Method for providing information about a wooded area, and system

Abstract

The invention relates to a computer-implemented method (100) for providing information about a wooded area, comprising the following method steps: - receiving (102) data which contain, according to coordinates, properties of the wooded area, - processing (104) the data, - creating (106) at least one map model for the wooded area from the data, - providing (108) first information about the wooded area by detecting, segmenting and classifying non-living objects and / or regions and living objects and / or regions of the wooded area in the map model, and - providing (110) second information about the wooded area in the form of data sets and / or databases on the basis of the first information and / or the map model.

Description

[0001] Method for providing information about a forest area and system

[0002] The invention relates to a method for providing information about a forest area or a forest region. In particular, the forest area is a disaster area. The invention further relates to a system.

[0003] In order for foresters to decide what to do with a forest area (e.g., how to seed and plant it), all objects on the forest floor must be quantified, identified, and classified (i.e., forest assessment). Of interest are the amount of biomass present in the forest area and the condition of the soil.

[0004] Furthermore, forest soil analysis and its quantification and assessment are required by most forest stakeholders for applications beyond replanting and seeding (e.g., estimation of biomass content, soil topology, soil condition, etc.).

[0005] The forestry sector is characterized by a traditional approach to forest management. Management practices have been established over centuries. Decision-making is based on proven experience. Climate change is creating site dynamics, increasing the need for data-driven decision-making support.

[0006] In view of the advances in computer technology, it is of interest to provide solutions, such as an improved method or system, for providing information.

[0007] The problem is solved in particular by the features of the independent claims. Preferred embodiments are specified in the subclaims and in the description, each of which, individually or in combination, may represent an aspect of the invention. A computer-implemented method for providing information about a forest area is proposed, comprising the following method steps:

[0008] - Process step: Receiving data containing coordinate-based properties of the forest area. Receiving can take place in a data processing system, particularly a computer-implemented one.

[0009] - Method step: Processing the data. The processing can take place in the data processing system, in particular in a data processing module thereof. The processing includes, in particular, aggregation and / or filtering, and is, in particular, grid-related, raster-related, area-section-related, pixel-related, and / or property-related to the forest area.

[0010] - Method step: Creating at least one map model or multiple map models for the forest area from the processed data. This creation can be done using a mapping system, in particular a computer-implemented one. The mapping system can preferably communicate with the data processing system. In particular, the creation takes place directly or indirectly from the processed data.

[0011] - Method step: Providing initial information about the forest area by detecting, segmenting, and classifying non-living objects and / or areas, as well as living objects and / or areas of the forest area in the map model. In particular, the initial information is provided by a preferably computer-implemented assessment system, which operates according to predefined analysis criteria. The assessment system can preferably communicate with the mapping system.

[0012] - Process step: Providing second information about the forest area in the form of datasets and / or databases based on the first information and / or the map model. In particular, this provision is carried out by the assessment system, in particular by processing the first information and / or the map model.

[0013] - Optional method step: Providing third information about the forest area in the form of a visualization of the processed data and / or the first information and / or the second information and / or the data sets or databases, in particular by a user device that can preferably communicate with the data processing system, the mapping system and / or the assessment system.

[0014] In other words, a method is proposed in which data relating to a forest area or forest region is first received, in which the received data is processed, in which a virtual map of the forest area is generated based on this data, in which the virtual map is analyzed with regard to objects and areas contained therein, e.g., through detection / segmentation / classification, to provide first information, and in which the first information and / or the virtual map are processed to provide second information that is as organized as possible. Optionally, third information can be created through visualization.

[0015] The method can provide high-quality and detailed data and analyses on the forest area. Foresters and users can be assisted in making decisions regarding the conservation and / or restoration of the forest area. Foresters can advantageously draw on first, second, and / or third-order information. This information is particularly useful when it is organized and / or displayed or displayable.

[0016] The invention enables innovation through the higher level of detail of the data and information. The invention solves the challenges of data acquisition and processing at a high level of detail. Conventional forest management systems only work with data from stands (i.e., the forest in an area), such as stock density and other aggregated information. The innovation aims at high-resolution collection of this data down to individual trees and plants.

[0017] In particular, foresters can now better decide what to do with a forest area, e.g., how to reseed and replant it. They can learn about the amount of biomass present and the soil condition. Objects on the forest floor can be quantified, detected, and / or classified (i.e., forest assessment). Furthermore, forest soil analysis, its quantification, and assessment can be used by most forest stakeholders for applications beyond replanting and seeding (e.g., estimating biomass content, soil topology, soil condition, etc.).

