A method for remote sensing data asset value assessment and pricing using AI

Abstract

The present application relates to the field of business and financial data processing, and discloses a method for remote sensing data asset value evaluation and pricing by using AI, which comprises: establishing a set of computer data processing procedures, which combines three utility components representing information novelty task scene matching degree and data structure clarity to generate a standardized differential utility fingerprint, the present application discloses a black box price prediction, and converts data asset value judgment into objective observation on a utility profile which can be traced and understood, thereby providing transparent and reliable technical basis with inherent logic self-checking ability for pricing decision in the field of finance and asset management.

Description

TECHNICAL FIELD

[0001] The present application relates to a method for remote sensing data asset value assessment and pricing using AI, belonging to the field of business and financial data processing technology. BACKGROUND

[0002] A current mainstream technical approach is to build a complex prediction model based on deep learning, which attempts to directly regress a price or value level by inputting multi-dimensional technical features of remote sensing data and historical transaction data. This to some extent realizes the automatic preliminary quantification of data asset value.

[0003] However, when this approach is applied to professional fields such as financial insurance or critical asset management that have strict requirements for decision-making timeliness and result reliability, its inherent limitations begin to appear. In these scenarios, the value of data assets is not an isolated objective scalar that can be accurately predicted, but rather a reflection of the utility it can provide in terms of key decision-making information under specific business tasks. The existing approach, which focuses excessively on fitting a dynamic and subjective price variable endogenous to the market, generally overlooks the data task utility itself. This oversight often leads to situations in actual business where high-priced data cannot be used due to mismatched information dimensions and task requirements, resulting in decision-making errors and economic losses.

[0004] In the face of the dilemma of disconnection between evaluation results and actual utility, a direct improvement idea is to further increase the complexity of the model or introduce more market-related factors, hoping to improve the accuracy of the prediction. However, this approach does not address the root of the problem and instead exacerbates the black box nature of the evaluation process and computational costs, making it even more difficult for professionals in the financial sector to base their core business decisions on an algorithmic model whose logic is not traceable and whose results are not interpretable. Specifically, the existing technology has the following deficiencies: 1. Simplifying the core problem of data task utility into a prediction problem of market price, leading to a disconnection between evaluation goals and business essence; 2. The evaluation process is highly dependent on complex black box models, whose internal logic is not interpretable, making it difficult for evaluation results to gain the trust of professional decision-makers; 3. Lack of a standardized data utility measurement that can span different data sources and different task scenarios, leading to a lack of objective and reliable basis for data procurement and risk management. Therefore, how to break free from the mindset of predicting a single price and establish an evaluation and pricing method that can transparently and standardizedly reveal the core utility of remote sensing data for specific business tasks at a low cost has become a technical problem to be solved by the present application. SUMMARY

[0005] The application provides a method for remote sensing data asset value evaluation and pricing by using AI, which mainly aims to solve the problem that the existing technology simplifies the data utility evaluation error as price prediction, resulting in the black-box evaluation process and the untrustworthy and lack of standardized utility measurement.

[0006] To achieve the above-mentioned purpose, the application provides a method for remote sensing data asset value evaluation and pricing by using AI, which establishes a set of computer data processing procedures, and the procedures include the following steps:

[0007] Step a, generating a differential utility fingerprint representing the decision utility of remote sensing data under a specific task, which specifically includes: after obtaining the remote sensing data to be evaluated and a known reference remote sensing data, performing a reference data reliability gating review: applying a signal processing method to the known reference remote sensing data to obtain its own structural clarity component, and judging whether the structural clarity component meets the reliability threshold; if the judgment result is not satisfied, another known reference remote sensing data is selected and the review is repeated until the effective reference data meeting the reliability threshold is found; after the review determines the effective reference data, the difference between the to-be-evaluated data and the effective reference data in the information amount distribution is calculated to obtain the first utility component representing the information novelty of the to-be-evaluated data; based on a task template defining the requirements of a specific task, the metadata of the to-be-evaluated data is logically matched with the task template to obtain the second utility component representing the scene matching degree of the to-be-evaluated data; a signal processing method is applied to the to-be-evaluated data to calculate the ratio of its internal structured information and unstructured information to obtain the third utility component representing the structural clarity of the to-be-evaluated data itself;

[0008] Step b, combining the first utility component, the second utility component and the third utility component to form the differential utility fingerprint of the to-be-evaluated remote sensing data;

[0009] Step c, outputting the differential utility fingerprint as a standardized technical data structure to support the value evaluation and pricing decision of the remote sensing data asset.

