Author_Institution :
Coll. of Comput. Sci. & Inf. Technol., China Three Gorges Univ., Yichang, China
Abstract :
The UAV remote sensing images due to its access to convenient, high resolution, low cost, low risk advantage has been more widely studied and applied to various field. However, due to the characteristics of complex, gray inconsistent, the larger distortion of UAV remote sensing image texture itself, which results the extraction of its characteristics to become one of the difficulties. In this paper, we take the UAV remote sensing image as the study object, analyzes several mainstream feature extraction algorithm. Through experiments, from the aspects of extraction rate, number, stability and distribution of the features, we analyzed and compared the performance, the advantages and disadvantages of various algorithms quantitatively and qualitatively, then proposed the feature extraction strategies for UAV remote sensing image, which have important significance for processing UAV remote sensing image.
Keywords :
autonomous aerial vehicles; feature extraction; image resolution; image texture; remote sensing; UAV remote sensing image processing; UAV remote sensing image texture distortion; feature distribution; feature extraction rate; feature number; feature point extraction methods; feature stability; image resolution; unmanned aerial vehicles; Algorithm design and analysis; Correlation; Data mining; Feature extraction; Remote sensing; Robustness; Stability analysis; UAV; feature extraction; remote sensing images;