[0018] In addition to foresters, other stakeholders, such as a public authority, also benefit from the invention. For example, funding measures for forest conversion can be decided or approved based on the information. Estimates of bound CO2 can be determined for a certificate. It can be used by responsible authorities for species mapping of plant populations. It can be used to monitor natural succession. Fire risk can be assessed, for example, for the fire department.

[0019] A further solution is proposed with the system according to the invention. The system comprises a computer-readable medium and / or a computer program product with instructions that, when executed on a computer, cause the computer to execute the method. The system can comprise the computer.

[0020] The invention supports data-driven decision-making in times when foresters can no longer rely on past experience. This is beneficial for assisting foresters in decision-making and ensuring that certain measures in the forest produce the desired results. Furthermore, efficiency is increased and costs are reduced. Furthermore, the results of the method can serve as a digital record and digital twin of the forest area, which can help foresters and landowners evaluate the success of their silvicultural measures.

[0021] The invention, with its essential features as well as its optional embodiments, leads to a number of advantages. It enables a simplified and more precise assessment of the forest area and its contents. It can estimate the carbon stock in biomass and soil. It can determine groundwater supply potential. It can assess the eligibility of reforestation programs for funding. It can examine reforestation measures already implemented. It can develop parameters for modeling ecosystems. It can count, identify, and characterize plants and trees. It can phenotype species. In particular, it enables habitat assessment and modeling of flora and fauna. It can verify the work results of contractors in accordance with forest certification programs.Foresters can be given guidance in planning management measures. Finally, seed planting patterns can be calculated and suggested.

[0022] The method is further developed in that the data comprises at least sensor data, satellite data, and / or public data. The data can be processed in the data processing module of the data processing system.

[0023] Boundaries of the forest area may have been identified. Representative data may be obtained from the forest area. For example, public data of varying quality are collected from authorities, forest owners, and / or forestry companies.

[0024] For example, spectral band analyses from satellites as satellite data (e.g. Galileo L2c, L1 C, MODIS, WorldEye-3), federal or state-wide open data providers (BayernAtlas, Geobasis NRW, ATKIS, ALKIS, BKG, Waldinfo. nrw, Geological Service, IPCC, DWD) with products such as (digital orthophotos, digital elevation models, laser scanner data, cadastral data, soil, geosurvey data, water supply data, climate models, weather data) can be used.

[0025] In particular, the sensor data is acquired by a remote sensing system, preferably one or more aircraft, in particular one or more unmanned aircraft. In the method, the aircraft or the multiple aircraft can be operated to acquire the sensor data. To provide the sensor data, the aircraft or remote sensing system, e.g. a drone, preferably with a fixed wing and / or with multiple rotors, can be instructed to fly over the region or forest area in question. The aircraft follows, for example, a predefined flight path and / or can use a mode in which the flight path is updated, in particular live, during the flight. A sensor device of the aircraft is preferably attached to a gimbal. The sensors can include LiDAR, RADAR, visual cameras, multispectral, hyperspectral and infrared sensors, depth sensors, ground radar, etc.

[0026] The sensors may alternatively or additionally include the following sensors: color image sensor, chemical sensor, in particular VOC, passive radio sensor and / or radiometer sensor.

[0027] A sensor device is oriented downwards, preferably straight down, in order to collect information about the forest area, particularly from an orthographic perspective. The sensor device can also be aligned automatically. The sensor device has, for example, a spectral sensor with a zoom lens. The position of the spectral sensor and preferably also the zoom level are typically actively controlled during the flight. Areas or surfaces where vegetation is suspected or where higher resolution is desired are preferably identified in the data stream of the first sensor. The sensor device preferably aligns the sensors to the areas of interest and preferably records image data with higher spatial resolution.

[0028] Data processing can include, for example, geographic and / or pixel-precise alignment and registration. The data can be fused to increase fidelity. Fusion methods used include pixel-based fusion and / or feature-based fusion. Fusion can also occur at a temporal level, for example, when an area is revisited by the aircraft.

[0029] Processed data can be used together with other data sources (e.g., via ground survey points, ground control points, or similar) to create high-quality maps as map models of the forest floor or forest area. The map model includes digital terrain models, digital surface models, orthomosaics, probabilistic models, maps with vegetation indices (NDVI, SAVI, etc.), probabilistic soil maps, 3D elevation models, and / or 3D models. In particular, temperature maps of the forest floor are created and / or the correlation with the detected objects is quantified. The maps may include the mapping of objects and factors below the forest floor.