[0010] Preferably, the way of obtaining the first utility component, the second utility component and the third utility component in step a is further defined as: the first utility component is obtained by calculating the gray level histogram of the predetermined waveband of the to-be-evaluated data and the effective reference data respectively, and using K-L divergence to measure the difference between the two gray level histograms; the second utility component is obtained by performing logical comparison and weighted summation on the waveband information, spatial resolution and imaging time information contained in the metadata of the to-be-evaluated data, and the key waveband requirement, optimal resolution range requirement and imaging time requirement defined in the task template for the specific task; and the third utility component is obtained by applying Laplacian operator to the to-be-evaluated data for filtering, and calculating the variance ratio of the filtered image and the original image.

[0011] Preferably, in step a, the way of calculating the structural clarity component is precisely defined as: wherein, is the pixel value distribution variance of the image after applying Laplacian operator to the remote sensing data, is the pixel value distribution variance of the original image of the remote sensing data.

[0012] Preferably, the credibility threshold is a numerical value determined according to statistical analysis of historical remote sensing data, which is used to identify low-quality data with large-area cloud and fog shielding or serious sensor noise.

[0013] Preferably, in step b, the differential utility fingerprint further includes a fourth utility component, and the generation method of the fourth utility component includes: obtaining the spatial resolution of the to-be-evaluated remote sensing data and the effective reference remote sensing data; calculating the ratio of the spatial resolutions, and performing nonlinear mapping on the ratio through an arctangent function to generate a fourth utility component representing the degree of scale inconsistency between the two, as a confidence index of the first utility component.

[0014] Preferably, when step a is performed, the method further includes the following steps in parallel: recording the mismatch information between the metadata of the to-be-evaluated data and each requirement in the task template; and combining all the non-zero mismatch information to generate a structured diagnostic mismatch fingerprint representing the capability short board of the to-be-evaluated data under the specific task requirement.

[0015] Preferably, the method further includes: based on the content of the diagnostic mismatch fingerprint, invoking data search instructions to search other remote sensing data products in the data directory whose metadata match one or more mismatch information recorded in the diagnostic mismatch fingerprint, and generating a data combination suggestion containing information of the other remote sensing data products.

[0016] ​Preferably, the method further comprises a set of task template dynamic optimization mechanism, which comprises the following steps: recording the differential utility fingerprints of the completed remote sensing data associated with a specific task; periodically performing statistical analysis on a set of differential utility fingerprints associated with the same specific task, determining the consensus utility pattern reflecting market preferences by calculating the statistical median of each utility component dimension; and comparing the consensus utility pattern with the task template of the specific task, and generating a set of optimization data for optimizing the task template based on the comparison result.

[0017] Preferably, the task template is stored in JSON format, which allows the task template to be updated or supplemented through text editing.

[0018] Compared with the prior art, the present application has the following advantages:

[0019] 1. The present application generates a first utility component representing information increment, a second utility component representing task fit degree, and a third utility component representing signal stability for a remote sensing data through parallel processing, wherein the information increment is obtained based on the difference in information distribution between the data to be evaluated and the known reference, the task fit degree is determined by logical comparison between data metadata and a preset task template, and the signal stability is quantified from the internal ratio of structured information and unstructured information of the data itself; this method unifies three independent analysis results from different dimensions into a standardized differential utility fingerprint, so that the value judgment of data assets is no longer dependent on isolated interpretation of a single technical indicator or prediction of market price, but is transformed into an inspection of a utility profile composed of information novelty, scene matching degree and structural clarity, which can be traced and understood.

[0020] 2. The present application calls a signal processing step for evaluating the structural clarity of the data itself before evaluating the information novelty, processes the known data as a measurement reference, and determines whether the reference data is valid according to its own structural clarity component; this way of prepositioning and reusing the internal functional modules of the evaluation system to examine the reliability of the premise itself establishes an internal logic self-checking loop for the entire evaluation method, which avoids the distortion of evaluation results caused by the introduction of contaminated reference data, and enables the evaluation system to resist input source pollution and maintain the reliability of its own conclusions when facing uncertain data quality in the real world.

[0021] 3、The method of the present application also includes recording the differential utility fingerprints of the closed remote sensing data associated with specific tasks, and periodically performing statistical analysis on these fingerprints derived from real market behavior to determine a consensus utility pattern reflecting market preferences; it compares this consensus utility pattern with the preset task template, and provides the differences as optimization suggestions to the managers, this mechanism builds a feedback path from discrete market transaction results to structured technical rules, so that the original evaluation system based on static expert knowledge can absorb and reflect the dynamic changes of market cognition, thereby having the ability of self-improvement and continuous evolution.

