Mars weathered rock image generation method and device, electronic equipment and storage medium

Abstract

The invention particularly relates to a Mars weathered rock image generation method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining a Mars weathered rock sample; generating a Mars rock data set based on the Mars weathered rock sample, and migrating knowledge of a preset CGAN model to the Mars rock data set to obtain a migrating optimized Mars rock data set; and generating a Mars weathered rock microscopic image based on the Mars rock data set after migration optimization, comparing the Mars weathered rock microscopic image with the Mars weathered rock sample, and evaluating the generation quality of the Mars weathered rock microscopic image according to a comparison result. Therefore, through earth weathered rock database transfer learning, construction of the conditional generative adversarial network and generation of the Mars weathered rock microscopic image, the problem that depicting of Mars rock characteristics in related technologies has great limitation is solved, and technical support is provided for Mars probe landing area site selection, drilling strategy making and in-situ resource utilization.

Description

Technical Field

[0001] The present application relates to the fields of digital rock cores, artificial intelligence (AI), and numerical simulation technology, and in particular to a method, device, electronic device, and storage medium for generating images of weathered rocks on Mars. Background Art

[0002] The microstructure of Martian rocks directly influences their physical and mechanical properties, and is of great scientific value for understanding Martian surface weathering, geological evolution, and planetary resource assessment. However, due to the limitations of Mars exploration missions, Martian rock research in related technologies has primarily focused on mineral composition analysis, while systematic research on their microstructure, physical and mechanical properties, and weathering mechanisms remains relatively scarce. The lack of microscopic image data limits the quantitative analysis of the pore structure, fracture evolution, and mechanical properties of Martian rocks, which in turn affects data interpretation, landing site selection, and resource development strategies for Mars exploration missions. Traditional laboratory testing methods are difficult to directly apply due to the extreme scarcity of Martian rock samples and limited experimental conditions. While analogous studies based on weathered rocks on Earth provide a reference, they are significantly limited in characterizing Martian rock properties due to the significant differences in environmental conditions between Earth and Mars (such as gravity, temperature, and atmospheric composition). Therefore, developing microscopic image generation methods suitable for Martian rocks is crucial for Martian rock mechanics research.

[0003] Among related technologies, generative AI technologies such as generative adversarial networks, variational autoencoders, and diffusion models have made significant progress in Earth rock image generation, pore structure modeling, and rock physical parameter prediction.

[0004] However, research in related technologies is mainly based on training based on Earth rock samples, and the generalization ability of the model depends on a large amount of data. However, Martian rock samples are extremely limited, making it difficult for existing AI models to accurately reflect the characteristics of Martian weathered rocks in the absence of sufficient data, which urgently needs to be solved. Summary of the Invention

[0005] This application provides a method, device, electronic device and storage medium for generating images of weathered rocks on Mars, in order to solve the problems such as the large limitations in the relevant technologies for depicting the characteristics of Martian rocks, and to provide technical support for the site selection of landing areas for Mars rovers, the formulation of drilling strategies and the utilization of in-situ resources.

[0006] A first embodiment of the present application provides a method for generating an image of weathered rocks on Mars, comprising the following steps:

[0007] Obtain samples of weathered Martian rocks;

[0008] generating a Martian rock dataset based on the Martian weathered rock sample, and migrating knowledge of a preset CGAN model to the Martian rock dataset to obtain a migrated and optimized Martian rock dataset;

[0009] A microscopic image of weathered rocks on Mars is generated based on the migrated and optimized Martian rock dataset, and the microscopic image of weathered rocks on Mars is compared with the Martian weathered rock sample, and the generation quality of the microscopic image of weathered rocks on Mars is evaluated according to the comparison result.

[0010] Optionally, before obtaining the Martian weathered rock sample, the method further includes:

[0011] Obtain a variety of Earth rock samples;

[0012] performing microscale testing and macroscale testing on the multiple earth rock samples to obtain microscopic parameters and macroscopic parameters of the multiple earth rock samples, respectively, and performing standardization processing on the microscopic parameters and the macroscopic parameters to obtain processed microscopic parameters and macroscopic parameters;

[0013] The multi-scale database of earth weathered rocks is established based on the processed microscopic parameters and macroscopic physical and mechanical parameters.

[0014] Optionally, before obtaining the Martian weathered rock sample, the method further includes:

[0015] generating a plurality of rock feature labels of the plurality of earth rock samples based on the plurality of earth rock samples; wherein the rock feature labels include mineral composition labels, weathering degree labels, and microstructure feature labels;

[0016] A preset CGAN model is constructed based on the conditional generative adversarial network and the rock feature labels.

[0017] Optionally, the preset CGAN model includes a CGAN model generator and a CGAN model discriminator, wherein:

[0018] The CGAN model generator is obtained based on a random vector and the weathering degree label, and the CGAN model generator is used to generate a rock mineral crystal model based on the random vector, the rock feature label and a preset loss function;

[0019] The discriminator of the preset CGAN model is obtained based on the rock mineral crystal model and the weathering degree label, and the discriminator of the preset CGAN model is used to discriminate the image generated by the rock mineral crystal model.

