![\"Images](\"\/wp-content\/uploads\/2026\/08\/Images-produced-from-different-satellites-can-have-different-geometric-quality-and-may-fail-to-align.-Roads-buildings-and-other-features-can-appear-displaced-when-imagery-from-separate-spacecraft-is-overlaid.webp\")
<\/p>\nChina\u2019s Ministry of Natural Resources has released the country\u2019s first set of geometric reference imagery covering its entire land territory, establishing a common spatial reference for the production and geometric correction of satellite remote-sensing imagery.<\/p>\n
Announced in late August, the dataset has a spatial resolution better than 0.5 meters and is intended to address a growing problem in China\u2019s expanding Earth observation sector: imagery collected by different satellites can have inconsistent geometric positioning, preventing datasets from aligning accurately when overlaid.<\/p>\n
The ministry said the reference imagery will be made openly available to society. Beyond improving positioning accuracy, the common reference could reduce duplicated image-processing work and lower production costs for government agencies, commercial satellite operators and downstream geospatial companies.<\/p>\n
Satellite images are not immediately interchangeable simply because they cover the same location. Differences in spacecraft orbit determination, sensor geometry, attitude knowledge, terrain correction, imaging angles and processing chains can introduce geometric discrepancies between products.<\/p>\n
Zhou Ping, director of the Data Processing Department at the Land Satellite Remote Sensing Application Center of the Ministry of Natural Resources, said images produced from different satellites can have different geometric quality and may fail to align. Roads, buildings and other features can appear displaced when imagery from separate spacecraft is overlaid.<\/p>\n
<\/p>\n
Using the nationwide geometric reference imagery for calibration allows those datasets to be brought onto a consistent spatial framework. According to Zhou, measurements such as building areas and road lengths can consequently be represented much more accurately across imagery from different satellites.<\/p>\n
That capability becomes increasingly important as China\u2019s remote-sensing architecture expands beyond a relatively limited number of government spacecraft toward a mixed ecosystem of state and commercial constellations. More satellites increase imaging frequency and available coverage, but they also multiply the number of sensors, orbital configurations and processing pipelines from which geospatial products are generated.<\/p>\n
A common reference layer therefore functions as infrastructure for the downstream remote-sensing industry rather than as another conventional Earth observation product. It provides a shared geometric baseline against which imagery from different spacecraft can be corrected and subsequently combined.<\/p>\n
The Ministry of Natural Resources said development of the dataset addressed precision-control challenges in mountainous and plateau regions, where terrain can make geometric correction more difficult. The resulting reference has improved the overall positioning accuracy of remote-sensing imagery across China by more than a factor of two.<\/p>\n
Yu Dequan, deputy director of the Land Satellite Remote Sensing Application Center, described the dataset as the highest-overall-accuracy version of geometric reference imagery currently available in China and said it covers all of the country\u2019s land territory.<\/p>\n
The distinction between spatial resolution and geometric accuracy is important. A satellite may collect very high-resolution imagery while still exhibiting positional offsets relative to imagery from another sensor. Geometric calibration and orthorectification are therefore essential if data are to support precise mapping, change detection, infrastructure measurement or multi-source analysis.<\/p>\n
The new reference dataset is designed to provide that common positional foundation.<\/p>\n
For multi-temporal Earth observation in particular, improved registration can directly affect analytical quality. Detecting changes between images collected at different times requires the same buildings, roads, shorelines or terrain features to occupy consistent coordinates. Otherwise, geometric displacement can be misinterpreted by automated processing systems as physical change.<\/p>\n
The issue is becoming more consequential as remote-sensing companies increasingly apply artificial intelligence and automated image analytics to large datasets. Better geometric consistency can reduce one source of preprocessing error before imagery enters segmentation, object-detection, mapping or change-detection pipelines.<\/p>\n
The decision to make the reference imagery openly available may have broader commercial implications than the technical improvement alone.<\/p>\n
Without a common nationwide reference, satellite operators, imagery processors and application providers may need to build or acquire their own control datasets and perform additional geometric correction before combining imagery from different sources. Those activities consume computing resources, engineering time and geospatial reference data.<\/p>\n