[0022] 4、The present application also records the mismatch information between the metadata of the data to be evaluated and each requirement in the task template in parallel when performing scene matching degree evaluation, and combines these information into a diagnostic mismatch fingerprint representing the short board of the ability; this converts the intermediate information discarded in the matching process into a new information product structured and usable for subsequent business logic, which not only reveals why it does not match, but also indicates the direction of ability complement, so that the evaluation method is no longer limited to obtaining a single matching score, but is extended to a function capable of providing data combination suggestions. BRIEF DESCRIPTION OF DRAWINGS

[0023] Fig. 1 Process interaction diagram for generating diagnostic mismatch fingerprints and data combination suggestions for the present application;

[0024] Fig. 2 Correlation diagram for the overall data flow and mechanism of the present application;

[0025] Fig. 3 Use case diagram for the remote sensing data asset value evaluation system of the present application. DETAILED DESCRIPTION

[0026] In order to make the technical solutions and advantages of the present application more clear, the present application will be further described in detail below, but it should be understood that the following embodiments are intended to explain and illustrate, but not to limit the protection scope of the present application.

[0027] The method for remote sensing data asset value evaluation and pricing by using AI provided by the application establishes a set of computer data processing procedures aiming to abandon unreliable price prediction and having inherent logic self-checking and utility transparency characteristics. The core of the procedure is to generate a standardized differential utility fingerprint. The fingerprint is formed by combining three utility components representing information novelty, task scene matching degree and data structure clarity. In a specific application scenario, for example, when a financial institution needs to supervise the asset changes in a region, the existing technical methods often cause the purchased data to be inconsistent with the actual task requirements due to the non-transparent evaluation process or excessive dependence on a single indicator, thereby affecting the reliability of the decision. To address the utility evaluation dilemma existing in similar business management and supervision scenarios, the method of the application is configured to perform the following steps. The method first performs step a, that is, generating a differential utility fingerprint representing the decision utility of remote sensing data in a specific task. Specifically, when processing a remote sensing data to be evaluated, in order to objectively measure the information increment contained in the remote sensing data, the system needs to introduce a known benchmark remote sensing data as a reference. Considering that the known data may have quality uncertainty such as cloud cover or sensor noise in the real world, in order to avoid distortion of the evaluation result caused by the introduction of contaminated benchmark, the system performs a benchmark data credibility gating review. Before calculating the information novelty, the procedure first calls a signal processing method for evaluating the data itself clarity, processes the known benchmark remote sensing data to obtain its own structure clarity component. The calculation of the structure clarity component of the remote sensing data is limited to wherein, is the pixel value distribution variance of the image after applying the Laplacian operator to the remote sensing data, is the pixel value distribution variance of the original image of the remote sensing data. Subsequently, the system compares the calculated structure clarity component of the benchmark data with a credibility threshold, which is a numerical value determined in advance according to historical remote sensing data statistical analysis to identify low-quality data. If the component value fails to meet the threshold, it indicates that the benchmark data itself has quality problems. The system will select another known benchmark remote sensing data and repeat the review until an effective benchmark data meeting the credibility threshold is found.

[0028] ​After determining an effective benchmark data, the system obtains a first utility component representing the information novelty of the to-be-evaluated data by calculating the difference between the to-be-evaluated data and the effective benchmark data in the information quantity distribution. To convert the concept of information quantity distribution difference into a calculable engineering implementation, the obtaining manner of this step is limited as follows: the gray level histograms of the predetermined wave bands of the to-be-evaluated data and the effective benchmark data are respectively calculated, and the K-L divergence is used to measure the difference between the two gray level histograms. As an information theory measure, the K-L divergence can effectively measure the amount of information lost when one probability distribution is used to approximate another. The calculation result numerically reflects the deviation degree of the information distribution of the to-be-evaluated data from the public information benchmark, thereby representing the information novelty contained therein. Further, to judge the applicability of the to-be-evaluated data in a specific commercial task, the system performs logical matching between the metadata of the to-be-evaluated data and a task template defining the requirements of a specific task, thereby obtaining a second utility component representing the scene matching degree of the to-be-evaluated data. To ensure the transparency and scalability of this matching process, the task template is stored in the JSON format, allowing the template to be updated or supplemented through text editing. The template defines the key wave bands, the optimal resolution range and the imaging time information required by a specific task. The matching process is realized by logically comparing and weightedly summing the wave band information, spatial resolution and imaging time information contained in the metadata of the to-be-evaluated data with the requirements in the task template. This design converts the judgment of task matching degree into a series of traceable logical operations and arithmetic summations. Meanwhile, to quantify the signal quality of the data itself, the system applies a signal processing method to the to-be-evaluated data to obtain a third utility component representing the self-structural clarity of the to-be-evaluated data by calculating the ratio of the internal structured information and unstructured information. The specific implementation of this process is to apply a Laplacian operator to filter the to-be-evaluated data and calculate the variance ratio of the filtered image and the original image. The technical logic is that the Laplacian operator, as a second-order differential operator, is sensitive to high-frequency noise signals in the image and weakly responds to smooth structural information. Therefore, by calculating the ratio, the proportion of the structural part carrying effective information in the image can be quantified, thereby providing an objective technical index for the stability and maneuverability of the data itself. This ratio can quantify the proportion of the structural part carrying effective information in the image, thereby providing an objective technical index for the stability and maneuverability of the data itself.