[0020] Optionally, the preset loss function is:

[0021] L gen =Ladv +λ1L mineral-comp +λ2L mineral-shape +λ3L weathering-prod +λ4L porosity +λ5L SSIM ;

[0022] Among them, L adv is the adversarial loss of CGAN, λ1 is the weight hyperparameter of mineral composition loss, L mineral-comp is the mineral composition loss, λ2 is the weight hyperparameter of the mineral morphology loss, L mineral-shap is the mineral form loss, λ3 is the weight hyperparameter of weathering degree loss, L weathering-prod is the weathering loss, λ4 is the weight hyperparameter of porosity loss, L porosity is the porosity ratio loss, λ5 is the weight hyperparameter of the structural similarity loss, L SSIM is the structural similarity loss.

[0023] Optionally, the comparison results include structural similarity, perceptual loss, peak signal-to-noise ratio, and Frechet Inception Distance, and evaluating the generation quality of the Martian weathered rock microscopic image based on the comparison results includes:

[0024] Based on a preset quantitative evaluation method, respectively comparing the structural similarity, the perceptual loss, the peak signal-to-noise ratio, and the Frechet Inception Distance with corresponding thresholds to obtain multiple comparison results;

[0025] Based on the multiple comparison results, verify whether the microscopic image of the weathered rock on Mars meets the preset validity conditions, and evaluate the generation quality of the microscopic image of the weathered rock on Mars based on the verification results.

[0026] A second embodiment of the present application provides a device for generating an image of weathered rocks on Mars, comprising:

[0027] Acquisition module, used to obtain Martian weathered rock samples;

[0028] a migration module, configured to generate a Martian rock dataset based on the Martian weathered rock sample, and migrate knowledge of a preset CGAN model to the Martian rock dataset to obtain a migrated and optimized Martian rock dataset;

[0029] A generation module is used to generate a microscopic image of Martian weathered rock based on the migrated and optimized Martian rock dataset, compare the microscopic image of Martian weathered rock with the Martian weathered rock sample, and evaluate the generation quality of the microscopic image of Martian weathered rock based on the comparison result.

[0030] Optionally, before obtaining the Martian weathered rock sample, the acquisition module is further configured to:

[0031] Obtain a variety of Earth rock samples;

[0032] performing microscale testing and macroscale testing on the multiple earth rock samples to obtain microscopic parameters and macroscopic parameters of the multiple earth rock samples, respectively, and performing standardization processing on the microscopic parameters and the macroscopic parameters to obtain processed microscopic parameters and macroscopic parameters;

[0033] The multi-scale database of earth weathered rocks is established based on the processed microscopic parameters and macroscopic physical and mechanical parameters.

[0034] Optionally, before obtaining the Martian weathered rock sample, the acquisition module is further configured to:

[0035] generating a plurality of rock feature labels of the plurality of earth rock samples based on the plurality of earth rock samples; wherein the rock feature labels include mineral composition labels, weathering degree labels, and microstructure feature labels;

[0036] A preset CGAN model is constructed based on the conditional generative adversarial network and the rock feature labels.

[0037] Optionally, the preset CGAN model includes a CGAN model generator and a CGAN model discriminator, wherein:

[0038] The CGAN model generator is obtained based on a random vector and the weathering degree label, and the CGAN model generator is used to generate a rock mineral crystal model based on the random vector, the rock feature label and a preset loss function;

[0039] The discriminator of the preset CGAN model is obtained based on the rock mineral crystal model and the weathering degree label, and the discriminator of the preset CGAN model is used to discriminate the image generated by the rock mineral crystal model.

[0040] Optionally, the preset loss function is:

[0041] L gen =L adv +λ1L mineral-comp +λ2L mineral-shape +λ3L weathering-prod +λ4L porosity +λ5L SSIM ;

[0042] Among them, L adv is the adversarial loss of CGAN, λ1 is the weight hyperparameter of mineral composition loss, L mineral-compis the mineral composition loss, λ2 is the weight hyperparameter of the mineral morphology loss, L mineral-shap is the mineral form loss, λ3 is the weight hyperparameter of weathering degree loss, L weathering-prod is the weathering loss, λ4 is the weight hyperparameter of porosity loss, L porosity is the porosity ratio loss, λ5 is the weight hyperparameter of the structural similarity loss, L SSIM is the structural similarity loss.

[0043] Optionally, the comparison result includes structural similarity, perceptual loss, peak signal-to-noise ratio, and Frechet Inception Distance, and the generation module is specifically configured to:

[0044] Based on a preset quantitative evaluation method, respectively comparing the structural similarity, the perceptual loss, the peak signal-to-noise ratio, and the Frechet Inception Distance with corresponding thresholds to obtain multiple comparison results;

[0045] Based on the multiple comparison results, verify whether the microscopic image of the weathered rock on Mars meets the preset validity conditions, and evaluate the generation quality of the microscopic image of the weathered rock on Mars based on the verification results.

[0046] The third aspect of the present application provides an electronic device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are configured to execute the method for generating an image of weathered rocks on Mars as described in the above embodiment.

