trained CT image denoising network based on deep learning, that is, the Noise2Noise network model that saves the optimal weight coefficient, to obtain the CT images of the cast fossils with high signal-to-noise ratio, and to achieve denoising of the CT images of the cast fossils. The CT image denoising network uses an unsupervised Noise2Noise network model based on the UNet architecture. The Noise2Noise network model uses CT image pairs with different levels of noise and the L2 loss function to train the network model. Finally, the Noise2Noise network model that saves the optimal weight coefficient is used to generate a high signal-to-noise ratio CT image.

[0033] In one example, when training the Noise2Noise denoising network model, the deep learning-based denoising network for CT images of cast fossils is trained, verified and tested by inputting pairs of CT images of cast fossils with different levels of noise as experimental data sets. The original weight file of the pre-trained UNet architecture is loaded, and the L2 loss function is used to discriminate CT images of cast fossils with different levels of noise and perform back propagation. The deep learning-based denoising network for CT images of cast fossils is trained, the denoising accuracy of the Noise2Noise network model is adjusted, and the parameter weights of the optimized Noise2Noise network model are saved.

[0034] The Noise2Noise network model is based on the UNet architecture. The network model consists of three parts: encoder, decoder, and skip connection. The network model contains 18 convolutional layers with 3×3 convolution kernels and 1 convolutional layer with 1×1 convolution kernel, and the stride is set to 1. Except for the last convolutional layer, the remaining convolutional layers have Relu activation function. The encoder part can convert the input die-cast fossil CT image into a feature map with multiple channels. Its redundant information can be downsampled through a pooling layer with a stride of 2. After passing through 4 pooling layers, the feature map is input to the decoder part. The skip connection can connect the feature maps of the same level in the encoder and decoder, thereby improving the signal-to-noise ratio of the die-cast fossil CT image. The decoder part upsamples the feature map 4 times and then restores the feature map. Finally, the decoder outputs the denoised die-cast fossil CT image through a convolutional layer with a 1×1 convolution kernel.

[0035] Step 2.2: Perform semantic segmentation on the denoised CT image to obtain the semantic region of each molded fossil.

[0036] The present invention mainly adopts the cone-beam FDK algorithm of the filtered back-projection framework to realize the pre-reconstruction work of the cast fossil CT image, and applies the Noise2Noise network model based on the UNet architecture to reduce the noise of the cast fossil CT image.

[0037] This embodiment uses the Otsu method to confirm the intra-class variance of the cast fossils and surrounding rocks in the overall CT image after denoising, determines the optimal threshold value by traversing the maximum value of the intra-class variance, and performs the binarization image threshold segmentation of the overall CT image. The optimal threshold value can divide the grayscale histogram of the overall CT image into two categories (surrounding rocks and cast fossils) so that the variance between the two categories is maximized. The grayscale value of the overall CT image of this embodiment is 0-65535 levels, 16-bit grayscale image. In general, when the grayscale distribution histogram of the overall CT image is bimodal, the optimal threshold value is It should fall at the trough between the two peaks. The optimal threshold calculation formula for the overall CT image is as follows:

[0038] in, is the threshold value; When the threshold is The between-class variance at time ; , are the probabilities of molded fossils and surrounding rocks, respectively; and are the mean values ​​of grayscale of molded fossils and surrounding rocks, respectively; is the gray value of the overall CT image.

[0039] like Figure 2 As shown, applying the optimal threshold The whole CT image is binarized and segmented, and the die-cast fossil is selected as the region of interest. Morphological operation methods are used to remove small external particles and internal small gaps in the binarized die-cast fossil CT image through corrosion, expansion and other operations, and the main structure in the die-cast fossil CT image, that is, the partially or completely decomposed early vertebrate microfossils, is retained. The regional growing algorithm is used to select the seed point of the die-cast fossil to merge adjacent pixels with similar attributes and combine the manual circle selection method to group pixels with similar properties to form a semantic region, so as to achieve semi-automatic semantic segmentation of the die-cast fossil CT image.

[0040] Step 2.3: The semantic segmentation results are subjected to isosurface extraction and surface volume rendering to generate a three-dimensional model of the cast fossil.