[0029] After the above three utility components are independently calculated, the method performs step b to combine the first utility component, the second utility component and the third utility component to form a differential utility fingerprint of the remote sensing data to be evaluated, which, in the absence of conflicts, can also be configured to include a fourth utility component, the generation method of which includes: obtaining the spatial resolution of the remote sensing data to be evaluated and the effective reference remote sensing data; calculating the ratio of the spatial resolutions of the two, and performing a non-linear mapping of the ratio through an arctangent function to generate a fourth utility component representing the degree of scale mismatch between the two, which serves as a confidence indicator for the first utility component; accordingly, for the conversion of intermediate information generated during the evaluation process into data products with commercial value, the method also records the mismatch information between the metadata of the data to be evaluated and each requirement in the task template in parallel when performing step a, and combines all non-zero mismatch information to generate a structured diagnostic mismatch fingerprint representing the capability short board of the data to be evaluated under specific task requirements. This diagnostic mismatch fingerprint reveals the specific reasons for the data mismatch and indicates the direction of capability complement, which can be further used to invoke data search instructions to search for other remote sensing data products that can complement the capability short board in the data directory, and generate a data combination suggestion containing information of other remote sensing data products.

[0030] Finally, the method performs step c to output the differential utility fingerprint as a standardized technical data structure to support the value evaluation and pricing decision of the remote sensing data asset. To enhance its usability in commercial scenarios, this step can also include visualizing the differential utility fingerprint on a data trading platform, such as presenting the first utility component, the second utility component and the third utility component in the form of a radar chart as an objective technical reference. Furthermore, to enable the entire evaluation system to have the ability of self-evolution, the method can also include a task template dynamic optimization mechanism that records the differential utility fingerprints of the remote sensing data that have been traded in association with a specific task, and periodically performs statistical analysis on a set of fingerprints under the same task to determine a consensus utility pattern reflecting market preferences by calculating the statistical median of each utility component dimension. The system compares this consensus utility pattern with the preset task template and generates a set of optimization data for optimizing the task template based on the comparison results. In this way, a feedback path from market transaction results to technical rules is established, enabling the evaluation system to absorb and reflect the dynamically changing market cognition.

[0031] In a commercial decision scenario, an agricultural supply chain finance company needs to assess the future output of multiple plantations it has granted credit to in a region after a regional drought, in order to adjust its financial derivative positions. At this time, there are multiple types of remote sensing data available on the data market: one is an optical image with high price and 0.5-meter spatial resolution; another is a multispectral image with moderate price and 10-meter spatial resolution, but containing a red edge band; and a publicly available benchmark remote sensing data with the same 10-meter resolution. Conventional data procurement decisions often focus on spatial resolution alone, which may lead to the purchase of high-priced optical images that cannot reflect the internal physiological stress of crops before wilting, resulting in underestimation of risks. When the method of the present application is applied to this scenario, the system does not directly predict the prices or grades of the three types of data, but generates a differential utility fingerprint for each type of data to be evaluated. When evaluating the multispectral image with moderate price, the system obtains the data and a publicly available benchmark remote sensing data, and first performs a benchmark data reliability gate review on the benchmark data, i.e., calls the structural clarity component calculation method to calculate the value of the benchmark data itself. After confirming that the value meets the preset reliability threshold, indicating that the data quality of the benchmark data can be used for benchmark comparison, the system continues to perform subsequent steps. This collaborative mechanism uses the third utility component calculation method for evaluating the stability of the data to be evaluated to preliminarily review the reliability of the evaluation premise, so that the acquisition of the first utility component of information novelty is based on a reliable benchmark.

