Skip to Main content Skip to Navigation
Journal articles

Classification of soil and vegetation by kernel Fisher and kernel PCA

Abstract : Precision Agriculture is concerned with all sort of within-field variability, spatially and temporally, that reduces the efficacy of agronomic practices applied in a uniform way all over the field. Because of these sources of heterogeneity, uniform management actions strongly reduce the efficiency of the resource input to the crop (i.e., fertilization, water) or for the agrochemicals used for pest control (i.e. herbicide). In particular, weed plants are one of these sources of variability for the crop, as they occur in patches in the field. Detecting the location, size and internal density of these patches, along with identification of main weed species involved, open the way to a site-specific weed control strategy, where only patches of weeds would receive the appropriate herbicide (type and dose). Herein, the first stage of recognition method of vegetal species, the classification of soil and vegetation, is described and is based upon the kernel Fisher discriminant method (KFDM) and on Kernel Principal Analysis (KPCA).
Document type :
Journal articles
Complete list of metadata
Contributor : Cybu Recherche Connect in order to contact the contributor
Submitted on : Tuesday, October 27, 2020 - 10:45:38 AM
Last modification on : Friday, August 5, 2022 - 2:46:00 PM

Links full text




Michel Chapron. Classification of soil and vegetation by kernel Fisher and kernel PCA. Pattern Recognition and Image Analysis, 2013, 462, ⟨10.1134/S1054661811020192⟩. ⟨hal-02979488⟩



Record views