A marine photovoltaic seabed dynamic geological modeling method based on environment-geology-engineering multi-source data fusion

Abstract

The application discloses a kind of offshore photovoltaic seabed dynamic geological modeling methods based on environment-geology-engineering multi-source data fusion, belong to offshore photovoltaic engineering geological survey site selection technical field.This method collects environment, geophysical prospecting, geotechnical investigation and photovoltaic array arrangement engineering data, carries out standardization pretreatment;Refer to multi-source data fusion modeling framework, utilize open source library (PyVista) to build environment-geology-engineering integrated three-dimensional model, realize stratum, geotechnical mechanics parameter and environment constraint space coupling;Realize survey data update automatic trigger interpolation recalculation algorithm, real-time correction model;Based on analytic hierarchy process-entropy weight method, construct multi-factor evaluation model, intelligently divide red, yellow, green area, combine geological information real-time early warning scour, settlement risk and optimize scheme.Solve the problem of traditional geological modeling data single, static constraint environment, realize fine intelligent site selection, significantly improve engineering safety and economy, with high application value.

Description

Technical Field

[0001] This invention relates to the fields of geological exploration for marine photovoltaic projects, marine environmental monitoring, marine engineering site selection, and seabed three-dimensional geological modeling. Specifically, it relates to a method for marine photovoltaic seabed geological modeling that integrates multi-source environmental, geological, and engineering data and achieves dynamic updates, risk zoning, and intelligent early warning. Background Technology

[0002] With the acceleration of the global energy transition, offshore photovoltaic (PV) power, as a new type of clean energy development, is becoming a new hotspot for new energy development in coastal areas due to its advantages such as not occupying land resources, high power generation efficiency, and strong absorption capacity. Unlike offshore wind power in the deep sea, offshore PV power plants are mainly located in near-shore shallow waters. These areas often have complex hydrodynamic conditions, dense human activities, and are significantly affected by marine meteorological and hydrological conditions such as tides, waves, and wind speed.

[0003] Among the existing publicly available technologies, patent application CN120355860A proposes a large-scale marine engineering geological modeling system, which only realizes the interpolation and three-dimensional visualization of geological data, without involving the conversion of geological parameters to engineering parameters; patent CN121168320A only targets multi-physics field coupling calculation, without solving the problem of integrated modeling of geological and engineering data. In the traditional field of marine engineering geological modeling, existing technologies are mostly concentrated on offshore wind power scenarios. For example, Wang Deyuan, Peng Xiuzhong, Du Yuchen, et al. A three-dimensional geological modeling method for offshore wind farms based on multi-source data fusion [J]. Acta Oceanologica Sinica, 2025, 47(2):85-97. proposed a three-dimensional modeling method that integrates geotechnical investigation and engineering geophysical data. However, this method only considers geological and engineering investigation data and does not include environmental constraints such as tidal currents, aquaculture areas, and ecological red lines. Offshore photovoltaic projects are located near the shore, and these environmental factors play a decisive role in the compliance, construction feasibility, and subsequent operation and maintenance safety of the project. For example, conflicts in aquaculture areas may prevent projects from being approved, and scouring in strong tidal current zones may cause pile foundation instability. These are all issues that traditional wind power geological modeling methods overlook.

[0004] Furthermore, most existing technologies rely on static modeling, meaning they generate models from a single input of survey data. This prevents dynamic updates based on supplementary survey data and real-time environmental monitoring data as the project progresses. Consequently, once established, the models cannot be iteratively optimized, making it difficult to address the complex and ever-changing geological and environmental conditions at sea, and hindering real-time updates on the distribution of sweet spots and danger zones. Simultaneously, existing technologies lack a suitable zoning mechanism for offshore photovoltaic projects, failing to provide owners with intuitive red, yellow, and green site selection suggestions, and lacking geological risk early warning and key survey area recommendation functions based on multi-source data.

