CN-121980307-A - Ecological environment monitoring device based on remote sensing image

CN121980307ACN 121980307 ACN121980307 ACN 121980307ACN-121980307-A

Abstract

The invention relates to the technical field of ecological environment monitoring, in particular to an ecological environment monitoring device based on remote sensing images, which comprises a data acquisition and preprocessing module, a land coverage identification module, a reference area selection module, a weather-spectrum mapping model construction module, a spectrum anomaly detection and area correction module, a new target area and an ecological parameter inversion and output module, wherein the data acquisition and preprocessing module is used for acquiring current and historical remote sensing images and weather data of a target area and constructing a time sequence data set, the land coverage identification module is used for identifying the land coverage type of the target area, the reference area selection module is used for defining a reference area in a preset range, the weather-spectrum mapping model construction module is used for establishing a mapping function from the weather data to spectrum reflection characteristics, the spectrum anomaly detection and area correction module is used for comparing the difference between an actual spectrum and an expected spectrum output by a model, and conducting boundary cutting correction on the anomaly area to obtain a new target area after weather influence correction, and the ecological parameter inversion and mapping. The invention can improve the adaptability to different climatic conditions and enhance the accuracy of the monitoring result.

Inventors

WANG CHAO
YUAN CHUNXIA
HU ZHUO
Xu Baicui
FENG YAYA

Assignees

兰州城市学院

Dates

Publication Date: 20260505
Application Date: 20260116

Claims (10)

1. An ecological environment monitoring device based on remote sensing image, characterized by comprising: The data acquisition and preprocessing module is used for acquiring and preprocessing the current remote sensing image and the current meteorological data of the target area, synchronously acquiring the historical remote sensing image and the corresponding historical meteorological data, and constructing a time sequence data set; The land coverage identification module is used for identifying the land coverage type of the target area by adopting a classification algorithm based on the spectral reflection characteristic; The reference area selection module is used for selecting a fixed reference object from the historical remote sensing image, defining a reference area of a preset range by taking the reference object as a center, serving as a reference area for weather-spectrum relation learning, and defaulting the area land coverage type to be unchanged in time sequence; The weather-spectrum mapping model construction module is used for learning dynamic association between weather data and spectral reflection characteristics in a reference area by using a deep learning model and establishing a mapping function from the weather data to the spectral reflection characteristics; The spectrum anomaly detection and region correction module is used for inputting the current remote sensing image and the current meteorological data into a trained meteorological-spectrum mapping model, predicting an expected spectrum of a reference region under the current meteorological condition, comparing an actual spectrum of a target region in the current remote sensing image with the expected spectrum output by the model, judging that the spectrum-meteorological relation is not a corresponding region if the difference exceeds a set threshold, and carrying out boundary cutting correction on the region to obtain a new target region after meteorological influence correction; and the ecological parameter inversion and output module is used for carrying out inversion and drawing of ecological parameters based on the corrected new target area.
2. The remote sensing image based ecological environment monitoring device of claim 1, wherein the current weather data and the historical weather data include at least precipitation data, air temperature data, air humidity data, and cloud cover data.
3. The remote sensing image based ecological environment monitoring device of claim 2, wherein the ecological parameters include NDVI, leaf area index, surface temperature, and vegetation coverage.
4. The device of claim 3, wherein the fixed reference is a natural or artificial ground object with stable spectrum and spatial morphology in the remote sensing images of a long time sequence.
5. The remote sensing image based ecological environment monitoring device according to claim 4, further comprising a dynamic baseline construction module for establishing a dynamic ecological baseline library for different geographic units according to seasonal changes based on long time sequence historical data.
6. The remote sensing image-based ecological environment monitoring device according to claim 5, further comprising an intelligent early warning module, wherein the intelligent early warning module is used for comparing real-time ecological parameters with dynamic baselines of corresponding space-time positions, and carrying out grading evaluation and early warning on ecological anomalies by combining parameter change trend analysis.
7. The remote sensing image-based ecological environment monitoring device of claim 6, wherein the intelligent pre-warning module performs state comparison based on the following logic: Obtaining a parameter value P obtained by inversion of the current space-time position; a normal value range which is extracted from the dynamic ecological base line library and corresponds to the current pixel geographic position and the current time period; Judging whether the parameter value P falls within the normal value range section: if P is in the range of the normal value, marking that the state is normal, and not entering the subsequent early warning process; if P And marking a normal value range section as abnormal state, and calculating the deviation degree.
8. The remote sensing image based ecological environment monitoring device of claim 7, wherein the intelligent pre-warning module performs trend analysis based on the following logic: inputting a short time sequence parameter sequence, namely acquiring a time sequence parameter value corresponding to the abnormal pixel in the last N days, wherein a current parameter value P is used as an end point of the sequence; trend fitting, namely carrying out linear regression on the short-time sequence data, and calculating a trend slope S; and constructing a decision matrix, namely constructing the decision matrix by taking the deviation degree as a horizontal axis and taking the trend slope S as a vertical axis, and predefining the early warning level and the urgency of each cell of the matrix.
9. The remote sensing image based ecological environment monitoring device of claim 8, further comprising a model adaptive update module for continuously collecting remote sensing images and synchronized weather data during long-term monitoring, and triggering incremental learning or retraining of the weather-spectrum mapping model periodically or when it is detected that the weather-spectrum mapping model prediction error continuously exceeds a preset range.
10. The remote sensing image based ecological environment monitoring device of claim 9, wherein the model adaptive update module determines whether the weather-spectral mapping model prediction error continues to exceed a preset range based on logic: in the reference area, comparing an expected spectrum predicted by the model according to the current meteorological data with an actual spectrum extracted from the current remote sensing image; Calculating root mean square error between the two; Storing the root mean square error obtained by each processing together with the time stamp thereof to form a continuous error time sequence; And carrying out statistical test on the error time sequence, detecting that the error has a statistically rising trend, and judging that the performance is continuously degraded if the current error value is at a historical preset level.