[0047] The fourth embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon, which is executed by a processor to implement the method for generating an image of weathered rocks on Mars as described in the above embodiment.

[0048] Thus, the embodiment of the present application obtains weathered rock samples from Mars, generates a Martian rock dataset, transfers knowledge from a preset CGAN model to the Martian rock dataset to obtain a migrated and optimized Martian rock dataset, generates microscopic images of the Martian weathered rock, compares the microscopic images of the Martian weathered rock with the Martian weathered rock samples, and evaluates the quality of the generated microscopic images of the Martian weathered rock based on the comparison results. Thus, through transfer learning from the Earth's weathered rock database, a conditional generative adversarial network is constructed to generate microscopic images of Martian weathered rock, addressing the significant limitations of related technologies in characterizing Martian rock properties and providing technical support for Mars rover landing site selection, drilling strategy formulation, and in-situ resource utilization.

[0049] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0051] Figure 1 This is a flow chart of a method for generating an image of weathered rocks on Mars according to an embodiment of the present application;

[0052] Figure 2 A schematic diagram of a small sample of Martian meteorite data for a method for generating an image of weathered rocks on Mars according to one embodiment of the present application;

[0053] Figure 3 A schematic diagram of a CGAN model construction framework for a method for generating images of weathered rocks on Mars according to one embodiment of the present application;

[0054] Figure 4 A schematic diagram of a knowledge transfer strategy for a method for generating images of weathered rocks on Mars according to one embodiment of the present application;

[0055] Figure 5 This is a flow chart of a method for generating an image of weathered rocks on Mars according to one embodiment of the present application;

[0056] Figure 6 A schematic diagram of a device for generating an image of weathered rocks on Mars according to an embodiment of the present application;

[0057] Figure 7 A schematic diagram of the structure of an electronic device provided according to an embodiment of the present application. DETAILED DESCRIPTION

[0058] The following describes in detail embodiments of the present application, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0059] The following describes the Martian weathered rock image generation method, device, electronic device, and storage medium according to an embodiment of the present application with reference to the accompanying drawings. In response to the significant limitations in the depiction of Martian rock characteristics in the related art mentioned in the background art, the present application provides a Martian weathered rock image generation method. In this method, the present application embodiment obtains Martian weathered rock samples, generates a Martian rock dataset, transfers the knowledge of a preset CGAN model to the Martian rock dataset to obtain a migrated and optimized Martian rock dataset, generates a Martian weathered rock micro-image, compares the Martian weathered rock micro-image with the Martian weathered rock sample, and evaluates the generated Martian weathered rock micro-image based on the comparison results. Thus, through transfer learning from the Earth's weathered rock database, a conditional generative adversarial network is constructed to generate Martian weathered rock micro-images, solving the significant limitations in depicting Martian rock characteristics in the related art and providing technical support for the Mars rover landing area selection, drilling strategy formulation, and in-situ resource utilization.

[0060] Specifically, Figure 1 A schematic flow chart of a method for generating an image of weathered rocks on Mars provided in an embodiment of the present application.

[0061] like Figure 1 As shown, the method for generating an image of weathered rocks on Mars includes the following steps:

[0062] In step S101, a Martian weathered rock sample is obtained.

[0063] Specifically, if Figure 2 As shown, Figure 2 This is a schematic diagram of a small sample data of a Martian meteorite for a method for generating an image of weathered rocks on Mars according to an embodiment of the present application; the main diagenetic mineral composition of the Martian meteorite is determined based on the test of the fully automatic mineral analysis system (Tescan Integrated Mineral Analyzer, TIMA); the load-displacement curves of the main diagenetic minerals are obtained based on nanoindentation, and their elastic modulus and hardness are calculated accordingly; the surface roughness and elastic modulus of the intercrystalline phases of different mineral crystals are obtained based on atomic force microscopy; and the three-dimensional mineral and micropore distribution of the Martian meteorite is obtained based on computed tomography (CT) scanning technology.

[0064] Optionally, in some embodiments, before obtaining Martian weathered rock samples, the method further includes: obtaining a plurality of Earth rock samples; performing microscale testing and macroscale testing on the plurality of Earth rock samples to obtain micro parameters and macro parameters of the plurality of Earth rock samples respectively, and standardizing the micro parameters and macro parameters to obtain processed micro parameters and macro parameters; and establishing a multi-scale database of Earth weathered rocks based on the processed micro parameters and macroscopic physical and mechanical parameters.

[0065] It is understandable that earth rock samples with different weathering degrees and rock types are obtained for micro-scale and macro-scale testing; at the microscale, micro-parameters such as the mineral composition, microstructure, diagenetic minerals and intercrystalline phase mechanical properties of the earth rock samples are obtained through micro-rock experiments; at the macroscale, macro-physical and mechanical parameters such as elastic modulus, hardness, compressive strength, tensile strength, fracture toughness and wave velocity of earth rock samples with different weathering degrees are obtained through macro-experiments such as uniaxial compression, Brazilian splitting and wave velocity testing; combined with data standardization processing, a multi-scale database of weathered earth rocks is established; image enhancement and feature extraction methods (such as edge detection, Fourier transform, morphological analysis, etc.) are used to standardize the data, improve the quality of training data, reduce the impact of noise, and provide data support for subsequent weathering geological knowledge extraction, feature label construction and AI model development.