[0005] Therefore, given the characteristics of offshore photovoltaic power generation being nearshore, subject to multiple constraints, and dynamically changing conditions, there is an urgent need to develop a new geological modeling method that can integrate multi-source data from the environment, geology, and engineering, and support dynamic updates, integrated modeling, risk zoning, and intelligent early warning, in order to fill the gaps in existing technologies. Summary of the Invention

[0006] To address the problems of static modeling, single data dimension, lack of dynamic update mechanism, inaccurate risk identification, and insufficient coupling of environmental constraints in existing technologies, this invention provides a dynamic geological modeling method for marine photovoltaic projects based on the fusion of multi-source environmental, geological, and engineering data. This method achieves integrated modeling and dynamic correction of environmental constraints, geological structure, and engineering parameters, intelligently divides construction suitability zones, provides real-time early warning of geological risks, and recommends key exploration areas, thus providing high-precision decision support for the site selection, design, and construction of marine photovoltaic projects.

[0007] Technical solution

[0008] A dynamic geological modeling method for marine photovoltaic seabed based on multi-source data fusion of environment, geology, and engineering includes the following steps:

[0009] S1 Multi-Source Data Acquisition and Standardized Preprocessing

[0010] Three types of data sources were collected and integrated: 1) Geological data: including engineering geophysical data such as shallow seabed profiles, side-scan sonar, single / multi-channel seismic data, and resistivity imaging; engineering survey data such as seabed borehole sampling, standard penetration test, static cone penetration test, and geotechnical tests; and geotechnical parameters such as lithology, soil thickness, soil density, internal friction angle, cohesion, and bearing capacity characteristics. 2) Engineering data: offshore photovoltaic array layout schemes, pile foundation types and dimensions, anchoring system parameters, draft and operational requirements of construction vessels, and foundation depth design parameters. 3) Environmental data: including ocean current velocity and direction, wave elements, seabed erosion rate, wind speed and wind rose diagrams, aquaculture area boundaries, shipping routes, ecological protection red lines, marine functional zoning, and distribution of submarine pipelines and obstacles.

[0011] The above data is processed by coordinate unification, format conversion, outlier removal, spatiotemporal matching and interpolation gridding to construct a standardized multi-source database.

[0012] S2 Integrated Environmental-Geological-Engineering 3D Geological Modeling

[0013] Based on the seabed topography, an initial three-dimensional stratigraphic structure model is constructed using methods such as Kriging interpolation, sequential Gaussian simulation, and tetrahedral mesh generation. Geomechanical parameters are assigned to corresponding stratigraphic units to form a geomechanical parameter field model. Environmental constraint surfaces such as tidal current field, wind speed field, scour sensitivity, aquaculture area, and ecological red line are superimposed to construct an environmental constraint field model. Through spatial superposition, attribute mapping, and topological association, an integrated environmental-geological-engineering three-dimensional geological model is achieved, uniformly expressing stratigraphic distribution, geomechanical properties, hydrodynamic conditions, ecological constraints, and engineering constraints. Specifically, this invention utilizes the Python open-source library PyVista for the construction and generation of the three-dimensional mesh. The Delaunay triangulation algorithm transforms discrete interpolation points into a continuous three-dimensional stratigraphic mesh, achieving an efficient and automated modeling process.

[0014] S3 Dynamic Model Update and Iterative Correction

[0015] A dynamic model update mechanism is established: When new engineering geophysical data, supplementary borehole exploration data, or real-time environmental monitoring data (tidal currents, wind speed, scour changes, etc.) are added, model reconstruction is automatically triggered: 1) Spatial matching and consistency checks are performed on the new data; 2) Stratigraphic interfaces, soil layer thicknesses, and geotechnical parameters are corrected; 3) Environmental constraint fields and engineering constraint boundaries are updated; 4) The grid is re-divided and attributes are mapped, enabling real-time iterative updates of the dynamic geological model. This mechanism ensures that the model can continuously optimize its accuracy as engineering exploration deepens and environmental monitoring data accumulates, always maintaining the best fit to the actual field conditions.

[0016] S4 Construction Suitability Red-Yellow-Green Intelligent Zoning

[0017] Based on an integrated geological model, a multi-factor comprehensive evaluation system was constructed, including: geological stability indicators, seabed scour risk, pile foundation constructability, soil and rock bearing capacity, environmental constraint intensity, aquaculture and ecological conflict risk, and navigation interference. A combined weighting method of analytic hierarchy process (AHP) and entropy weighting was used to divide the area into three zones through spatial overlay analysis: 1) Red Zone (Prohibited Construction Zone): ecological red line, core aquaculture area, main navigation channel, dense submarine pipeline area, shallow gas area, landslide area, and area with thick distribution of extremely soft soil; 2) Yellow Zone (Engineering Risk Zone): medium bearing capacity soil, locally scour-sensitive area, marginal aquaculture area, area with large strata undulation, and area with high geological uncertainty; 3) Green Zone (Recommended Construction Zone): a sweet spot for engineering projects with uniform strata, sufficient bearing capacity, weak scour, no obvious environmental constraints, and suitable for pile foundation layout and anchoring.