[0032] ​In the calculation of the first utility component, the system compares the gray histogram of the multispectral image with the effective benchmark data in the red edge band, and uses K-L divergence to measure; because the change of chlorophyll content of crops under drought stress will be reflected in the red edge band, resulting in deviation of the information distribution from the known benchmark data under normal conditions, the system therefore calculates a higher information novelty component; when evaluating the scene matching degree, the red edge band is listed as a necessary band requirement in the JSON format task template for this drought stress monitoring task, and the optimal resolution range requirement is set between 5 meters and 15 meters, so that the metadata of the multispectral image obtains a higher second utility component after logical comparison and weighted summation with the task template; on the contrary, although the 0.5 meter resolution optical image far exceeds the requirement in resolution, its second utility component is lower because it does not contain the red edge band, this method resolves the contradiction between technical superiority and actual utility in data procurement; in the evaluation process, the system also records the mismatch information between the metadata of the data to be evaluated and the task template in parallel; for example, if the task template contains a necessary time requirement within 72 hours after the disaster, and the imaging time of the multispectral image exceeds this range, the system will record this mismatch information and generate a separate diagnostic mismatch fingerprint; this converts the intermediate information discarded in the matching process into a structured information product that can be used for subsequent business logic, which not only reveals the short board of the data in timeliness to the decision maker, but also drives the generation of data combination suggestions, such as calling data search instructions to find other data products that can make up for the timeliness short board; finally, the data analysis department of the financial company no longer faces a group of isolated and difficult to compare prices or technical parameters, but receives a differential utility fingerprint of the multispectral image composed of three utility components of high information novelty, high scene matching degree and high structure clarity, and a diagnostic mismatch fingerprint indicating the short board in timeliness, the basis of decision-making changes from the prediction of a single price to the review of a utility profile composed of information novelty, scene matching degree and structure clarity, which can be traced and understood, making the risk management and procurement decision of data assets have logical self-checking ability and transparent technical basis.

[0033] Example 2: To objectively verify the effectiveness of the method of the present invention in distinguishing the utility of different remote sensing data for a specific task compared with conventional methods, this experiment was designed. The purpose of the experiment is to quantitatively compare the differential utility fingerprint evaluation system used in the present invention with a conventional evaluation method that relies on a single technical indicator, and to determine the degree of conformity between the evaluation results and the preset true utility levels when faced with a complex dataset containing high-quality, low-quality, and pseudo-high-value data. To this end, a computer-generated remote sensing data simulation test platform was constructed. This platform includes a test dataset containing 10 remote sensing data with different characteristics, and a software environment for performing data utility evaluation. In the test dataset used in the experiment, different true utility levels were preset for different data, which represent their application in the specific management task of monitoring the addition of illegal buildings in a monitored ecological protection zone. The application value of this method is evident. The dataset contains various types of interference data, such as high-resolution pseudo-color data with high spatial resolution but whose imaging timeframe does not meet the task template requirements, and high-resolution, low-quality data with large areas of cloud cover. The experiment sets up a control group and an experimental group. The control group uses an evaluation method with spatial resolution as the primary weighting indicator, while the experimental group uses the complete evaluation method based on differential utility fingerprinting, as described in this invention. For the baseline data reliability gating review stage in the experimental group method, the reliability threshold is set to 0.25. This parameter aims to balance the rigor and usability of the review; values ​​below this typically indicate noise or occlusion in the data that may affect the evaluation. This value is based on statistical analysis of historical public datasets and is an engineering example determined for image data that can effectively filter out cloud cover exceeding 30%.

[0034] After the experiment began, the control group and the experimental group each processed all 10 remote sensing data points in the test dataset using their respective evaluation methods, and each generated a utility ranking. The results showed that the experimental group's ranking was highly correlated with the actual utility level, while the control group's ranking showed several discrepancies with the actual utility level, resulting in a utility level of 5. The experimental group's method, based on its scene matching component, assigned a ranking of 5 to D03, consistent with the actual situation. For data point D07 (high score, low quality - cloud), its actual utility level was 8, while the control group assigned a ranking of 6. The experimental group's method, based on its structural clarity component... The calculation shows that the data quality does not meet the requirements, and the ranking 8 consistent with the true utility level is given, in addition, the test group correctly identifies the data numbered D02 (medium multispectral - high quality) as the highest utility ranking, while the control group ranks it in the third place; the difference in the ranking results is due to the differential utility fingerprint mechanism of the application, which avoids the one-sidedness of a single technical index through the collaborative evaluation of the three independent dimensions of information novelty, scene matching degree and structural clarity, and is assisted by the benchmark data reliability gate review, can more comprehensively reflect the decision utility of the data under the specific task, the test results confirm that, compared with the conventional method relying on a single technical index, the method of the application can identify and distinguish the internal quality and task applicability of different remote sensing data, and the evaluation results are highly consistent with the application value of the data, which provides a data processing procedure for the value evaluation of remote sensing data assets in the fields of administrative management and supervision.

