Measurement of classify typical planting medicinal Fritillaria Pallidiflora Schrenk with WorldView-2 extraction problem imagery

Abstract

Aiming at the problem of the anomalous spectrum of WordView-2 high-resolution remote sensing image, this paper explores the method of extracting and interpreting Fritillaria Pallidiflora Schrenk information effectively, then provides reference for the use of similar problem images. Programming customized April 24, 2013 WorldView-2 Remote Sensing Image, based on the texture feature, PCA and texture features and object-oriented classification method were used to classify the whole area of study area and the sub-shell, and the range of Fritillaria Pallidiflora Schrenk was extracted, and the results were obtained by validating the sample set. Verification. The four classification methods showed different classification ability. Among them, the classification accuracy of the classification area was 43.49% and the kappa coefficient was 0.19, which showed good classification ability compared with other methods. For the Fritillaria, the result of the four classification method showed significant differences. The classification of the object-oriented classification method was 61.45% and the kappa coefficient was 0.45, which showed good classification ability compared with other classification methods.