[0066] Optionally, in some embodiments, before obtaining Martian weathered rock samples, the method further includes: generating multiple rock feature labels for multiple Earth rock samples based on multiple Earth rock samples; wherein the rock feature labels include mineral composition labels, weathering degree labels, and microstructure feature labels; and constructing a preset CGAN model based on a conditional generative adversarial network and the rock feature labels.

[0067] It is understandable that based on the study of multi-scale weathering, weathering has a significant impact on the mineral composition, distribution and micropore characteristics of rocks; extracting rock characteristics includes mineral composition, that is, extracting the main mineral components in rock samples, combining nanoindentation testing to obtain mineral crystal mechanical parameters, and classifying minerals into weathering-resistant and weathering-prone categories; mineral component labels quantitatively analyze the mineral composition of rock samples (including weathering-resistant minerals, weathering-prone minerals and weathering products) through X-ray diffraction, scanning electron microscopy and other technologies to obtain their volume fraction and mass fraction; extracting rock characteristics includes weathering products. The weathering process involves identifying weathering products (such as kaolin and hematite) formed during the weathering process and analyzing their formation and distribution patterns at different weathering stages. Weathering degree labels are based on the International Society for Rock Mechanics weathering scale, marking the weathering degree of each rock sample to ensure a correspondence between the selected weathering characteristics and the rock's physical and mechanical properties. Rock characteristics extracted include microstructure, recording the evolution of microstructures such as mineral distribution, mineral morphology, and micropores during weathering. Microstructural feature labels are based on CT scanning technology to extract microstructural features such as porosity and fracture density. Furthermore, the orientation distribution of mineral crystals is analyzed using microscopic techniques such as electron backscatter diffraction, serving as a "directionality" label to describe the arrangement and orientation of minerals in the rock. At the same time, a structural similarity index is introduced to measure the similarity of local details of rock mineral crystals; based on the extraction of weathering geological knowledge, rock feature labels with real physical properties are constructed; and based on these rock feature labels, the network parameters are initialized, real samples and labels are input, fake samples G(z|y) are generated, the discriminator parameters are updated, the generator parameters are updated, and a preset CGAN model is constructed based on the conditional generative adversarial network.

[0068] Optionally, in some embodiments, the preset CGAN model includes a CGAN model generator and a CGAN model discriminator, wherein the CGAN model generator is obtained based on a random vector and a weathering degree label, and the CGAN model generator is used to generate a rock mineral crystal model based on the random vector, the rock feature label and a preset loss function; the discriminator of the preset CGAN model is obtained based on the rock mineral crystal model and the weathering degree label, and the discriminator of the preset CGAN model is used to discriminate images generated by the rock mineral crystal model.

[0069] Optionally, in some embodiments, the preset loss function is:

[0070] L gen =L adv +λ1L mineral-comp +λ2L mineral-shape +λ3L weathering-prod +λ4L porosity +λ5L SSIM ;

[0071] Among them, L adv is the adversarial loss of CGAN, λ1 is the weight hyperparameter of mineral composition loss, L mineral-comp is the mineral composition loss, λ2 is the weight hyperparameter of the mineral morphology loss, L mineral-shap is the mineral form loss, λ3 is the weight hyperparameter of weathering degree loss, L weathering-prod is the weathering loss, λ4 is the weight hyperparameter of porosity loss, L porosity is the porosity ratio loss, λ5 is the weight hyperparameter of the structural similarity loss, L SSIM is the structural similarity loss.

[0072] It is understandable that if Figure 3 As shown, Figure 3 This is a schematic diagram of the CGAN model construction framework for a method for generating images of weathered rocks on Mars according to an embodiment of the present application. The embodiment of the present application adopts a conditional generative adversarial neural network (CGAN) structure. The input of the generator G is random noise and a weathering degree label. The conditional information is introduced into the generation network through a splicing vector injection mechanism. The preset loss function is integrated to generate the corresponding mineral crystal image under the weathering stage. The generator input includes the weathering degree label (wj): x'i = G(z,wj). Among them, G is the generator, and x'i is the generated weathered rock mineral crystal model. The generator generates a rock mineral crystal model with physical properties and structure consistent with the input feature label based on these conditional labels and random noise z. The discriminator D receives real samples or generated images and their corresponding labels, and uses the PatchGAN architecture to discriminate local areas of the image to improve the fidelity of details; the mineral composition labels and microstructural feature labels participate in the loss function calculation, which specifically includes adversarial loss, mineral composition loss, orientation loss, weathering product and pore ratio loss, and structural similarity index (SSIM) loss to generate a mineral crystal model with consistent physical properties; the discriminator input is the generated rock sample x' i and the real sample x i Composed of, and supplemented by the conditional tag w j , can be expressed as: D(x i ,w j ).

