Xiaoying Lian

Xiaoying Lian

Auto-generated profile, based on publicly available information

Xiaoying Lian

State Key Laboratory of Infrared Physics, Shanghai Institute of Technical Physics, Chinese Academy of Sciences, Shanghai, CHINA

Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
10 February 2025 in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Verified
Journal Article
Open access
XLXiaoying Lian
Xiaoying Lian
JLJingyang Li
Jingyang Li
DSDexin Sun
Dexin Sun
Hyperspectral imaging technology enables the simultaneous acquisition of geometric, radiometric, and spectral information about targets, making it an important tool for Earth-based remote sensing and target detection. However, owing to the inherent nonuniformity of detector responses and the influences of the imaging chain, the image data captured by push-broom hyperspectral imagers inevitably exhibit nonuniformity. This nonuniformity includes not only discrepancies between different pixels along the spatial dimension but also varying response deviations among different pixels across the spectral dimension. If these discrepancies are not effectively corrected, they can significantly impact the quantitative applications of hyperspectral image data. To address the limitations of existing methods—which often struggle to fully utilize spatial and spectral information, require manual parameter adjustments for each hyperspectral image, involve cumbersome coefficient calculations, and have limited generalization capabilities—this paper presents a scene-based nonuniformity correction method for push-broom hyperspectral imagers. The method can adaptively identify uniform regions in different scenes to derive correction coefficients and can utilize spectral-dimensional information for constraints, effectively ensuring satisfactory calibration results in both spatial and spectral dimensions while enabling automatic execution on each scene image. The proposed method shows strong adaptability to different hyperspectral imagers, scenes, and spectral bands, with a quality improvement factor of 42.55 and an average relative deviation of less than 1.04%. It preserves spatial detail without significant loss, maintains the integrity of the spectral curve, and requires less run time. This work provides a solid technical foundation for the real-time processing of terrestrial and on-orbit hyperspectral images.