The XVIII International Conference on Mine Surveying

177 / 2023-08-31 17:29:40

A novel real-time registration approach of remote sensing image with multi-source feature factors a case study in eastern China

Multi-source feature factor;Remote sensing image registration;Feature extraction;Feature matching.

Due to the large geometric and radiation distortion in the public area of remote sensing images obtained in different periods, there are some problems such as single object and uneven feature distribution in the public area. Therefore, an MSAC algorithm based on multi-source feature factor fusion and strong resistance is proposed for real-time registration of remote sensing images. The experimental results show that the multi-source feature factor fusion can effectively improve the accuracy of RMSE registration, and the MSAC algorithm can reduce the mismatching rate of registration. This method can deal with complex remote sensing image data such as multi-scene, low overlap rate and large rotation Angle, and effectively ensure the quality of real-time registration and stitching.