[0073] Furthermore, L adv is the adversarial loss of CGAN, and the calculation formula is:

[0074]

[0075] Among them, L adv is the adversarial loss of CGAN, For the real data distribution p data The expectation of logD(x|y) is the predicted logarithmic probability of the discriminator D for the input sample x being a true sample under the given condition y. For the noise distribution p z , D(G(z|y)) is the probability output of the discriminator D for the sample G(z|y) generated by the generator G as a real sample under the given condition y.

[0076] L mineral-comp The loss of mineral composition ensures that the mineral composition of the generated rock conforms to the law of weathering. The calculation formula is:

[0077]

[0078] in, To generate the predicted volume fraction of the i-th type of mineral in the image, M i is the volume fraction of type i minerals in real geological samples.

[0079] L mineral-shap To account for mineral morphology loss, Angular Variance and Cosine Similarity are used to measure the difference in orientation between the generated mineral crystals and the real samples. The calculation formula is:

[0080]

[0081] in, To generate the orientation distribution of mineral crystals, O i is the orientation distribution of real mineral crystals, λ θ and λ cos is a hyperparameter, V θ To control the variance of orientation, S cos To control the similarity of orientation, it is used to adjust the weight.

[0082] In particular, in order to ensure that the rock mineral structure in the weathered rock AI model generation process is consistent with the actual situation in terms of overall distribution and local details, L is introduced into the loss function. SSIM , which is used to control the local details of the generated rock mineral crystal graphics (such as the boundaries of mineral crystals); the preset loss function of the generator and discriminator consists of two parts: adversarial loss and physical property loss.

[0083] In step S102, a Martian rock dataset is generated based on Martian weathered rock samples, and the knowledge of the preset CGAN model is migrated to the Martian rock dataset to obtain a migrated and optimized Martian rock dataset.

[0084] Specifically, based on Martian weathered rock samples combined with data standardization processing, the Martian meteorite small data constraints include the introduction of Martian meteorite micro- and macro-data, including mineral composition, microstructure, mineral crystals and intercrystalline phase mechanical properties. Figure 4 As shown, Figure 4 This is a schematic diagram of a knowledge transfer strategy for a method for generating images of weathered Martian rocks, according to one embodiment of the present application. The method applies weathering knowledge learned by a source domain generative adversarial neural network (GAN) to a target domain (Martian weathered rock samples) through transfer learning, thereby generating Martian weathered rock images with realistic physical properties, given minimal Martian rock image data. Based on a knowledge transfer mechanism, the weathering characteristics of the source domain (Earth rock samples) are effectively transferred to the target domain (Martian weathered rock samples) through conditional batch normalization (BN) layer parameter migration to address the problem of scarce Martian data. Based on a knowledge sharing strategy, statistical features are shared between small sample categories of Martian rocks to improve the model's generalization ability under small data volumes. Regularization constraints on similarity and residuals are introduced into the GAN training objective to prevent overfitting and ensure the stability and diversity of the generated results.

[0085] In step S103, a Martian weathered rock microscopic image is generated based on the migrated and optimized Martian rock dataset, and the Martian weathered rock microscopic image is compared with a Martian weathered rock sample, and the generation quality of the Martian weathered rock microscopic image is evaluated based on the comparison results.

[0086] Specifically, the CGAN model optimized by knowledge transfer was used to generate microscopic images of weathered rocks on Mars, and compared with actual Martian meteorite samples to verify the physical validity and detail fidelity of the generated images.

[0087] Optionally, in some embodiments, the comparison results include structural similarity, perception loss, peak signal-to-noise ratio and Frechet Inception Distance, and the generation quality of the microscopic image of the weathered rock on Mars is evaluated according to the comparison results, including: based on a preset quantitative evaluation method, comparing the structural similarity, perception loss, peak signal-to-noise ratio and Frechet Inception Distance with corresponding thresholds to obtain multiple comparison results; based on the multiple comparison results, verifying whether the microscopic image of the weathered rock on Mars meets the preset validity conditions, and evaluating the generation quality of the microscopic image of the weathered rock on Mars based on the verification results.

[0088] It can be understood that the following quantitative indicators are used to evaluate the model performance by comparing the generated images with the real images; when the mean value of the Structural Similarity Index Measure (SSIM) reaches 0.87, it indicates that the image structure is consistent; when the mean value of the Learned Perceptual Image Patch Similarity (LPIPS) is lower than 0.12, it indicates that the model has good detail restoration capabilities; when the mean value of the Peak Signal-to-Noise Ratio (PSNR) is higher than 30dB, it indicates that the generated image has higher visual quality than the original image, with less noise and better image clarity; when the mean value of the Frechet Inception Distance (FID) is lower than 25, it indicates that the distribution difference between the generated image and the real image is small, and the generated image is close to the real image in visual quality.

[0089] Therefore, in response to the problems of extreme scarcity of Martian rock samples, limitations of traditional experimental methods, and insufficient generalization capabilities of existing AI models, the embodiments of the present application utilize a multi-scale database of Earth's weathered rocks and construct a generative adversarial neural network model through a transfer learning method, so that it can effectively learn the microstructural characteristics of Martian rocks under small sample data of Martian meteorites, thereby generating microscopic images of Martian weathered rocks with real physical meaning; the embodiments of the present application can be used for numerical simulation of Martian weathered rocks, prediction of physical and mechanical properties, and data support for deep space exploration missions, providing technical support for the site selection of Mars rover landing areas, formulation of drilling strategies, and in-situ resource utilization.