[0030] In particular, a forest analysis and forest assessment are carried out, preferably to provide the initial information, especially with regard to the map model. This preferably includes the detection of all non-living objects such as debris (dead wood or logging debris, tree stumps, trunks, branches, etc.), rocks, bodies of water, roads, etc. Detection preferably includes living objects such as vegetation (grass, weeds, naturally growing seedlings, etc.). Classification and segmentation are preferably performed to classify the type of detected objects using, preferably supervised and / or unsupervised, machine learning or artificial intelligence methods, such as probabilistic data-based methods, few-shot learning methods, instance segmentation, multi-instance learning, etc.Quantifying the extent of damage in a disaster area is also estimated and may include the number and type of trees dead or felled, the size of the disaster area, etc.

[0031] A morphological analysis of the forest soil can be performed, which can quantify, for example, the terrain characteristics (e.g., slope) and soil properties (e.g., moisture content, available nutrients, pH, etc.). The analysis usually also includes a probabilistic assessment of the seeding potential of a given area based on a seed type and planting time, as well as a probabilistic quantification and analysis of the depths of the forest layers (humus, mineral soil, etc.) and subterranean obstacles. Based on the extensive analyses above and below the forest floor, specific forestry information can be generated upon request, e.g., optimal seeding locations, compliance with forestry regulations (e.g., FSC).

[0032] In particular, the results of the analysis and evaluation, or rather, their provision, can be summarized in usable and presentable databases and data representation layers. These include, for example, geospatial databases, probabilistic layers, integration into existing global maps with different coordinate systems, etc., which can take the probability of the analysis and evaluation into account.

[0033] Preferably, the data, including, for example, raw data, collected data, assessment maps, and results, can be stored and visualized on a user-friendly platform with a graphical user interface. The platform can be hosted, for example, in the cloud, on a local server, or other storage medium, accessible by logging in and interacting with the maps, digital twins, and assessment information. Additionally, the data and assessment results can be used for an augmented reality module with which users can interact. This would provide foresters and other forest stakeholders with the information they need for their applications (e.g., reforestation and seeding, carbon biomass estimation, damage quantification, etc.). The data processing system can receive sensor data from one or more airborne remote sensing platforms or aircraft.The satellite data can refer to various satellites, e.g., spectral or image data. The open public data refers in particular to various data on forests, their soil, and topology. Public data includes, among others, public government sources (e.g., LUCAS dataset, BayernAtlas, Geobasis NRW, ATKIS, ALKIS, BKG, Waldinfo.nrw, Geological Service, IPCC, DWD, etc.) and / or private sources. The data processing module can, in particular, combine all of the above-mentioned data pixel-wise and / or feature-wise. The dynamics and movement commands of the aircraft can be used to improve the quality of the data. The data processing system can interact with the mapping system to exchange data or criteria.

[0034] The mapping system can contain or work with various components or contents, for example:

[0035] - the processed data and / or the map model,

[0036] - an information-gain-based path planning module,

[0037] - a 2D mapping module,

[0038] - a 3D mapping module, and / or

[0039] - a multilayer map mapping module or map display module.

[0040] The processed data can be provided or displayed in a grid- or raster-based, pixel-based and / or feature-based manner.

[0041] The path planning module may include an information metric to be optimized, preferably one that generates the paths to be followed by the aircraft, preferably one that can maximize mapping efficiency and reduce costs.

[0042] The 2D mapping module can contain or create orthophotos, surface or terrain models, soil maps, and / or data. Ground control points, data, and / or features can be used to create the maps. The 3D mapping module can use 3D point clouds from sensor data, fused data, and / or photogrammetric methods to represent the terrain in 3D. Soil data can be used to create the digital twin of the forest floor.

[0043] The multilayer map mapping module can contain multiple layers of different maps based on different sensor data or features, or pixels or 3D point clouds. The mapping system can interact with the assessment system to exchange data and / or criteria.

[0044] The assessment system can contain or work with various components or modules, for example:

[0045] - the map model and / or at least one map or maps,

[0046] - a module for object recognition and object classification,

[0047] - a module for detection, segmentation and classification,

[0048] - a module for forest area attributes or soil attributes,

[0049] - an expert knowledge module, and / or

[0050] - an aggregation module for analysis and evaluation.

[0051] Analysis criteria can form the basis of the assessment system or be fed into it. These analysis criteria can be factors and / or variables that are of interest depending on the intended application, such as a desired resolution and / or confidence interval. The analysis criteria can be modified and / or predefined, for example, at the user's device.

[0052] The map can be 2D or 3D, including but not limited to vegetation and soil maps, as well as soil data and points.

