Abstract
Abstract. Imagery geometry models (IGMs) of the high-resolution satellite images (HRSIs) are always of great interest in the photogrammetry and remote sensing community for the raising new kinds of sensors and imaging systems. Especially the generalized sensor models (GSMs) have been widely used for positioning of satellite images, and the accuracy are already validated. Since Back propagation (BP) neural network is a better choice for the two key reasons of the replacement of physical sensor models by generalized sensor models, numerous mathematical estimations for every specialized sensor, and secret equations of the IGMs. Experiments are carried out to test the approximation accuracy of the new generalized sensor model. And the experimental results show that, the BP neural network is of extremely high accuracy for satellite imagery photogrammetric restitution.
Highlights
Geospatial information of high-resolution satellite images (HRSIs) are extracted via the functional relationship between the image space and object space, which is usually called imagery geometry model (IGM) (Tao and Hu, 2001; Yue et al, 2013; Yavari et al, 2013)
The IGMs can be mainly categorized into two types: rigorous sensor models (RSM), which are physical sensor models of satellite ephemeris data (Radhadevi et al, 1998; Jeong and Kim, 2015), and generalized sensor models, such as rational polynomial coefficients (RPCs) model provided by the imagery vendor(Fraser et al, 2002; Grodecki and Dial, 2003; Zhang, G. and Yuan, X., 2006)
Rational function model (RFM) parameters (RPCs) instead of RSMs are provided to the end users as the standard auxiliary data attaching to the imagery by the commercial vendors
Summary
Geospatial information of high-resolution satellite images (HRSIs) are extracted via the functional relationship between the image space and object space, which is usually called imagery geometry model (IGM) (Tao and Hu, 2001; Yue et al, 2013; Yavari et al, 2013). The IGMs can be mainly categorized into two types: rigorous sensor models (RSM), which are physical sensor models of satellite ephemeris data (Radhadevi et al, 1998; Jeong and Kim, 2015), and generalized sensor models, such as rational polynomial coefficients (RPCs) model provided by the imagery vendor(Fraser et al, 2002; Grodecki and Dial, 2003; Zhang, G. and Yuan, X., 2006). The experiments of BP neural network fitting capability of the satellite physical sensor models, and the accuracy comparison experiment of BP neural network IGM and RFM are presented and analyzed. The experimental results demonstrate that, the BP neural network achieve a high fitting accuracy instead of the classical RFM algorithm, and can be applied to HRSIs geopositioning
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