[0090] To facilitate those skilled in the art to further understand the method for generating a weathered rock image on Mars according to the embodiment of the present application, the following is a description of the method. Figure 5 The illustrated embodiment will be described in detail.

[0091] Specifically, if Figure 5 As shown, Figure 5A flowchart of a method for generating images of weathered rocks on Mars provided in one embodiment of the present application. In response to the extreme scarcity of Martian samples and the difficulty in generalizing traditional experimental methods and AI models, the embodiment of the present application uses the Earth's weathered rock database to construct a conditional generative adversarial network (CGAN) to extract weathering labels such as mineral composition, weathering products, and microstructure. Through transfer learning, the Earth's weathering knowledge is transferred to the target domain of Martian rocks to achieve physically credible generation of microscopic images under the conditions of a small amount of Martian meteorite data; the introduction of multi-label physical constraints and regularization mechanisms significantly improves the structural fidelity and weathering evolution consistency of the model in a small sample environment; the embodiment of the present application is suitable for scenarios such as prediction of Martian weathered rock structure, landing area analysis, and image completion and auxiliary interpretation of deep space exploration missions. It has the advantage of efficient modeling under data scarcity conditions, and specifically includes the following steps:

[0092] S501: Data collection and preprocessing Collect earth rock samples for microscopic and macroscopic testing to establish a standardized multi-scale database.

[0093] S502: Extract the mineral composition, weathering products, and microstructure of earth rock samples to construct physical property labels.

[0094] S503: Construct a CGAN model and use the feature labels as conditional input to design the generator and discriminator to define a multi-constraint loss function.

[0095] S504: Based on the Martian meteorites NWA 13190 and NWA 12564, the research was carried out to obtain the mineral, mechanical and pore data of Martian meteorite 13190 from Martian meteorite samples through TIMA, nanoindentation, CT scanning and other technologies.

[0096] S505: Conditional batch normalization and feature sharing are used to introduce regularization on the parameters of the Earth rock CGAN model to prevent overfitting and perform knowledge transfer. The weathering knowledge learned by the source domain CGAN model is applied to the target domain (Martian meteorite 13190) through transfer learning.

[0097] S506: Based on the Martian meteorite 13190 image data, generate an image of the Martian weathered meteorite 13190 with real physical properties, and quantitatively verify it through indicators such as SSIM and PSNR to ensure physical validity and detail fidelity.

[0098] Therefore, the embodiment of the present application breaks through the data bottleneck and, through transfer learning and small data constraints, enables the GAN model to still generate high-quality microscopic images under the conditions of limited Martian rock samples; combines physical constraints and weathering simulation to improve the rationality of the mineral structure of the generated image, so that it has higher physical credibility and enhances the image authenticity; by combining generative AI algorithms with rock physics modeling, a multi-scale Martian weathered rock database is constructed to provide data support for numerical simulation and exploration missions, and support multi-scale analysis; the embodiment of the present application can be used for Mars landing area site selection, drilling analysis and resource assessment, improve the scientific efficiency and engineering feasibility of exploration missions, and enhance Mars exploration applications.

[0099] According to the method for generating images of weathered Martian rocks proposed in the embodiments of this application, a sample of weathered Martian rock is obtained to generate a Martian rock dataset. The knowledge of a preset CGAN model is transferred to the Martian rock dataset to obtain a migrated and optimized Martian rock dataset. A microscopic image of the Martian rock is generated, and the microscopic image is compared with the sample. The quality of the generated microscopic image is evaluated based on the comparison results. Thus, by using transfer learning from an Earth weathered rock database and constructing a conditional generative adversarial network to generate microscopic images of Martian rock, this method addresses the significant limitations of related technologies in characterizing Martian rock properties and provides technical support for Mars rover landing site selection, drilling strategy formulation, and in-situ resource utilization.

[0100] Next, the device for generating an image of weathered rocks on Mars proposed in accordance with an embodiment of the present application will be described with reference to the accompanying drawings.

[0101] Figure 6 It is a block diagram of a device for generating an image of weathered rocks on Mars according to an embodiment of the present application.

[0102] like Figure 6 As shown, the Mars weathered rock image generation device 10 includes: an acquisition module 100, a migration module 200 and a generation module 300.

[0103] The acquisition module 100 is used to obtain Martian weathered rock samples;

[0104] A migration module 200 is configured to generate a Martian rock dataset based on Martian weathered rock samples, and to migrate the knowledge of a preset CGAN model to the Martian rock dataset to obtain a migrated and optimized Martian rock dataset;

[0105] The generation module 300 is used to generate a microscopic image of Martian weathered rock based on the migrated and optimized Martian rock dataset, compare the microscopic image of Martian weathered rock with a Martian weathered rock sample, and evaluate the generation quality of the microscopic image of Martian weathered rock based on the comparison results.