[0018] S5 Geological Risk Real-time Early Warning and Key Exploration Area Recommendations

[0019] Based on the geomechanical parameters, stratigraphic structure, and environmental dynamic conditions in the dynamic geological model, the system automatically identifies and alerts geological risks, including: shallow landslide risk, scour and erosion risk, uneven settlement risk of pile foundations, large deformation risk of soft soil, shallow gas overflow risk, and insufficient anchor chain holding power risk. Based on data sparsity, stratigraphic abrupt change points, high-risk areas, and model uncertainty distribution, the system intelligently recommends key areas for investigation, guides intensified drilling and supplementary geophysical exploration, and improves model accuracy and engineering safety.

[0020] S6 Results Visualization and Engineering Application Output

[0021] The platform integrates dynamic geological models, red, yellow, and green zoning, risk warning information, and key exploration recommendations into a 3D visualization platform. It supports real-time querying, cross-sectional analysis, engineering quantity estimation, and design scheme optimization, providing an integrated decision-making basis for offshore photovoltaic site selection, basic design, construction organization, and operation and maintenance safety.

[0022] Innovation

[0023] 1) Multi-dimensional data fusion innovation: For the first time, environmental data such as tidal currents, wind speed, aquaculture areas, and ecological red lines are systematically integrated with geological and engineering data to construct an integrated environmental-geological-engineering three-dimensional geological model of the seabed for marine photovoltaic projects, breaking through the limitations of traditional modeling that relies solely on geophysical and exploration data.

[0024] 2) Innovative dynamic modeling mechanism: Establish a dynamic geological model that supports real-time correction of newly added geophysical, exploration, and environmental data, enabling continuous iterative optimization of the model as data grows, thereby improving long-term applicability and accuracy.

[0025] 3) Innovation in red, yellow and green zoning and intelligent early warning: A construction suitability zoning system for offshore photovoltaics is formed, and geological risk is alerted in real time by combining geomechanical information. Key exploration areas are intelligently recommended, which significantly improves the level of intelligence in project site selection.

[0026] 4) Innovative adaptation to offshore photovoltaic scenarios: Unlike offshore wind power, which focuses on deep foundation modeling, this method focuses on adapting to the characteristics of near-shallow seabed, such as shallow strata, hydrodynamic sensitivity, and dense ecological constraints, and addresses key issues such as erosion, aquaculture conflicts, and shallow geological risks.

[0027] Beneficial effects

[0028] 1) More comprehensive data utilization: Deep coupling of multi-source information from environment, geology, and engineering is achieved, which fully reflects the constraints of offshore photovoltaic construction and avoids design errors caused by the omission of environmental factors.

[0029] 2) The model is more timely: The dynamic update mechanism enables the geological model to be continuously optimized as the project progresses, reducing exploration redundancy and improving the reliability of geological knowledge.

[0030] 3) More precise risk management: Clear red, yellow and green zoning, timely risk warnings, and scientific key survey areas can significantly reduce risks such as pile foundation instability, construction conflicts, and ecological violations.

[0031] 4) Higher engineering economics: Accurately identify the sweet spot of the project, optimize the pile position and array layout, reduce the cost of foundation reinforcement and construction risks, and improve the investment benefits of offshore photovoltaic projects.

[0032] 5) Wider range of applications: The marine dynamic geological modeling method that integrates environmental, geological and engineering multi-source data can be directly used for the construction of shallow sea wind farms, guiding the site selection and exploration engineering design of wind farms. Attached Figure Description

[0033] Figure 1 This is a flowchart of the dynamic geological modeling process for marine photovoltaic seabed based on multi-source environmental-geological-engineering data fusion in an embodiment of the present invention.

[0034] Figure 2 This is a diagram of the integrated structured digital model architecture of environment-geology-engineering in an embodiment of the present invention;

[0035] Figure 3 This is a schematic diagram of the red, yellow, and green zoning and risk warning of the marine photovoltaic seabed geological model in this embodiment of the invention. Detailed Implementation

[0036] The present invention will be further described in detail below with reference to specific embodiments.

