Search

Land-cover data are used for the enforcement of quota restrictions, yield estimations and agricultural statistics. With the launch of the European Remote sensing Satellite (ERS-1) the first long-term spaceborne synthetic aperture radar (SAR) has become available for the mapping of land-cover. This paper assesses the capability of ERS-1 SAR precision data to (i) discriminate between the crop types for land-cover inventory purposes at (ii) the earliest stage in the growing season, using field-based classification. The objectives were tested for an agricultural region in The Netherlands where 12 crop types are found. Fourteen ERS-1 SAR images were available for this area, covering the 1992 growing season between May and November. The field-based classification yielded an overall classification accuracy of 80 per cent with the optimal data set. The stage at which the crop types could be assessed is crop dependent. Most crop types could not be distinguished before the month of August.

A methodology is proposed for a post-processing assessment of the performance of training set selections and classification procedures. This paper focuses on the necessity and usefulness of a critical validation of classification results. A statistical procedure is proposed to evaluate the algorithms for the numerical classification of images. The approach is based on the derivation of performance indicators from measurements of signature separability and thresholding analysis. Although these measurements are not new in image processing techniques, they are used in this study in an original way for the comparison of outputs resulting from different classification criteria. The theoretical description of the suggested method is followed by their practical application to a case-study for mapping crop coefficients in an irrigation district.