[0053] Specifically, the object detection and classification module utilizes data from the map model or maps, which can be run through computer vision algorithms or pre-trained artificial intelligence algorithms to detect objects. Classifying objects into their respective classes can include tree stumps, trunks, twigs, branches, rocks, etc. The object detection and classification module can incorporate map data and / or other data, as well as vegetation indices (e.g., NDVI, SAVI, etc.) to detect and classify plant types. The segmentation method can segment patches of vegetation and / or branches on the forest floor.

[0054] Both the object detection and classification module and the detection, segmentation, and classification module can use supervised and unsupervised machine learning and artificial intelligence methods, such as probabilistic data-based methods, few-shot learning, instance segmentation, multi-instance learning, etc. Within supervised or unsupervised learning, multiple pre-trained models can be used to improve detection, classification, and segmentation.

[0055] Forest area attributes are defined as attributes relating to the properties of the forest area. Forest area attributes include various properties relating to the ground itself (soil, caves, etc.) and the features above ground (plants, objects, buildings, bodies of water, etc.).

[0056] The forest area attributes module or soil attributes module can use the maps, data, and / or open databases, for example, to create a probabilistic model of the terrain (such as slope, elevation, etc.) and soil factors (such as water content, nutrients, etc.). Probabilistic attributes can include both attributes above the forest floor and attributes and factors below the forest floor, such as subsurface rocks, root systems, water, etc.

[0057] The forest area attributes or soil attributes module can use vegetation maps and indices and / or other assessment results such as object detection. The expert knowledge module can incorporate expert knowledge related to forests; for example, the presence of a particular plant species can indicate the presence of water and nutrients. This expert knowledge can be modified and / or predefined or entered at the user facility.

[0058] The aggregation module for analysis and assessment can combine the results of the previous modules into a specific data type or representation, which can be specified by the analysis criteria. The assessment system can interact with the mapping system to exchange data or criteria.

[0059] The user device can contain or work with various components or modules, for example

[0060] - a visualization module, and / or

[0061] - includes an augmented reality module.

[0062] The user facility may receive data from the data processing system, the mapping system and / or the assessment system.

[0063] The visualization module can visualize the first, second, and / or third information, or the data, data sets, and / or databases. In particular, it is a graphics module and / or graphics program.

[0064] The augmented reality module can use data and assessment results to display them in maps and / or use them with handheld devices, e.g. smartphones or computers.

[0065] The user device can be hosted on a cloud, a remote server, a local server, and / or another type of data storage. The user(s) can log in to the user device and visualize and / or interact with data, maps, and / or analysis results. The results and / or other custom products can be shared, exported, and / or integrated into existing solutions. The user device interacts, in particular, with the data processing system, the mapping system, and / or the assessment system, particularly to exchange data or criteria.

[0066] The modules, systems and / or other process components mentioned above and below can be implemented individually or jointly by computer in order to reduce effort.

[0067] The method is advantageously further developed by operating the aircraft over the forest area to provide the sensor data via a sensor device of the aircraft. The sensor device preferably provides the sensor data using one or more of the following sensors: LiDAR, RADAR, multispectral, hyperspectral, visibility, depth, GPS, RTK GPS, IMU, infrared, ground-penetrating radar, and / or electromagnetic sensor. In this respect, the sensor device can comprise the sensor(s) to generate the sensor data during flight. The aircraft can capture the sensor data in a targeted manner, thus providing further improvement in information in specific areas or regions.

[0068] The method is advantageously further developed by creating a flight path in / on the forest area or forest region using an information metric to be optimized. The flight path can be sent to the aircraft to provide additional sensor data with improved resolution along the flight path. The aircraft can fly or follow the flight path and generate the additional sensor data. In particular, the mapping system has an information acquisition flight path planning module that creates the flight path.

[0069] The method is advantageously further developed by generating the map model as a multilayer map model. In particular, the mapping system comprises the multilayer map mapping module, which can generate or generates the multilayer map model. In this way, a large amount of information on specific areas can be provided in aggregated form.

[0070] The multilayer map model can be two-dimensional, or a 2D model. The multilayer map model can be two-and-a-half-dimensional, or a 2.5D model. The multilayer map model can also be three-dimensional, or a 3D model.

[0071] In a multilayer map model, especially a 2D model, information is presented primarily on a flat plane. This typically includes maps that display geographical information. In a multilayer model, different data layers can be superimposed, for example, a base map with layers above it containing additional information. Each layer can be shown or hidden to analyze different data combinations.