Description

Ecological environment monitoring device based on remote sensing image Technical Field The invention relates to the technical field of ecological environment monitoring, in particular to an ecological environment monitoring device based on remote sensing images. Background Along with the increasing prominence of global ecological environment problems, dynamic and accurate monitoring of a large-scale surface ecological system is realized, and the dynamic and accurate monitoring becomes an important foundation for environment management, ecological protection and sustainable development. Traditional ground investigation methods are limited by manpower, material resources and space-time coverage, and are difficult to meet the monitoring requirements of large-scale, high-frequency and continuous monitoring. Remote sensing technology has become an indispensable means for ecological environment monitoring by virtue of its macroscopic, rapid, objective and periodic observation. At present, monitoring based on remote sensing images mainly depends on multispectral, hyperspectral, radar and other image data acquired by satellites or aviation platforms, and key ecological parameters such as vegetation coverage, land utilization, water quality, surface temperature and the like are estimated through interpretation analysis. The nature of remote sensing monitoring relies on the interaction of surface targets with electromagnetic waves (primarily from solar radiation or active emissions from sensors). Different ground features (such as vegetation, water, soil, artificial buildings and the like) show unique absorption, reflection and emission characteristics to incident electromagnetic waves due to the differences of material compositions, structures, moisture contents and the like, so that identifiable spectral fingerprints are formed. The sensor forms pixel brightness values of the remote sensing images by recording the differentiated radiation signals, so that various ecological parameters are inverted. However, the process of remotely sensing data acquisition is highly dependent on atmospheric conditions and surface transients. Climate change and its derived fluctuations in meteorological conditions introduce significant uncertainty and interference from the data source. For example, the surface wetting after precipitation can obviously change the spectral reflectance characteristics of vegetation leaves and soil, namely the reflectance of vegetation in a near infrared band is reduced due to moisture, the vegetation is easily misjudged to be reduced in growth condition or biomass reduction, and the reflectance of wetted soil is obviously lower than that of dry soil, so that the surface classification and parameter inversion accuracy are affected. The existing monitoring method is mostly dependent on remote sensing images of specific time phases or uses simple vegetation indexes (such as NDVI), and lacks an adaptive modeling and compensation mechanism for seasonal changes and annual fluctuations. The monitoring result is easy to fluctuate between vegetation growing season and dry season and arid year and humid year, long-term ecological trend and short-term climate noise are difficult to effectively separate, comparability of data on time sequence and decision support value are reduced, and continuous monitoring capability under different meteorological conditions and precision and dimension of ecological parameter inversion are restricted. In view of the defects existing in the prior art, a novel and intelligent ecological environment monitoring device based on remote sensing images is needed to improve the adaptability to different climatic conditions and enhance the accuracy and reliability of monitoring results. Disclosure of Invention In order to solve the problems, the invention provides an ecological environment monitoring device based on remote sensing images, which is used for improving the accuracy of monitoring results. In order to achieve the purpose, the technical scheme of the invention is as follows, an ecological environment monitoring device based on remote sensing images comprises: The data acquisition and preprocessing module is used for acquiring and preprocessing the current remote sensing image and the current meteorological data of the target area, synchronously acquiring the historical remote sensing image and the corresponding historical meteorological data, and constructing a time sequence data set; The land coverage identification module is used for identifying the land coverage type of the target area by adopting a classification algorithm based on the spectral reflection characteristic; The reference area selection module is used for selecting a fixed reference object from the historical remote sensing image, defining a reference area of a preset range by taking the reference object as a center, serving as a reference area for weather-spectrum relation learning, and defaulting the area land