[0106] Optionally, in some embodiments, before obtaining Martian weathered rock samples, the acquisition module 100 is also used to: obtain a variety of Earth rock samples; perform microscale testing and macroscale testing on the variety of Earth rock samples to obtain micro parameters and macro parameters of the variety of Earth rock samples, respectively, and perform standardization on the micro parameters and macro parameters to obtain processed micro parameters and macro parameters; and establish a multi-scale database of Earth weathered rocks based on the processed micro parameters and macroscopic physical and mechanical parameters.

[0107] Optionally, in some embodiments, before obtaining Martian weathered rock samples, the acquisition module 100 is further used to: generate multiple rock feature labels for multiple Earth rock samples based on multiple Earth rock samples; wherein the rock feature labels include mineral composition labels, weathering degree labels, and microstructure feature labels; and construct a preset CGAN model based on the conditional generative adversarial network and the rock feature labels.

[0108] Optionally, in some embodiments, the preset CGAN model includes a CGAN model generator and a CGAN model discriminator, wherein the CGAN model generator is obtained based on a random vector and a weathering degree label, and the CGAN model generator is used to generate a rock mineral crystal model based on the random vector, the rock feature label and a preset loss function; the discriminator of the preset CGAN model is obtained based on the rock mineral crystal model and the weathering degree label, and the discriminator of the preset CGAN model is used to discriminate images generated by the rock mineral crystal model.

[0109] Optionally, in some embodiments, the preset loss function is:

[0110] L gen =L adv +λ1L mineral-comp +λ2L mineral-shape +λ3L weathering-prod +λ4L porosity +λ5L SSIM ;

[0111] Among them, L adv is the adversarial loss of CGAN, λ1 is the weight hyperparameter of mineral composition loss, L mineral-comp is the mineral composition loss, λ2 is the weight hyperparameter of the mineral morphology loss, L mineral-shap is the mineral form loss, λ3 is the weight hyperparameter of weathering degree loss, L weathering-prod is the weathering loss, λ4 is the weight hyperparameter of porosity loss, L porosity is the porosity ratio loss, λ5 is the weight hyperparameter of the structural similarity loss, L SSIM is the structural similarity loss.

[0112] Optionally, in some embodiments, the comparison results include structural similarity, perception loss, peak signal-to-noise ratio and Frechet Inception Distance, and the generation module 300 is specifically used to: based on a preset quantitative evaluation method, respectively compare the structural similarity, perception loss, peak signal-to-noise ratio and Frechet Inception Distance with corresponding thresholds to obtain multiple comparison results; based on the multiple comparison results, verify whether the microscopic image of Martian weathered rock meets the preset validity conditions, and evaluate the generation quality of the microscopic image of Martian weathered rock according to the verification results.

[0113] It should be noted that the above explanation of the embodiment of the method for generating an image of weathered rock on Mars is also applicable to the device for generating an image of weathered rock on Mars in this embodiment, and will not be repeated here.

[0114] According to the Martian weathered rock image generation device proposed in the embodiment of the present application, the embodiment of the present application obtains Martian weathered rock samples, generates a Martian rock dataset, transfers the knowledge of the preset CGAN model to the Martian rock dataset to obtain a migrated and optimized Martian rock dataset, generates a Martian weathered rock micro-image, compares the Martian weathered rock micro-image with the Martian weathered rock sample, and evaluates the generated Martian weathered rock micro-image quality based on the comparison results. Thus, through transfer learning from the Earth's weathered rock database, a conditional generative adversarial network is constructed to generate Martian weathered rock micro-images, solving the significant limitations of related technologies in characterizing Martian rock characteristics, and providing technical support for Mars rover landing site selection, drilling strategy formulation, and in-situ resource utilization.

[0115] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. The electronic device may include:

[0116] Memory 701 , processor 702 , and computer programs stored in the memory 701 and executable on the processor 702 .

[0117] When the processor 702 executes the program, the method for generating an image of weathered rocks on Mars provided in the above embodiment is implemented.

[0118] Furthermore, the electronic device further includes:

[0119] The communication interface 703 is used for communication between the memory 701 and the processor 702 .

[0120] The memory 701 is used to store computer programs that can be run on the processor 702 .

[0121] The memory 701 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0122] If the memory 701, processor 702, and communication interface 703 are implemented independently, the communication interface 703, memory 701, and processor 702 can be interconnected via a bus and communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus. Buses can be divided into address buses, data buses, control buses, etc. For ease of representation, Figure 7 Only one thick line is used in the diagram, but this does not mean that there is only one bus or one type of bus.

[0123] Optionally, in a specific implementation, if the memory 701, the processor 702 and the communication interface 703 are integrated on a chip, the memory 701, the processor 702 and the communication interface 703 can communicate with each other through an internal interface.

[0124] The processor 702 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application.

[0125] An embodiment of the present application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-mentioned method for generating an image of weathered rocks on Mars.

[0126] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or N embodiments or examples in a suitable manner. In addition, those skilled in the art can combine and combine different embodiments or examples described in this specification and features of different embodiments or examples without contradiction.

[0127] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of technical features indicated. Thus, a feature specified as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of this application, "N" means at least two, for example, two, three, etc., unless otherwise specifically defined.

