Skip to main content
ARS Home » Research » Publications at this Location » Publication #256963
Title: Unsupervised linear unmixing of hyperspectral image for crop yield estimation

Author
item LUO, BIN - Grenoble Institute Of Technology
item Yang, Chenghai
item CHANUSSOT, JOCELYN - Grenoble Institute Of Technology

Submitted to: Institute of Electrical and Electronics Engineers Proceedings Fuzzy Systems
Publication Type: Proceedings
Publication Acceptance Date: 8/20/2010
Publication Date: 12/20/2010
Citation: Luo, B., Yang, C., Chanussot, J. 2010. Unsupervised linear unmixing of hyperspectral image for crop yield estimation. Institute of Electrical and Electronics Engineers Proceedings Fuzzy Systems. CDROM.

Interpretive Summary: Hyperspectral imagery can be used for estimating crop yield and its spatial variation. This paper describes a new spectral technique, unsupervised linear unmixing, to estimate crop yield from hyperspectral images. The spectral unmixing scheme was used to convert the hyperspectral images of two grain sorghum fields to multiple abundance maps, which were then related to yield data. Statistical analysis showed crop yield was significantly related to the abundance map corresponding to the crop. These results indicate that unsupervised linear unmixing is a useful tool to convert hyperspectral imagery of crop fields to relative yield maps.

Technical Abstract: Multispectral and hyperspectral imagery are often used for estimating crop yield. This paper describes an unsupervised unmixing scheme of hyperspectral images to estimate crop yield. From the hyperspectral images, the endmembers and their abundance maps are computed by unsupervised unmixing. The abundance maps are then compared with the crop yield data. The results showed there existed high correlations between crop yield data and the abundance maps of the endmembers corresponding to the crop, even though the unsupervised unmixing scheme does not need any prior knowledge to extract the abundance maps of the endmembers.