Xiguang Series Hyperspectral Remote Sensing Satellite Constellation
Committed to building ChinaLargest and most feature-rich satellite constellation, easy to use and practical.Hyperspectral remote sensing constellation
Hyperspectral imaging technology
Space doctors performing "CT scans" of Earth
Ordinary satellite cameras can only show the surface color (such as a green forest), while hyperspectral cameras are like CT scanners in hospitals. They can "slice" objects for analysis, not only showing the color of the forest, but also determining whether the trees are sick, lack water, or even what minerals are hidden underground by analyzing hundreds of subtle color differences in the sunlight reflected by the leaves.
Compared to imagery primarily used for visualization (RGB) and semi-quantitative classification/monitoring (multispectral), hyperspectral imagery, with its tens of times greater data information (hundreds of continuous narrow bands vs. a few discrete wide bands), achieves incredibly detailed capture of the spectral characteristics of ground features. This leap in information density directly translates into a revolutionary improvement in analytical precision, giving it irreplaceable core value in fields requiring precise material identification and quantitative analysis, such as precision agriculture, mineral exploration, environmental monitoring, and military target identification.
Hyperspectral imaging has the following significant advantages
◎ It has nearly continuous spectral information of ground features.After spectral reflectance reconstruction, hyperspectral images can obtain a continuous spectral reflectance curve that approximates the object being detected, matching its measured value, thus applying the spectral analysis model of the object being detected in the laboratory to the imaging process.
◎The ability to detect and identify land cover has been greatly improved.Hyperspectral data can detect substances with diagnostic spectral absorption characteristics and accurately distinguish ground cover types, road surface materials, etc.
There are various methods for classifying and identifying geographic features.Image classification can employ pattern recognition methods such as Bayesian discriminant analysis, decision trees, neural networks, and support vector machines, or it can use methods based on spectral matching from a spectral database of the object being examined. Classification and recognition features can be derived from either spectral diagnostic features or feature selection and extraction.
◎ Quantitative and semi-quantitative classification and identification of ground features will become possible.Hyperspectral images can estimate the state parameters of various objects being detected, greatly improving the accuracy and reliability of high-quantitative analysis of imaging.
Hyperspectral satellite remote sensing data possesses the characteristic of "integrated image and spectrum," enabling it to acquire more information about ground features than ordinary satellites, and can be widely applied to various uses, including military, government, and civilian sectors. In marine environment, ecological environment, and atmospheric monitoring, hyperspectral data can achieve more precise monitoring results than multispectral data.
-
Full-spectrum hyperspectral observation -
Wide amplitude, wide spectrum -
High spatial resolution, high spectral resolution
High signal-to-noise ratio and high calibration accuracy -
Covering time, space,
Spectrum, radiation and polarization -
Conduct observations from multiple angles.