[0041] The present invention applies a surface volume rendering method of a cast fossil CT image based on a stepping cube, performs isosurface extraction and surface volume rendering on the semantic segmentation result of the cast fossil CT image, and generates an equal-threshold triangular patch model of the cast fossil CT image. The stepping cube algorithm is one of the most commonly used surface volume rendering methods. By finding voxels that intersect with the isosurface in the three-dimensional volume data, the intersection points of the isosurface and the edges of the intersecting voxels are solved by linear interpolation, and finally all the intersection points are connected to extract the isosurface, and the equal-threshold triangular patch model of the CT image is generated. The stepping cube algorithm can extract voxels with the same grayscale threshold in the cast fossil CT image stack and connect them into cast fossil triangular patches in a topological form.

[0042] In one example, the visualization toolkit VTK is used to provide a support environment for generating a three-dimensional model of a cast fossil. The present invention uses the vtkMarchingCubes class in VTK to realize the surface volume rendering of the cast fossil CT image. The output cast fossil three-dimensional model file is usually in the format of STL, OBJ, PLY, etc. The cross-platform application tool Blender is used to optimize the cast fossil three-dimensional model, improve the rendering quality and visualization effect. The cast fossil three-dimensional model is composed of voxels containing color and measurement value information. The appearance of the three-dimensional model can be changed arbitrarily (such as color, transparency, etc.), and can be used for research on three-dimensional data visualization. At present, STL files are one of the most popular three-dimensional model format files for sharing. They can be read by three-dimensional data visualization software such as Meshlab or ImageJ, and can be applied to digital-material tasks such as 3D printing.

[0043] Step 3: extracting local features of the three-dimensional model of the cast fossil, combining the local features, and classifying the cast fossil according to the combined overall features.

[0044] The three-dimensional model of the molded fossil to be classified is input into the trained three-dimensional model classification network of the molded fossil based on deep learning, the classification result of the three-dimensional model of the molded fossil is obtained, the identification of the three-dimensional model of the molded fossil is realized, the identification result of the three-dimensional model of the molded fossil is output, and the types and quantity of the molded fossils are counted. This step mainly includes: three-dimensional model data preprocessing, feature extraction, feature fusion, modeling classification and type, quantity statistics and other aspects.

[0045] The 3D model data preprocessing includes posture normalization, mean curvature map calculation and multi-view rendering functions. The present invention adopts the PCA method to perform posture normalization on the cast fossil 3D model, applies the mean curvature map to enhance the local geometric morphological features of the cast fossil 3D model, and uses the multi-view rendering tool to collect multi-view views of the cast fossil 3D model.

[0046] The 3D model feature extraction is mainly implemented by using the transfer learning technology, using the VGG16 network model trained in the ImageNet dataset as the transfer network. Figure 3 As shown in the figure, the low-level feature map output by the front part of the migration network is such as the edge structure information of the cast fossil. The intermediate feature map in the middle part of the migration network is a recombination of low-level features. The intermediate features are close to the local properties of the cast fossil, such as the tumor points, radial ridges, medullary cavity holes and other structural features of the cast fossil. The high-level feature map output by the back part of the migration network is a recombination of the intermediate features to obtain the overall features. The overall features can highly summarize the overall properties of the cast fossil. Finally, the long short-term memory architecture in the recurrent neural network model is used to realize the multi-view overall feature fusion of the three-dimensional model.

[0047] The three-dimensional model modeling classification selects the support vector machine multi-classifier to classify the three-dimensional model modeling of the cast fossil. The support vector machine multi-classifier calculates a series of hyperplanes to distinguish the predicted class of the cast fossil three-dimensional model from other classes. The hyperparameters of each hyperplane are represented by w j , the set of hyperparameter vectors is represented as W={w1,w2,…,w k}, k represents the type of cast fossils in the experimental data set. The SVM classifier needs to calculate k hyperparameters to complete the classification task of the cast fossil 3D model. The quadratic optimization method is used to solve the unknown hyperparameter vector set W of the SVM multi-class classifier.

[0048] When training the classification network model, the main shape features of the cast fossils are combined, and the corresponding genus and species names are used to mark and classify the cast fossil 3D models to obtain genus and species name labels. The cast fossil 3D models and corresponding genus and species name labels are input as experimental data sets to train, verify and test the network model. The original weight file of the pre-trained improved MVCNN network model is loaded, and the classification results of the cast fossil 3D model are compared with the true labels, and the cast fossil 3D model classification network based on deep learning is trained through continuous regression. The accuracy of the cast fossil 3D model classification is adjusted, and the parameter weights of the optimized improved MVCNN network model are saved.