[0035] Embodiment 3: This embodiment combines Figs. 1 to 3 to realize and illustrate a method for evaluating and pricing remote sensing data assets by using AI, as shown in Fig. 1 The diagram starts with the user submitting the data to be evaluated to the evaluation system, and then the evaluation system starts the scene matching degree evaluation, and then the scene matcher reads the task template requirements, and then compares the band information, spatial resolution and imaging time item by item, in the process, the mismatch recorder records the mismatch information synchronously, and combines the information to generate a diagnostic mismatch fingerprint after the comparison is completed; when the system determines that there is a capability gap, an alternative process is triggered, that is, a data search instruction is sent to the data search engine, which searches for complementary data in the data product directory to make up for the gap, and generates a data combination suggestion according to the returned matching product list, finally, the evaluation system outputs the diagnostic mismatch fingerprint and the combination suggestion to the user.

[0036] As shown in Fig. 2 , the diagram shows that the data to be evaluated, the benchmark data and the JSON format task template are the core inputs, which enter the differential utility fingerprint evaluation core module together, the module performs comprehensive evaluation on information novelty, scene matching degree and structural clarity, on the one hand, generates a differential utility fingerprint as a standardized objective utility profile to support value evaluation and pricing decisions, on the other hand, drives the generation of a diagnostic mismatch fingerprint, that is, records the mismatch information between the metadata and the task template to reveal the capability gap, so as to form a structured capability gap information, and a data combination suggestion can be generated accordingly; at the same time, the diagram also reveals a set of task template dynamic optimization mechanism, which analyzes the differential utility fingerprint of the completed remote sensing data to calculate the utility mode reflecting the market consensus, and then makes optimization suggestions for the task template, realizing the self-evolution of the system.

[0037] As shown in Fig. 3As shown, the figure shows the use case relationship of the remote sensing data asset value evaluation system of the present application from the perspective of system participants, among which the data analyst or decision maker is the core user, interacts with the system to perform the two main functions of evaluating data asset value and obtaining diagnosis and enhancement suggestions; and the system administrator is responsible for performing the function of managing and optimizing the task template to ensure the accuracy and timeliness of the evaluation system. In addition, the figure also shows the data providing source as the source of system evaluation object.

[0038] Example 4: When the method of the present application is applied to a new commercial task, i.e. financial risk pre-monitoring of suspected illegal felling activities in a specific tropical rainforest area, a calibrated task template and evaluation procedure need to be constructed for this task. This embodiment describes the systematic calibration process for completing the configuration of this task. Before starting the calibration process, a reference data set for the illegal felling task in the tropical rainforest is first obtained, which contains 20 to 30 positive sample images annotated by analysts, which clearly show the recent felling patches. The data set also contains a comparable number of negative sample images, which are seasonal leaf fall or non-target changes such as farmland rotation; The first step of the process is to calibrate the internal weights of the scene matching degree component in the task template. The system adjusts the relative weights of the three requirements of key band, optimal resolution range and necessary phase in an iterative manner, and uses these weights to calculate the second utility component of all samples in the reference data set. The objective function of this iterative process is to find a combination of weights that maximizes the mean of all positive sample scores while minimizing the mean of all negative sample scores, and solidifies this weight combination into the JSON format task template specific to this task.