[0072] A 2.5D model extends the 2D model by, for example, adding additional visual depth and / or elevation information without creating full 3D models. This can be achieved through techniques such as shadows, exaggerations, value markings, and other effects that create an illusion of depth or indicate depth quantitatively. One example is topographic maps, which can represent elevation differences using color gradients or hatching. This can be useful for intuitively understanding the geography of an area or for having elevation information at hand.

[0073] In a 3D model, data can be represented in an at least essentially fully simulated three-dimensional environment. This can take the form of terrain features and other structures that are modeled in their actual physical form. Different levels of 3D data can be integrated in a 3D model, such as the modeling of geometric structures, the overlaying of different information in a three-dimensional environment, or the representation of patterns or flight paths. The method is further developed by providing physical, chemical and / or biological soil attributes or forest area attributes. Error estimates or confidence intervals can be provided for the forest area attributes. In particular, probability-based forest area attributes can be provided in this regard. For this purpose, the assessment system can use the module for forest area attributes orHave soil attributes that can or does provide forest area attributes.

[0074] The method is further developed by providing expert knowledge-based forest area attributes or soil attributes, in particular, these include: planting distances, planting density, number of seeds per capsule, seed material per seeding site, seeding sites per nest, and / or soil depths. In particular, the assessment facility comprises the expert knowledge module, which can provide the expert knowledge-based forest area attributes.

[0075] The forest area attributes include, in particular, soil slope, soil water content, soil nutrient content, subsurface rock occurrence and / or subsurface root system occurrence.

[0076] The method is further developed by using the visualization module and / or the augmented reality module for visualization. In particular, the user device has the visualization module and uses it to visualize. This allows the information or data to be provided in a more user-friendly manner.

[0077] The method is further developed by operating the method in at least one step on a server-based and / or terminal-based basis. In particular, the user device is operated on a server-based and / or terminal-based basis. Several or all steps of the method can be operated in this way. This allows for local resource savings. Self-sufficiency can also be achieved. Insemination means, in particular, planting the soil with at least one seed.

[0078] Soil includes, in particular, the earth's surface, including the subsoil. Soil includes, in particular, all objects on the earth's surface, such as trees, dead trees, rocks, and the like. Soil is part of the forest area.

[0079] The aircraft is, in particular, a flight-capable device. The aircraft is, in particular, wirelessly remote-controlled and / or unmanned. The aircraft may be a drone. The aircraft may be radio-capable.

[0080] In the context of the disclosure, the abbreviation “resp.” is a short form for “respectively” and is intended to indicate alternative, essentially equivalent and / or synonymous features or terms in order to better convey the idea or meaning of a feature or term.

[0081] “Respectively” can always be replaced with “and / or”.

[0082] The invention will be explained below by way of example with reference to the accompanying drawings using preferred embodiments, wherein the features presented below can represent an aspect of the invention both individually and in combination. They show:

[0083] Fig. 1 shows a method according to the invention in a schematic view,

[0084] Fig. 2 shows a system according to the invention in a schematic view, and

[0085] Fig. 3 shows the method according to the invention in a further schematic view. Fig. 1 shows a sequence of a computer-implemented method 100 according to the invention. The method is shown as an example in Fig. 3, which will be discussed in more detail below.

[0086] At reference numeral 102, various data about a specific forest area whose boundaries have been determined are determined. The data includes sensor data 202, satellite data 204, and public data 206. The data relates to coordinate-based properties of the forest area. The data is collected by authorities, forest owners, and / or forestry companies. The data includes spectral band evaluations from satellites and / or digital orthophotos, digital elevation models, laser scanner data, cadastral data, soil and geosurvey data, water supply data, climate models, and / or weather data, in particular obtained from federal and / or state-wide open data providers.

[0087] Step 102 may take place in a particularly computer-implemented data processing system 210. The data processing system 210 may include a data processing module 208.

[0088] Some of the data is collected using sensors from a remote sensing system, namely an unmanned aerial vehicle such as a drone. The aircraft follows a custom-created flight path or can use a mode in which the flight path is updated live during the flight. The flight path is within the forest area, i.e., in the air above the corresponding ground level of the forest area.

[0089] The aircraft's sensor array is attached to a gimbal for data acquisition. The sensor array includes an infrared sensor and a radar. The aircraft can provide additional sensor data with improved resolution along the flight path.

[0090] The sensor device is specifically directed straight down from the aircraft to collect information about the forest area from an orthographic perspective. The sensor device has a spectral sensor with a zoom lens. The position of the spectral sensor and the zoom level are actively controlled during flight. Objects or areas, particularly vegetation, that require higher resolution are identified in the data stream of the first sensor. The sensor device aligns the sensors to the areas of interest and records image data with higher spatial resolution.

