Proposal of a Geometric Calibration Method Using Sparse Recovery to Remove Linear Array Push-Broom Sensor Bias.

Area of Science:

• Remote Sensing
• Geospatial Analysis
• Satellite Imaging

Background:

• The rational function model (RFM) is standard for Earth observation satellite calibration but struggles with short-period errors and limited ground control points (GCPs).
• Existing RFM methods show reduced calibration performance with image distortions from factors like attitude jitter.
• Insufficient or unevenly distributed GCPs further degrade the precision of RFM-based geometric calibration.

Purpose of the Study:

• To develop an improved geometric calibration method for linear array push-broom sensors.
• To address the limitations of the rational function model (RFM) concerning short-period errors and GCP requirements.
• To enhance the accuracy and efficiency of satellite image geometric calibration.

Main Methods:

• A novel geometric calibration method utilizing sparse recovery to mitigate linear array push-broom sensor bias.
• Approximation of imaging process errors to equivalent bias angles for calibration.
• Identification and removal of short-period errors by detecting periodic wavy patterns in image data.
• Significant reduction in the number and distribution requirements for ground control points (GCPs) through sparse recovery.

Main Results:

• The proposed sparse recovery method effectively removes short-period errors, improving calibration accuracy.
• The method significantly reduces the dependency on a large number of ground control points (GCPs).
• Experimental validation using data from Earth Observing 1 (EO-1) and Advanced Land Observing Satellite (ALOS) confirmed the method's effectiveness.

Conclusions:

• The developed geometric calibration method offers a robust solution for sensor calibration in Earth observation satellites.
• The technique enhances calibration precision by effectively handling short-period errors and optimizing GCP usage.
• The method demonstrates broad applicability and effectiveness across different satellite platforms like EO-1 and ALOS.