[0049] The invention provides a process for creating a classification network of a three-dimensional model of a molded fossil based on deep learning. The three-dimensional models of molded fossils in the training set, validation set and test set of the experimental data set are input into the data preprocessing module, and the posture normalization of the three-dimensional model of the molded fossil is ensured by translation, scaling and rotation operations. The local microstructure of the three-dimensional model of the molded fossil is enhanced by the average curvature, and the multi-view views of the three-dimensional model of the molded fossil are collected using an application developed in combination with OpenGL and OpenCV. The multi-view views of the three-dimensional model of the molded fossil are input into the improved MVCNN network model training network. First, the VGG16FT network model for feature extraction is trained, the number of iterations is set to 100, the optimizer is the Adam function, the batch size is set to 4, the initial learning rate is set to 1e-4, and the loss function of the VGG16FT network model for feature extraction is the cross entropy function. Then, the LSTM network model for multi-view overall feature fusion of the three-dimensional model of the fossil is trained, the number of iterations is set to 100, the optimizer is the Adam function, the batch size is set to 36, the initial learning rate is set to 1e-3, and the loss function of the LSTM network model for feature fusion is the cross entropy function. Finally, the SVM multi-classifier was trained, the number of iterations was set to 3500, and the regularization parameter was set to 0.1 to obtain the optimal deep learning-based three-dimensional model classification network for molded fossils. The three-dimensional model classification network for molded fossils based on deep learning was used to count the types and quantities of molded fossils in the surrounding rocks. The surrounding rocks mainly include three-dimensional models of bone fragments, teeth and scale fossils of five early ancient fish (jawless fish, armored fish, placoderms, chondrichthyes, acanthichthyes and bony fish). Among them, jawless fish (armored fish, 105; Asian paraflorid fish, 57; Liaokuo paraflorid fish, 100; Wang's paraflorid fish, 100), placoderms (placoderms, 106), bony fish (ghost fish, 145; Yang's fish, 163; spotted fish, 170), cartilaginous fish (Guo's fish, 888), acanthichthyes (acanthichthyes, 34; dorsal acanthichthyes, 142), the detailed statistics are as follows Figure 4 shown.

[0050] Step 4: Generate a digitally materialized cast fossil based on the three-dimensional model of the cast fossil and the classification results of the cast fossil.

[0051] Import the surrounding rock and the molded fossils as a whole into the 3D model optimization software Materials Magics or RapidForm XOV. In this embodiment, Materials Magics is used as an example to perform repair work on the 3D model by inverting reverse triangles, stitching gaps, filling holes, deleting bad edges, crossing and overlapping triangles, and interfering with shells. Import the repaired 3D model of the surrounding rock and the molded fossils into the slicing software GrabCAD print. The surrounding rock is given a transparent color material, such as Fullcure720 photosensitive resin. Different types of molded fossils can be given different colors, such as VeroBlue, VeroYellow, VeroCyan, VeroMagenta and VeroGray photosensitive resins. The direction and position of the 3D model on the tray are optimized to shorten the printing time. The scaling ratio of the 3D model is adjusted to achieve the best effect of observing the distribution of the molded fossils in the surrounding rock.

[0052] In summary, in order to solve the problem that geologists are unable to losslessly restore the true distribution of cast fossils buried in the strata, the present invention collects cast fossil CT images through micro-CT, uses the unsupervised Noise2Noise network model built on the UNet architecture to remove cast fossil CT image noise, and uses a semi-automatic method to complete the semantic segmentation of cast fossil CT images, uses a deep learning-based cast fossil three-dimensional model classification network to classify cast fossils and conduct type and quantity statistics, and uses 3D printing technology to realize the materialization of cast fossils. The present invention can intuitively show geologists the in-situ preservation of cast fossils in the strata from all-round and multi-angle perspectives such as digital visualization and materialization. Figure 5 The figure shows the in-situ preservation visualization of molded fossils before and after the surrounding rock is made transparent in the embodiment. The left figure sets the surrounding rock transparency threshold to 0, and the distribution of molded fossils inside cannot be displayed. The right figure sets the surrounding rock transparency threshold to 50%, showing the distribution of different types of molded fossils inside the surrounding rock. The present invention is expected to indicate potential directions for determining geological ages, dividing and comparing strata, and searching for mineral resources.

