Unmanned Airship Based Multiple Spectrum Image Mosaic with SIFT Feature Matching

Unmanned airship-based remote sensing is widely used in agriculture, remote sensing, environmental monitoring and detection of military camouflage works. A multi-spectral remote sensing image mosaic technique with SIFT feature matching is proposed to deal with images from weak wind, poor stability unmanned airship. Firstly, the characteristics of multi-spectrum image from unmanned airship are analyzed. A multi-spectrum image mosaic flowchart is designed. Secondly, a SIFT based multi-spectrum image mosaic algorithm is presented. SIFT feature vectors with spectral information is designed. The spectrum information is adopted to improve robustness of mosaic algorithm. The BBF algorithm and RANSAC methods are presented for coarse, fine matching processing and error removal. The optimal transformation matrix from SIFT feature matching calculation is adopted to realize image mosaic. Finally, mosaic experiment results of multi-spectrum images from unmanned airship show that the algorithm can obtain a large number of matching feature points which can obtain stable transformation matrix for further image mosaic. The mosaic accuracy and effectiveness can meet the needs of interpretation.