[0128] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code comprising one or more executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may be performed out of the order shown or discussed, including performing functions in a substantially simultaneous manner or in reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present application belong.

[0129] It should be understood that various parts of the present application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiment, the N steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used to implement: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0130] Those skilled in the art will understand that all or part of the steps in the method of the above embodiment can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiment.

Claims

1. A method for generating an image of weathered rocks on Mars, characterized in that: The method comprises the following steps: Obtain samples of weathered Martian rocks; generating a Martian rock dataset based on the Martian weathered rock sample, and migrating knowledge of a preset CGAN model to the Martian rock dataset to obtain a migrated and optimized Martian rock dataset; A microscopic image of weathered rocks on Mars is generated based on the migrated and optimized Martian rock dataset, and the microscopic image of weathered rocks on Mars is compared with the weathered rock sample on Mars, and the generation quality of the microscopic image of weathered rocks on Mars is evaluated according to the comparison result.

2. The method according to claim 1, characterized in that Before obtaining the Martian weathered rock sample, the method further includes: Obtain a variety of Earth rock samples; performing microscale testing and macroscale testing on the multiple earth rock samples to obtain microscopic parameters and macroscopic parameters of the multiple earth rock samples, respectively, and performing standardization processing on the microscopic parameters and the macroscopic parameters to obtain processed microscopic parameters and macroscopic parameters; The multi-scale database of earth weathered rocks is established based on the processed microscopic parameters and macroscopic physical and mechanical parameters.

3. The method according to claim 2, characterized in that Before obtaining the Martian weathered rock sample, the method further includes: generating a plurality of rock feature labels of the plurality of earth rock samples based on the plurality of earth rock samples; wherein the rock feature labels include mineral composition labels, weathering degree labels, and microstructure feature labels; A preset CGAN model is constructed based on the conditional generative adversarial network and the rock feature labels.

4. The method according to claim 3, characterized in that The preset CGAN model includes a CGAN model generator and a CGAN model discriminator, wherein, The CGAN model generator is obtained based on a random vector and the weathering degree label, and the CGAN model generator is used to generate a rock mineral crystal model based on the random vector, the rock feature label and a preset loss function; The discriminator of the preset CGAN model is obtained based on the rock mineral crystal model and the weathering degree label, and the discriminator of the preset CGAN model is used to discriminate the image generated by the rock mineral crystal model.

5. The method according to claim 4, characterized in that The preset loss function is: L gen =L adv +λ1L mineral-comp +λ2L mineral-shape +λ3L weathering-prod +λ4L porosity +λ5L SSIM ; Among them, L adv is the adversarial loss of CGAN, λ1 is the weight hyperparameter of mineral composition loss, L mineral-comp is the mineral composition loss, λ2 is the weight hyperparameter of the mineral morphology loss, L mineral-shap is the mineral form loss, λ3 is the weight hyperparameter of weathering degree loss, L weathering-prod is the weathering loss, λ4 is the weight hyperparameter of porosity loss, L porosity is the porosity ratio loss, λ5 is the weight hyperparameter of the structural similarity loss, L SSIM is the structural similarity loss.

6. The method according to claim 1, wherein The comparison results include structural similarity, perception loss, peak signal-to-noise ratio, and Frechet Inception Distance. The generation quality of the microscopic image of the weathered Martian rock is evaluated based on the comparison results, including: Based on a preset quantitative evaluation method, respectively comparing the structural similarity, the perceptual loss, the peak signal-to-noise ratio, and the Frechet Inception Distance with corresponding thresholds to obtain multiple comparison results; Based on the multiple comparison results, verify whether the microscopic image of the weathered rock on Mars meets the preset validity conditions, and evaluate the generation quality of the microscopic image of the weathered rock on Mars based on the verification results.

7. A device for generating images of weathered rocks on Mars, characterized in that: include: Acquisition module, used to obtain Martian weathered rock samples; a migration module, configured to generate a Martian rock dataset based on the Martian weathered rock sample, and migrate knowledge of a preset CGAN model to the Martian rock dataset to obtain a migrated and optimized Martian rock dataset; A generation module is used to generate a microscopic image of Martian weathered rock based on the migrated and optimized Martian rock dataset, compare the microscopic image of Martian weathered rock with the Martian weathered rock sample, and evaluate the generation quality of the microscopic image of Martian weathered rock based on the comparison result.

8. The device according to claim 7, characterized in that Before obtaining the Martian weathered rock sample, the acquisition module is further configured to: Obtain a variety of Earth rock samples; performing microscale testing and macroscale testing on the multiple earth rock samples to obtain microscopic parameters and macroscopic parameters of the multiple earth rock samples, respectively, and performing standardization processing on the microscopic parameters and the macroscopic parameters to obtain processed microscopic parameters and macroscopic parameters; The multi-scale database of earth weathered rocks is established based on the processed microscopic parameters and macroscopic physical and mechanical parameters.

9. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method for generating an image of weathered rocks on Mars as described in any one of claims 1 to 6.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: The program is executed by a processor to implement the method for generating an image of weathered rocks on Mars as described in any one of claims 1 to 6.

Images