[0053] Please refer to Figure 6 , which shows a block diagram of a digital materialization system for molded fossils provided by an embodiment of the present invention. The device includes the following modules: an image acquisition module 610 , a model generation module 620 , a fossil classification module 630 , and a digital materialization module 640 .

[0054] An image acquisition module 610 is used to acquire a CT image of the molded fossil; A model generation module 620, for generating a three-dimensional model of the molded fossil based on the CT image of the molded fossil; The fossil classification module 630 is used to extract local features of the three-dimensional model of the mold-cast fossil, combine the local features, and classify the mold-cast fossil according to the combined overall features; wherein the local features of the three-dimensional model of the mold-cast fossil include: tumor point structure features, radial ridge structure features and medullary cavity foramen structure features; The digital materialization module 640 is used to generate a digital materialized cast fossil based on the cast fossil three-dimensional model and the classification result of the cast fossil.

[0055] For the detailed description of the execution process and beneficial effects of the system module, please refer to the description of the above method embodiment, which will not be elaborated here.

[0056] In an exemplary embodiment, a computer device is also provided, the computer device comprising a memory and a processor, the memory storing a computer program, the computer program being loaded and executed by the processor to implement the above-mentioned method for digital materialization of a cast fossil.

[0057] In an exemplary embodiment, a computer-readable storage medium is further provided, on which a computer program is stored. When the computer program is executed by a processor, the method for digitally materializing a molded fossil as described above is implemented.

[0058] In an exemplary embodiment, a computer program product is also provided. When the computer program product is run on a computer device, the computer device is caused to execute the method for digitally materializing a molded fossil as described above.

[0059] The above description is only an embodiment of the present invention and is not intended to limit the present invention. Any modification, equivalent substitution, improvement, etc. made within the spirit and principle of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for digital materialization of a molded fossil, characterized in that: The method comprises: Obtain CT images of the molded fossils; Generating a three-dimensional model of the mold-cast fossil based on the CT image of the mold-cast fossil; Extracting local features of the three-dimensional model of the mold-cast fossil, combining the local features, and classifying the mold-cast fossil according to the combined overall features; wherein the local features of the three-dimensional model of the mold-cast fossil include: tumor point structure features, radial ridge structure features, and medullary cavity foramen structure features; Based on the three-dimensional model of the cast fossil and the classification results of the cast fossil, a digitally materialized cast fossil is generated.

2. The method according to claim 1, characterized in that Generating a CT image of the molded fossil, comprising: Cut the rock sample according to the set size to obtain the molded fossil; The X-ray detector is used to image the molded fossil at different angles in a 360° step-by-step manner; The images at each angle were reconstructed, and the filtered back-projection algorithm was used to obtain the CT images of the molded fossil.

3. The method according to claim 1, characterized in that Based on the CT image of the molded fossil, a three-dimensional model of the molded fossil is generated, including: De-noising the CT images of the molded fossils to obtain high signal-to-noise ratio CT images; Perform semantic segmentation on high signal-to-noise ratio CT images to obtain semantic segmentation results of molded fossils; The semantic segmentation results are used for isosurface extraction and surface volume rendering to generate a three-dimensional model of the cast fossil.

4. The method according to claim 3, characterized in that The denoising of the CT image of the mold cast fossil to obtain a high signal-to-noise ratio CT image includes: A filter is used to denoise the CT image of the molded fossil to obtain a CT image with a high signal-to-noise ratio; wherein the filter includes: a median filter or a mean filter; or, A CT image denoising network based on deep learning is used to denoise the CT image of the cast fossil to obtain a high signal-to-noise ratio CT image; wherein the CT image denoising network based on deep learning includes: a Noise2Noise network model or a Noise2Void network model.