[0039] The second step of the process is to calibrate the trust threshold used in the reference data trust review. Due to the differences in atmospheric and noise characteristics of images from different geographical regions and commonly used sensors, a matching threshold needs to be set. The calibration procedure is to randomly select 100 images from the historical public data archives associated with the target region, and divide them into three categories of clear, light cloud or haze pollution and heavy cloud or haze pollution by analysts or auxiliary tools; Then, the system calculates the structural clarity component of each of the 100 images and analyzes the score distribution of the three categories of images. The value of the trust threshold is determined as the score of the clear category The 10th percentile of the score distribution, which is designed to ensure that the threshold has discrimination ability while avoiding excluding most available data due to overly strict standards; The third step of the process is to refine the calculation method of the fourth utility component, which represents the degree of scale inconsistency between the data to be evaluated and the reference data. The ratio of the spatial resolution of the two is nonlinearly mapped through an arctangent function to generate a scale inconsistency factor with a value range of [0, 1] The calculation formula is determined as Wherein is the ratio of the larger value to the smaller value of the spatial resolution of the effective reference data and the data to be evaluated, that is This function maps the value of to 0 when , and as the value of increases, the value of nonlinearly approaches 1; By performing the above calibration process, the system finally generates a complete evaluation procedure for the illegal deforestation of tropical rainforests, including accurate internal weights, localized credibility thresholds, and deterministic calculation formulas. This process converts the setting of all key parameters in the evaluation method from reliance on experience to a set of traceable and reproducible engineering calibration steps, providing a standardized deployment path for the application of the invention to new commercial tasks.

[0040] Embodiment 5: Task template dynamic optimization mechanism of the method of the invention, in a continuously running commercial environment, its operation is as follows: when the system is monitoring the drought stress task, after recording a statistically significant number of differential utility fingerprints of the traded remote sensing data within a predetermined period, an optimization analysis procedure is triggered, which calculates the statistical median of the consensus utility pattern in the information novelty, scene matching degree and structure clarity three dimensions of this set of fingerprint collection, to form a consensus utility pattern reflecting market preferences; Then, the system compares the consensus utility pattern with the current artificial preset task template for this task, if there is a persistent deviation in any dimension that exceeds the preset statistical threshold, a template optimization suggestion report is generated, which does not automatically modify the system parameters, but submits the data comparison results and optimization suggestions presented in the report to the designated system administrator for review, and makes a decision on whether to update the task template, all decision-making processes are recorded by the system log.

[0041] In the daily operation of the method, when a remote sensing data to be evaluated fails to fully meet the requirements of the specified task template due to its metadata, a diagnostic mismatch fingerprint representing its capability short board is generated. The system initiates a business opportunity exploration mechanism. If the diagnostic mismatch fingerprint indicates that the current data has a short board in meeting the spatial resolution requirement, the system automatically converts the mismatch information into a structured data search instruction, which searches for other remote sensing data products that can make up for the capability short board in the data product catalog, and generates a data combination suggestion containing the respective differential utility fingerprints of the original data and the recommended supplementary data, as well as an expected combination utility fingerprint calculated based on the combination of the two features, to provide the data user with a quantifiable and procurement decision-making basis for the combination scheme.

[0042] In the case of a boundary condition where there is no effective public reference remote sensing data, the system is configured to execute a preset fault-tolerant procedure. When the system cannot obtain effective reference data that meets the correlation requirements in time and space and passes the reference data reliability gate review in all configured public data archives after a preset number of consecutive searches for a piece of data to be evaluated, the system determines that it is in a reference unavailable state. In this state, to ensure the continuity of the evaluation process, the system automatically sets the first utility component of the data to be evaluated, i.e., information novelty, to a preset neutral non-zero value of 0.5, and adds a status identifier indicating that the information novelty is not effectively calculated in the differential utility fingerprint generated in this evaluation.

[0043] To determine the internal weight of the scene matching degree component in the task template so that the utility evaluation has reproducible quantitative basis, the system is configured to execute an offline optimization parameter search process. The objective function of the process is defined as finding a combination of weights that can maximize the mean value of the scene matching degree scores calculated from a set of positive sample reference data, and minimize the mean value of the scores calculated from a set of negative sample reference data. To solve the objective function, the system uses a coordinate ascent iterative optimization algorithm. The initial state of the algorithm is to set the internal weights of the key band requirement, the optimal resolution range requirement, and the necessary time requirement under the scene matching degree component to the same initial value. In each iteration, the algorithm fixes all other weights except one, and traverses the weight in the [0, 1] interval with a preset step size to find the optimal weight value that optimizes the objective function value. This process is applied to all weights in turn until a fixed number of iterations is completed, or the gain of the objective function value in two consecutive iterations is less than a preset threshold, and the final weight combination is output.

[0044] It is apparent for a person skilled in the art that the application is not limited to the details of the above-described exemplary embodiments, but that the application can be implemented in other concrete forms without departing from the spirit or essential characteristics of the application.

[0045] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application but not to limit the present application, and although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present application.