[0037] This embodiment provides a method for dynamic geological modeling of the seabed for marine photovoltaic applications based on the fusion of multi-source environmental, geological, and engineering data. The process is as follows: Figure 1 As shown, the integrated structured digital model architecture is as follows: Figure 2 As shown, the red, yellow, and green zoning and risk warning diagram of the seabed geological model for marine photovoltaic applications are as follows: Figure 3 As shown. The specific implementation steps are as follows:

[0038] Step 1: Multi-source heterogeneous data acquisition and preprocessing

[0039] For a planned nearshore offshore photovoltaic site in western Guangdong, multi-source data collection was first conducted.

[0040] Geological data: Data from 40 boreholes, 10 CPT static cone penetration tests, multibeam bathymetry data covering the entire area, and 150 km of 2D seismic survey lines were collected. Interface elevations of various strata, including silt, clay, loess, and sand, as well as mechanical parameters such as cohesion, internal friction angle, and bearing capacity, were extracted from these data.

[0041] Environmental data: The following data were collected: tidal field simulation data (flow velocity, flow direction), wind speed statistics for the past 10 years, vector range of aquaculture areas released by the marine department, ecological protection red line range, coordinates of nearby shipping channels, and existing submarine pipeline data.

[0042] Engineering data: The parameters of the proposed pile foundation (diameter 2m, burial depth 15m) and the draft requirements of the construction vessel were collected.

[0043] All data were uniformly converted to the WGS84 coordinate system, outliers in the borehole data were removed using the 3σ criterion, and the environmental raster data was resampled to a 50m resolution grid, thus constructing a standardized multi-source database.

[0044] Step 2: Integrated 3D Geological Modeling

[0045] First, the seismic data underwent time-depth conversion, transforming the two-way travel time of the seismic profile into depth, with an average velocity of 1600 m / s used for the velocity model. Then, the seismic profile was comprehensively interpreted in conjunction with borehole data, identifying the boundaries of four main strata. Next, ordinary kriging was used to spatially interpolate the boundaries of each stratum, generating regular grid surfaces with 50 m spacing, thus obtaining the continuous distribution interfaces of each stratum.

[0046] Subsequently, using Python's PyVista library, these stratigraphic interface data were read, and Delaunay triangulation was performed to construct a three-dimensional volume model containing hexahedral meshes. Simultaneously, geotechnical parameters (such as bearing capacity) were interpolated onto the mesh nodes to form a geological parameter field; environmental data such as tidal flow velocity and aquaculture area masks were also mapped onto the mesh, ultimately constructing an integrated environmental-geological-engineering three-dimensional model containing 1 million mesh cells.

[0047] Step 3: Dynamic Model Update

[0048] In the early stages of the project, the model had high uncertainty due to limited exploration data. As the exploration work progressed, five supplementary borehole data points and high-resolution shallow seismic profile data were added. Upon detecting the new data, the system automatically triggered an update mechanism.

[0049] (1) The stratigraphic data from the new boreholes were compared with the original model to verify the consistency of the data;

[0050] (2) Add the new stratigraphic boundary points to the input dataset of the interpolation algorithm;

[0051] (3) The Kriging interpolation algorithm was rerun to update the interface elevations of each stratum;

[0052] The 3D mesh was re-divided using PyVista, generating a new dynamic geological model. The updated model reduced the formation prediction error near the newly added borehole from 1.2m to 0.3m, significantly improving accuracy.

[0053] Step 4: Red, Yellow, and Green Intelligent Partitioning

[0054] Based on the established integrated model, an evaluation index system was constructed. The weights of each index were calculated using the analytic hierarchy process (AHP) with entropy weighting, where environmental constraints accounted for 0.4, geological bearing capacity for 0.35, and scour risk for 0.25. Through spatial overlay analysis, the final zoning results were generated.

[0055] Red Zone: This mainly includes the aquaculture area on the southwest side and the navigation channel area on the northeast side. These areas are marked as no-construction zones.

[0056] Yellow Zone: This mainly includes the edge of the aquaculture area, areas where the local soft soil thickness exceeds 20m, and scour-sensitive areas with tidal current velocities greater than 2m / s. These areas are marked as engineering risk zones, and special foundation reinforcement is required if construction is to be carried out.