[0091] In the next method step, at reference numeral 104, processing is performed. Here, in particular, the data received or provided in step 102 are filtered and processed, in particular aggregated. The data are further processed, in particular, using computer implementation to ensure high data quality. Such processing can include, for example, geographical and pixel-precise alignment and registration. The data can be merged to increase fidelity.

[0092] Step 104 may occur in data processing system 210.

[0093] In the next method step, at reference numeral 106, a map model, particularly a multilayer map model, is created. The processed or merged data, preferably together with other data (e.g., from ground survey points, ground control points, manually entered data, etc.), are used to create high-quality maps of the forest area. Temperature maps of the forest area are created, and the correlation with the detected objects can be quantified.

[0094] Step 106 can be performed in a particularly computer-implemented mapping system 212.

[0095] In the next method step, at reference numeral 108, initial information about the forest area is provided, particularly with reference to the map model. A forest analysis and assessment is performed. This includes the detection or recognition of all non-living objects such as debris, deadwood, or logging debris. Detection also includes living objects such as vegetation. Furthermore, segmentation and classification are performed to classify the type of detected objects using machine learning methods. The quantification of the extent of damage in a disaster area is also estimated. Predefined analysis criteria 222 can be applied.

[0096] The step 108 can be used in a particularly computer-implemented assessment system 220, which can operate according to the predefined analysis criteria 222.

[0097] In addition, during provision 108, a morphological analysis of the forest area is performed, which can quantify, for example, the terrain characteristics (e.g., slope) and soil properties (e.g., moisture content, available nutrients, pH, etc.). The analysis also includes a probabilistic assessment of the seeding potential of a given area based on the seed type and planting time, as well as a probabilistic quantification and analysis of the depths of the forest layers (humus, mineral soil, etc.) and subterranean obstacles. Based on the extensive analyses above and below the forest area, specific forestry information can be generated upon request, e.g., optimal seeding locations, compliance with forestry regulations (e.g., FSC).

[0098] The provision 108 includes physical, chemical and / or biological soil attributes including error estimates or confidence intervals.

[0099] The provision 108 comprises expert knowledge-based forest area attributes, including planting distances, planting densities, number of seeds per capsule (i.e., per capsule that can or should be introduced into the soil to inseminate the forest area during reforestation), seed material per seeding location, seeding locations per clump, and / or soil depths. Soil water content, soil nutrient content, subsurface rock occurrence, and subsurface root system occurrence can also be provided. In the next method step at reference numeral 110, second information about the forest area is provided. The second information is provided as data sets based on the first information and the map model. In other words, the results of the analysis and evaluation by the provision 108 are summarized in usable and presentable databases and data representation layers. These include, for example,geospatial databases, probabilistic layers, integration into existing global maps with different coordinate systems, etc., especially those that can take into account probability of analysis and evaluation.

[0100] Step 110 can be performed in the assessment system 220.

[0101] In the next method step at reference numeral 112, third information about the forest area is provided in the form of a visualization of the particularly processed data, optionally the map model, the first information, the second information, the data sets, and the databases. Essentially, the data or information from at least one, several, or all of the previous steps 102, 104, 106, 108, 110 are stored on a user-friendly platform, in particular a user device 234 with a graphical user interface, and visualized via a visualization module 236. The user device 234 can, in particular, communicate bidirectionally with the data processing system 210, the mapping system 212, and the assessment system 220.

[0102] The user device 234 can be hosted, for example, in a cloud or on a local server or other storage medium that can be accessed by logging in. The data or the first / second / third information can be used for an augmented reality module 238 with which users, e.g., foresters, can interact.

[0103] Fig. 2 shows a system 300 with a computer 302 and with a computer-readable medium 304. The medium 304 contains a computer program product 306 with instructions which, when the computer program product 306 or the instructions are executed on a computer 302, cause the computer 302 to execute the method 100 described above.

[0104] Fig. 3 shows schematically and exemplarily some links and processes that lead to the realization of the method described above.

[0105] The computer-implemented data processing system 210 includes the sensor data 202, the satellite data 204, and the public data 206.

[0106] The sensor data 202 is received from the aircraft. The satellite data 204 in this case refers to several products from different satellites, in this case spectral and image data. The public data 206 represents various data about forests, their soil, and topology, including, in particular, the relevant forest area.

[0107] In the data processing module 208, the data 202, 204, 206 can be processed, in particular aggregated, filtered, merged, and / or combined. The data processing module 208 can combine the data 202, 204, 206 pixel-wise or feature-wise. The aircraft can be used to improve the quality of the data through new data acquisition.