Claims

1. A method for remote sensing data asset value assessment and pricing using AI, characterized in that, The method establishes a set of computer data processing procedures, which comprises the following steps: Step a, generating a differential utility fingerprint representing the decision-making utility of remote sensing data under a specific task, which specifically comprises: after obtaining the remote sensing data to be evaluated and a known reference remote sensing data, performing a reference data reliability gating review: applying a signal processing method to the known reference remote sensing data to obtain its own structural clarity component, and judging whether the structural clarity component meets the reliability threshold; if the judgment result is not satisfied, selecting another known reference remote sensing data and repeating the review until valid reference data meeting the reliability threshold is found; after the review determines the valid reference data, a first utility component representing the information novelty of the to-be-evaluated data is obtained by calculating the difference between the to-be-evaluated data and the valid reference data in the information quantity distribution, and the first utility component is obtained by calculating the gray level histogram of each predetermined band of the to-be-evaluated data and the valid reference data respectively, and using K-L divergence to measure the difference between the two gray level histograms; based on a task template defining the requirements of a specific task, the metadata of the to-be-evaluated data is logically matched with the task template to obtain a second utility component representing the scene matching degree of the to-be-evaluated data, and the second utility component is obtained by logically comparing and weighted summing the band information, spatial resolution and imaging time information contained in the metadata of the to-be-evaluated data with the key band requirement, optimal resolution range requirement and timeliness requirement defined for the specific task in the task template; a signal processing method is applied to the to-be-evaluated data to calculate the ratio of its internal structured information and unstructured information to obtain a third utility component representing the structural clarity of the to-be-evaluated data, and the third utility component is obtained by applying a Laplacian operator to filter the to-be-evaluated data and calculating the variance ratio of the filtered image and the original image; Step b, combining the first utility component, the second utility component and the third utility component to form a differential utility fingerprint of the to-be-evaluated remote sensing data; Step c, outputting the differential utility fingerprint as a standardized technical data structure to support the value evaluation and pricing decision of the remote sensing data asset.

2. The method for remote sensing data asset value assessment and pricing using AI according to claim 1, wherein, In step a, the structural clarity component is calculated as: wherein the calculation of the structural clarity component is defined as: wherein, is the variance of the pixel value distribution of the image after applying the Laplacian operator to the remote sensing data, is the variance of the pixel value distribution of the original image of the remote sensing data.

3. The method for remote sensing data asset value assessment and pricing using AI according to claim 1, wherein, The reliability threshold is a value determined according to statistical analysis of historical remote sensing data to identify low-quality data with large-area cloud cover or severe sensor noise.

4. The method for remote sensing data asset value assessment and pricing using AI according to claim 1, wherein, In step b, the differential utility fingerprint further comprises a fourth utility component, and the generation method of the fourth utility component comprises: obtaining the spatial resolutions of the to-be-evaluated remote sensing data and the valid reference remote sensing data; calculating the ratio of the spatial resolutions, and nonlinearly mapping the ratio through an arctangent function to generate a fourth utility component representing the degree of scale mismatch between the two, as a confidence indicator for the first utility component.

5. The method for remote sensing data asset value assessment and pricing using AI according to claim 1, wherein, In performing step a, the method further comprises, in parallel, the steps of: recording mismatch information between the metadata of the data to be evaluated and each requirement in the task template; and combining all non-zero mismatch information to generate a diagnostic mismatch fingerprint that characterizes the capability shortfalls of the data to be evaluated under the specific task requirements.

6. The method for remote sensing data asset value assessment and pricing using AI according to claim 5, wherein, The method further comprises the steps of: based on the content of the diagnostic mismatch fingerprint, invoking a data search instruction to search for other remote sensing data products in the data catalog whose metadata match one or more pieces of mismatch information recorded in the diagnostic mismatch fingerprint, and generating a data combination suggestion containing information of the other remote sensing data products.

7. The method for remote sensing data asset value assessment and pricing using AI according to claim 1, wherein, The method further comprises a set of task template dynamic optimization mechanisms, which include the following steps: recording the differential utility fingerprints of the remote sensing data associated with a specific task; periodically performing statistical analysis on a set of recorded differential utility fingerprints associated with the same specific task, determining a consensus utility pattern reflecting market preferences by calculating the statistical median of each utility component dimension; and comparing the consensus utility pattern with the task template of the specific task, and based on the comparison result, generating a set of optimization data for optimizing the task template.

8. The method for remote sensing data asset value assessment and pricing using AI according to claim 1, wherein, The task template is stored in JSON format, which allows the task template to be updated or supplemented through text editing.

Images