A shared high-accuracy reference can move part of that work into common national infrastructure. Commercial providers can use the same baseline to calibrate raw satellite imagery rather than independently constructing equivalent reference layers.<\/p>\n
The potential benefits extend beyond individual image products. Consistent geometry can simplify the creation of mosaics assembled from multiple satellites and improve interoperability between imagery supplied by different operators. It can also make it easier for downstream customers to integrate new commercial satellite sources without rebuilding their entire spatial reference workflow.<\/p>\n
This is particularly relevant to high-frequency commercial constellations. Their value depends not only on how often satellites revisit a location but also on whether large numbers of images can be processed into standardized products efficiently. As constellation sizes increase, repeatable calibration and automated production become increasingly important to controlling the cost per delivered image.<\/p>\n
The nationwide reference imagery could therefore act as an enabling layer for scaling remote-sensing production, although individual operators will still need their own sensor calibration, orbit and attitude processing, radiometric correction and product-quality assurance.<\/p>\n
The Ministry of Natural Resources plans to promote use of the dataset in the Real-Scene 3D China program, natural-resource surveys and monitoring, and commercial space-based remote sensing.<\/p>\n
These applications place demanding requirements on geometric consistency.<\/p>\n
Large-scale 3D geographic datasets depend on imagery and other geospatial information being registered to a stable coordinate framework. Natural-resource monitoring similarly requires reliable comparison of observations acquired at different times, particularly when measuring changes in land use, vegetation, coastlines or infrastructure.<\/p>\n
Commercial Earth observation adds another dimension because customers increasingly expect imagery from multiple spacecraft to behave as standardized data products rather than isolated scenes tied to individual satellite systems.<\/p>\n
The reference layer could also support data fusion between optical satellite imagery and other geospatial datasets. In practical production systems, however, the quality of the final fused product will continue to depend on the characteristics and calibration of each sensor and on the processing methods used.<\/p>\n
The release is expected to be followed by a policy framework governing how the reference imagery is used.<\/p>\n
Sun Hua, an official with the Remote Sensing Information Division of the Ministry of Natural Resources\u2019 Department of Land Surveying and Mapping, said the ministry will study and formulate documents in 2026 to standardize application of the geometric reference imagery and ensure that the dataset is used efficiently while remaining secure and controllable.<\/p>\n
The ministry also plans to work with other government departments on measures intended to encourage adoption, including by commercial satellite companies.<\/p>\n
According to Sun, the objective is not only to improve utilization of the reference dataset but also to support new business models and a broader satellite remote-sensing industrial ecosystem.<\/p>\n
That policy dimension will determine part of the dataset\u2019s eventual commercial impact. Open access can reduce the cost of obtaining a common geometric reference, but widespread benefits depend on how consistently operators adopt it, how derived products are standardized and what requirements apply to different categories of geospatial data.<\/p>\n
For China\u2019s rapidly expanding commercial Earth observation sector, the release represents a shift toward shared infrastructure at the image-production layer. Increasing the number and resolution of satellites expands the supply of raw observations; establishing a common geometric foundation can make those observations easier to combine, process and turn into interoperable products at scale.<\/p>\n
China\u2019s expanding commercial space sector is bringing more satellite manufacturing capacity, Earth observation constellations and payload options to the global market. For international customers, that translates into a broader and increasingly competitive supply of satellite imagery, remote-sensing payloads and related space capabilities. STARPATH GLOBAL connects customers with this growing Chinese supply base, helping international projects identify competitive solutions beyond the traditional satellite and geospatial supply chain.<\/p>\n
For satellite imagery, higher resolution is not always the most economical choice. A project that can meet its technical objectives with 1-meter or 3-meter imagery may gain little from purchasing substantially more expensive sub-meter data. STARPATH GLOBAL helps customers match resolution, sensor type and imagery products to specific industry requirements, whether for agriculture, infrastructure monitoring, environmental assessment or change detection. The goal is to deliver the performance the application actually needs without paying for unnecessary capability. Customers can explore available options in the STARPATH GLOBAL\u00a0Satellite Imagery Catalog<\/a>.<\/p>\n