[0108] Once the data 202, 204, 206 have been processed, they are passed as processed data 214 to the mapping system 212, which creates the map model.

[0109] The mapping system 212 receives the processed data 214.

[0110] The data processing system 210 can interact with the mapping system 212, in particular communicate bidirectionally, to exchange data or criteria.

[0111] In this case, the mapping system shows 212

[0112] - the processed data 214,

[0113] - an information acquisition trajectory planning module 215,

[0114] - a 2D mapping module 216, a 3D mapping module 218, and a 3D or multilayer map imaging module 219.

[0115] The mapping system 212 is computer-implemented. Modules 215, 216, 218, and 219 are computer-implemented.

[0116] The processed data 214 can be displayed or mapped in the map model in a grid- or raster-based, pixel-based or feature-based manner.

[0117] The trajectory planning module 215 includes an information metric to be optimized that generates paths to be followed by the aircraft that can maximize mapping efficiency and reduce costs.

[0118] The 2D mapping module 216 may contain orthophotos, surface or terrain models, soil maps, and data. Ground control points, data, and features are used to create the map model.

[0119] The 3D mapping module 218 may use 3D point clouds from the sensor data 202, from the processed data 214 and / or from photogrammetric methods to represent the terrain in 3D, in particular as the map model or as part thereof.

[0120] The multilayer map mapping module 219 can contain one, two, or more layers of different maps based on different sensor data 202, features, and / or pixel or 3D point clouds. The mapping system 212 can interact with the assessment system 220 to transmit data or criteria, in particular to exchange them bidirectionally.

[0121] The Assessment System 220 includes

[0122] - Maps 224 including the map model,

[0123] - a module for object recognition and object classification 226,

[0124] - a module for detection, segmentation and classification 228,

[0125] - a module for forest area attributes or soil attributes 230, - an expert knowledge module 231 , and

[0126] - an aggregation module for analysis and evaluation 232 or works with it.

[0127] The modules are computer-implemented. Analysis criteria 222 are also incorporated into the assessment system 220. The assessment system 220 can also communicate with the data processing system 210 or receive data 214 from it.

[0128] The assessment system 220 provides initial information about the forest area by detecting, segmenting and classifying non-living objects and / or areas as well as living objects and / or areas of the forest area in the map model.

[0129] The assessment system 220 provides second information in the form of data sets based on the first information and the map model.

[0130] The maps 224 are provided in particular on the basis of the map model from the mapping system 212.

[0131] The object detection and classification module 226 uses the data from the maps 224 to detect objects. The classification of the objects into classes can include tree stumps, trunks, twigs, branches, rocks, etc.

[0132] The detection, segmentation, and classification module 228 deals with the vegetation in the respective forest area. Module 228 can contain maps 224 and vegetation indices to detect and classify the type of plants. Segmentation can segment patches of vegetation or branches on the forest floor or within the forest area.

[0133] Machine learning is used particularly in modules 226 and 228.

[0134] In particular, in the method in the present or another embodiment, a growth simulation can be carried out with respect to the plants, trees, etc., for example, by the assessment system 220, in particular by means of one or more of the modules 226, 228, 230, 231, and / or 232. Preferably, the growth simulation can characterize or include a time-dependent progression of a stand. The growth simulation can provide information, for example, first, second, and / or third information.

[0135] The forest area attributes or soil attributes module 230 may use the data, the processed data, and the map model(s) 224 to create a probabilistic model of the terrain and soil factors.

[0136] The forest area attributes can include both above-ground and below-ground attributes and factors. The attributes relate to physical, chemical, and / or biological properties of the forest area. Error estimates are provided for the attributes.

[0137] The expert knowledge module 231 can use vegetation maps and indices, as well as other assessment results such as object detection. The module 231 can incorporate expert knowledge regarding forests as needed; for example, the presence of a particular plant species can indicate the presence of water and nutrients. Expert knowledge-based forest area attributes can be provided, such as planting distances. The forest area attributes can be probability-based.

[0138] Forest area attributes include, for example, soil slope, soil water content, soil nutrient content, subsurface rock occurrence, and subsurface root system occurrence.

[0139] The analysis and evaluation aggregation module 232 can combine the results of the preceding modules 226, 228, 230, and 231 into a data type and / or representation that can be specified by the analysis criteria 222. The assessment system 220 can communicate or interact with the mapping system 212. The user device 234 can receive content directly or indirectly from each of the systems 210, 212, and 220. Communication with the systems 210, 212, and 220 is optionally bidirectional.