5. The method according to claim 3, characterized in that: The semantic segmentation of the high signal-to-noise ratio CT image to obtain the semantic segmentation result of the molded fossil includes: Get the grayscale mean of the molded fossil , grayscale mean of surrounding rock and the overall grayscale value of high signal-to-noise ratio CT images ; By calculating the optimal threshold The between-class variance , get the optimal threshold ;in, is the probability of a molded fossil, is the probability of surrounding rock; Applying the best threshold Performing binary image segmentation on the high signal-to-noise ratio CT image to obtain a binary mold-cast fossil CT image; After removing the external small particles and internal small gaps in the binarized cast fossil CT image, the seed points of the cast fossil are selected to merge adjacent pixels with similar attributes and combined with manual circle selection to group pixels with similar properties into semantic regions, thus obtaining the semantic segmentation result of the cast fossil.

6. The method according to claim 3, characterized in that The semantic segmentation results are extracted by isosurface extraction and surface volume rendering to generate a three-dimensional model of the molded fossil, including: Obtain the three-dimensional volume data corresponding to each semantic segmentation result; Find the voxels that intersect the isosurface in the three-dimensional volume data and obtain the voxel edges; Applying linear interpolation to solve the intersection of the isosurface and the voxel edge; All intersection points are connected to extract isosurfaces to generate a three-dimensional model of the molded fossil.

7. The method according to claim 1, characterized in that The step of extracting the features of the three-dimensional model of the molded fossil and classifying the molded fossil based on the features comprises: Preprocessing the three-dimensional model of the molded fossil; The VGG16 network model of transfer learning is used to extract the overall features of the pre-processed three-dimensional model of the molded fossil; wherein the features of the pre-processed three-dimensional model of the molded fossil extracted by the VGG16 network model of transfer learning include: Using the front part of the network in the VGG16 network model, a low-level feature map of the three-dimensional model of the molded fossil is obtained, wherein the low-level feature map includes edge structure information of the three-dimensional model of the molded fossil; The low-level feature maps are combined using the middle part network in the VGG16 network model to obtain local features of the three-dimensional model of the molded fossil; The local features are combined by using the latter part of the network in the VGG16 network model to obtain the overall features of the three-dimensional model of the molded fossil; The long short-term memory architecture in the recurrent neural network model is used to fuse the overall features of multiple perspectives; The SVM multi-classifier is selected to classify the fused features to obtain the classification result of the mold cast fossil.

8. The method according to claim 7, characterized in that The three-dimensional model of the molded fossil is preprocessed, including: The PCA method is used to normalize the posture of the three-dimensional model of the molded fossil; and, Using the mean curvature map to enhance the local geometric features of the three-dimensional model of the molded fossil; and, A multi-view rendering tool is used to collect multi-view views of the cast fossil three-dimensional model.

9. The method according to claim 1, characterized in that: Based on the three-dimensional model of the cast fossil and the classification results of the cast fossil, a digitally materialized cast fossil is generated, which also includes: Repairing the three-dimensional model of the molded fossil, the repairing comprising: inverting reverse triangular facets, stitching gaps, filling holes, deleting bad edges, crossing and overlapping triangular facets, and interfering with shells; The repaired three-dimensional model of the cast fossil is imported into the 3D printing slicing software, and the surrounding rock is given a transparent material, and the cast fossil is given a color material based on the classification result of the cast fossil, so as to obtain a solid three-dimensional model of the cast fossil.

10. A digital materialization system for molded fossils, characterized in that: The system comprises: An image acquisition module, used for acquiring a CT image of the molded fossil; A model generation module, used for generating a three-dimensional model of the molded fossil based on the CT image of the molded fossil; A fossil classification module is used to extract local features of the three-dimensional model of the die-cast fossil, combine the local features, and classify the die-cast fossil according to the combined overall features; wherein the local features of the three-dimensional model of the die-cast fossil include: tumor point structure features, radial ridge structure features and medullary cavity foramen structure features; The digital materialization module is used to generate digital materialized cast fossils based on the three-dimensional model of the cast fossils and the classification results of the cast fossils.

Citation Information

Patent Citations

  • Ancient biological fossil positioning and recognition method based on deep learning

    CN110211173A

  • Micro paleontology fossil image detection, classification and discovery method, system and application

    CN113128335A

  • Lumbar fusion cage and preparation method thereof

    CN113952085A

  • Three-dimensional digital model stitching method

    CN115375882A

  • Micro fossil CT image analysis and classification method and device based on deep learning

    CN115546134A