[0057] Green Zone: Located in the central part of the site, this area has uniform strata, with silty clay bearing capacity reaching 180 kPa, gentle tidal current velocity, and no environmental conflicts. It is marked as the recommended sweet spot for construction. Zoning results are as follows: Figure 3 As shown.

[0058] Step 5: Risk Warning and Survey Recommendations

[0059] The system identified risks based on model parameters:

[0060] In the northwest corner of the yellow zone, a soft soil layer with a thickness of 25m and a compression modulus of only 2MPa was identified, and the system issued an early warning of "risk of uneven settlement of pile foundation".

[0061] In the eastern part of the site, a tidal erosion rate of 5 cm / year was identified, and the system issued a warning of "risk of seabed erosion and hollowing".

[0062] Meanwhile, the system analyzed the uncertainty of the model and found that in an ancient river channel area in the middle of the field, due to the large spacing of the original survey lines, high data sparsity, and abrupt changes in strata, the system designated this area as a key exploration area and suggested adding 3 boreholes and 2 high-resolution geophysical survey lines in this area to reduce the uncertainty of the model.

[0063] Step 6: Visualize the output

[0064] Ultimately, all results were integrated into a WebGL-based 3D visualization platform, allowing design units to intuitively view the 3D geological structure, click on red, yellow, and green zones to view detailed constraints, and view the specific parameters of risk warning points, thereby guiding subsequent design and exploration work.

[0065] This embodiment demonstrates that the method of the present invention can effectively integrate multi-source data to construct a high-precision dynamic geological model, providing scientific and reliable decision support for the site selection and design of offshore photovoltaic projects.

Claims

1. A method for dynamic geological modeling of the seabed for marine photovoltaic applications based on the fusion of environmental, geological, and engineering multi-source data, characterized in that, Includes the following steps: S1: Multi-source heterogeneous data acquisition and standardized preprocessing: Collect and integrate three types of heterogeneous data sources: environmental data, geological data and engineering data. Perform coordinate system unification, spatiotemporal matching, outlier removal and gridded interpolation on the three types of heterogeneous data sources to construct a standardized multi-source database. S2: Integrated 3D geological modeling of environment, geology and engineering. Based on the seabed topography, an initial 3D stratigraphic structure model is constructed. Geomechanical parameters are assigned to the corresponding stratigraphic units to form a geomechanical parameter field. Tidal field, wind speed field, scour sensitivity, aquaculture area and ecological red line are superimposed to form an environmental constraint field. An integrated 3D geological model is constructed through spatial superposition and attribute mapping. S3: Dynamic model update and iterative correction based on Python open source library. When new engineering geophysical data, supplementary borehole exploration data or real-time environmental monitoring data are detected, the new data is automatically spatially matched and consistent. Stratigraphic interfaces, geotechnical parameters and environmental constraint boundaries are corrected. The ordinary Kriging interpolation algorithm is called to regenerate smooth stratigraphic interfaces. The PyVista library is used to re-subdivide the three-dimensional mesh to realize the real-time iterative update of the model. S4: Intelligent red-yellow-green zoning for construction suitability. Based on an integrated geological model, a multi-factor comprehensive evaluation system is constructed, which includes geological stability, seabed scour risk, pile foundation constructability, and environmental constraint intensity. The combined weighting method of analytic hierarchy process and entropy weighting is used to divide the study area into red, yellow and green zones through spatial overlay analysis. S5: Real-time geological risk warning and key exploration area recommendation. Based on the geomechanical parameters and environmental dynamic conditions in the dynamic geological model, it automatically identifies and prompts geological risks such as shallow landslides, scour and hollowing, uneven settlement, and shallow gas overflow. At the same time, based on data sparsity, stratigraphic abrupt change points and high-risk areas, it intelligently recommends areas that need to be focused on exploration.

