Translator Disclaimer
15 November 2007 Performance evaluation of pattern painting camouflage based on the texture characteristic of multi-spectral images
Author Affiliations +
Proceedings Volume 6787, MIPPR 2007: Multispectral Image Processing; 67871Y (2007) https://doi.org/10.1117/12.750578
Event: International Symposium on Multispectral Image Processing and Pattern Recognition, 2007, Wuhan, China
Abstract
Multi-spectral images have a high degree of spectral resolution. Through distilling the spectrum features as well as texture features between target and background, and calculating the Mahalanobis Distance of spectrum features as well as texture features data vector, then we can accordingly analyze the blend performance of pattern painting camouflage quantitatively. Utilizing the generating pixel spectrum curve function by the tool of Z Profile of ENVI, the remote sensing imaging processing system, we can draw the spectrum curve of target and typical background, and choose the spectral bands with small data relativity and obvious spectrum value difference, combining with principal component analysis. According to the result of spectrum analysis, we can establish the gray level co-occurrence of chosen image or area, and get texture characteristic, then take the Mahalanobis Distance calculated by spectrum features as well as texture features data vector between target and background of typical spectrum band images as the foundation of performance evaluation of pattern painting camouflage.
© (2007) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Xiaopeng Li, Xuliang Lv, Youjun Wang, Chuan'an Cui, and Yv Huang "Performance evaluation of pattern painting camouflage based on the texture characteristic of multi-spectral images", Proc. SPIE 6787, MIPPR 2007: Multispectral Image Processing, 67871Y (15 November 2007); https://doi.org/10.1117/12.750578
PROCEEDINGS
6 PAGES


SHARE
Advertisement
Advertisement
Back to Top