[0140] The user device 234 provides third information about the forest area in the form of a visualization of the processed data and the first information and the second information or the data sets.

[0141] The user device 234 has

[0142] - the visualization module 236 and

[0143] - the augmented reality module 238.

[0144] The augmented reality module 238 can receive or use the data, data such as the processed data, the map model, the first information, the second information and the third information to visualize them.

[0145] The user device 234 is hosted on a cloud. Users can log in to the cloud, for example, using a terminal device, and visualize and interact with the data, the processed data, the map model, the first information, the second information, and / or the third information.

[0146] The invention underlying this patent application was developed in the project called “Garrulus” funded by the MULNV.

[0147]

[0148] 100 procedures

[0149] 102 Receive

[0150] 104 Process

[0151] 106 Create

[0152] 108 Provision

[0153] 110 Provision

[0154] 112 Provision

[0155] 202 sensor data

[0156] 204 satellite data

[0157] 206 public data

[0158] 208 Data processing module

[0159] 210 Data processing system

[0160] 212 mapping system

[0161] 214 processed data

[0162] 215 Information Gathering Trajectory Planning Module

[0163] 216 2D mapping module

[0164] 218 3D mapping module

[0165] 219 Multilayer map mapping module

[0166] 220 Assessment System

[0167] 222 analysis criteria

[0168] 224 Map

[0169] 226 Module for object recognition and object classification

[0170] 228 Module for detection, segmentation and classification

[0171] 230 Module for forest area attributes or soil attributes

[0172] 231 Expert knowledge module

[0173] 232 Aggregation module for analysis and evaluation

[0174] 234 User setup

[0175] 236 Visualization module

[0176] 238 Augmented-Reality-Modul

[0177] 300 System

[0178] 302 Computer

[0179] 304 Medium

[0180] 306 Computerprogrammprodukt

Claims

Patent claims 1 . A computer-implemented method (100) for providing information about a forest area, comprising the following method steps: - receiving (102) data containing coordinate-based properties of the forest area, - Processing (104) of the data, - creating (106) at least one map model of the forest area from the data, - providing (108) first information about the forest area by detecting, segmenting and classifying non-living objects and / or areas as well as living objects and / or areas of the forest area in the map model, and - Providing (110) second information about the forest area in the form of data sets and / or databases based on the first information and / or the map model.

2. Method (100) according to the preceding claim, comprising the method step: - Providing (112) third information about the forest area in the form of a visualization of the data, the first information, the second information, the data sets and / or the databases.

3. Method (100) according to one of the preceding claims, wherein the data comprises at least sensor data (202), satellite data (204) and public data (206), wherein the sensor data (202) is acquired by an aircraft.

4. Method (100) according to the preceding claim, comprising the following method step: Operating the aircraft over the forest area to provide the sensor data (202) by a sensor device of the aircraft, wherein the sensor device preferably provides the sensor data (202) by means of one or more which provides the following sensors: LiDAR, RADAR, Multispectral, Hyperspectral, Visibility, Depth, GPS, RTK GPS, IMU, Infrared, Ground Penetrating Radar and / or Electromagnetic Sensor.

5. Method (100) according to one of the preceding two claims, wherein a flight path in the forest area is created by means of an information metric to be optimized, and wherein the flight path is sent to the aircraft to provide further sensor data (202) with improved resolution along the flight path, and the aircraft flies the flight path and generates the further sensor data (202).

6. Method (100) according to one of the preceding claims, wherein the map model is generated as a multilayer map model.

7. Method (100) according to one of the preceding claims, wherein physical, chemical and / or biological soil attributes and preferably error estimates for the soil attributes are provided.

8. Method (100) according to one of the preceding claims, wherein expert knowledge-based forest area attributes are provided, in particular wherein these include: planting distances, planting density, number of seeds per capsule, seed material per seeding location, seeding locations per clump and / or soil depths.

9. The method (100) according to any one of the preceding claims, wherein the probability-based forest area attributes and / or the expert knowledge-based forest area attributes comprise a soil gradient, a soil water content, a soil nutrient content, a subsurface rock occurrence and / or a subsurface root system occurrence.

10. Method (100) according to one of the preceding claims, wherein a visualization module (236) and / or an augmented reality module (238) is / are used for visualization.

11. Method (100) according to one of the preceding claims, which is operated in at least one method step in a server-based and / or terminal-based manner.

12. System (300) comprising a computer-readable medium (304) and / or a Computer program product (306) with instructions which, when the computer program product (306) is executed on a computer (302), cause the computer (302) to execute the method (100) according to one of the preceding claims, in particular wherein the system comprises the computer (302).

Images