2. The method according to claim 1, characterized in that, In S1, the environmental data includes ocean current velocity and direction, wave elements, seabed erosion rate, wind speed and wind rose diagram, aquaculture area range, shipping routes, ecological protection red lines, marine functional zoning, and distribution of submarine pipelines and obstacles; the geological data includes engineering geophysical data such as seabed shallow strata profiles, side-scan sonar, single / multi-channel seismic data, and resistivity imaging, as well as engineering survey data such as seabed borehole sampling, standard penetration test, static cone penetration test, and geotechnical test, and also includes geotechnical parameters such as stratum lithology, soil layer thickness, soil density, internal friction angle, cohesion, and bearing capacity characteristic values; the engineering data includes offshore photovoltaic array layout schemes, pile foundation types and dimensions, anchoring system parameters, draft and operational requirements of construction vessels, and foundation depth design parameters.

3. The method according to claim 1, characterized in that, In step S2, constructing the initial three-dimensional stratigraphic structure model specifically includes: performing time-depth conversion on the engineering geophysical data, converting the vertical coordinate of the seismic image from two-way travel time to depth values; identifying stratigraphic interfaces and extracting stratigraphic boundary data points through comprehensive interpretation of geotechnical investigation data and engineering geophysical data; and using ordinary kriging spatial interpolation algorithm to interpolate the stratigraphic boundary data points to generate a continuous and smooth stratigraphic interface grid surface.

4. The method according to claim 1, characterized in that, In S2, the construction of an integrated three-dimensional geological model specifically includes: using the PyVista open-source library to perform Delaunay triangulation on the interpolated stratigraphic interfaces to generate a hybrid three-dimensional mesh of hexahedrons and wedges; mapping the attribute information of the geomechanical parameter field and the environmental constraint field to the unit nodes of the three-dimensional mesh to realize the integrated expression of stratigraphic structure, soil and rock properties and environmental constraints.

5. The method according to claim 1, characterized in that, In S3, the dynamic model update and iterative correction specifically include: establishing a data-driven triggering mechanism to trigger the model reconstruction process when the amount of new data reaches a preset threshold or when a manual update instruction is received; performing consistency checks on the new data points and the original model data to remove abnormal data; re-executing the spatial interpolation algorithm to update the elevation data of each stratigraphic interface; and re-executing the three-dimensional mesh subdivision and attribute mapping to generate the updated dynamic geological model.

6. The method according to claim 1, characterized in that, In S4, the red zone is a prohibited construction zone, including ecological red line areas, core aquaculture areas, main navigation channels, areas with dense submarine pipelines, shallow gas areas, landslide areas, and areas with thick layers of extremely soft soil; the yellow zone is an engineering risk zone, including areas with medium bearing capacity soil, areas sensitive to local scour, peripheral aquaculture areas, areas with large strata undulations, and areas with high uncertainty in geological conditions; the green zone is a recommended construction zone, including sweet spots in engineering projects where the strata are uniform, the bearing capacity meets design requirements, the scour effect is weak, there are no obvious environmental constraints, and the pile foundation layout and anchoring are suitable.

7. The method according to claim 1, characterized in that, In S5, the real-time geological risk early warning specifically includes: calculating the scour and hollowing risk index based on the seabed scour rate and stratum scour resistance parameters; calculating the uneven settlement risk index of the pile foundation based on the soft soil layer thickness and compression modulus; calculating the shallow gas overflow risk index based on the shallow gas detection results and stratum pressure parameters; and issuing a corresponding risk warning when any risk index exceeds a preset threshold.

8. The method according to claim 1, characterized in that, In S5, the recommended key exploration areas specifically include: the spatial uncertainty distribution of the calculation model to identify areas where the data sparsity is greater than a preset threshold; the identification of stratigraphic abrupt change points and geological structure boundary areas; the identification of distribution areas of high-risk warning values; and, by combining the above three types of areas, delineating key exploration areas that require intensified drilling and supplementary geophysical exploration.

9. The method according to claim 1, characterized in that, Also includes: S6: Output of results visualization and engineering application, integrating dynamic geological models, red, yellow and green zoning, risk warning information, and key exploration suggestions into a 3D visualization platform, supporting real-time query, cross-sectional analysis, engineering quantity estimation and design scheme optimization.

10. The method according to claim 1, characterized in that, Unlike geological modeling of offshore wind farms, this method is tailored to the characteristics of near-shallow marine photovoltaics. It focuses on coupling the strata parameters of the shallow seabed (0-30m) with hydrodynamic environmental constraints, prioritizing the resolution of unique engineering problems of marine photovoltaics such as scour, aquaculture conflicts, and deformation of shallow soft soil.

Images