Detection and evaluation of mixed pixels in Landsat agricultural scenesA major problem area encountered in the identification and estimation of agricultural crop proportions in Landsat imagery involves the large proportion of the pixels which are mixed pixels, whose spectral response is influenced by more than one ground cover type. The development of methods for the detection and estimation of crop proportions in mixed pixels is presently reported. The procedure designated CASCADE, based on the estimation of the gradient image for the detection of mixed pixels, considers the consequences of a linear mixing model and is found to provide a method for the allocation of mixed pixels to the surrounding homogeneous region.
Document ID
19840032890
Acquisition Source
Legacy CDMS
Document Type
Conference Paper
Authors
Merickel, M. B. (Lockheed Engineering and Management Services Co., Inc. Houston, TX, United States)
Lundgren, J. C. (Lockheed Engineering and Management Services Co., Inc. Houston, TX, United States)
Lennington, R. K. (Lockheed Engineering and Management Services Co., Inc. Houston, TX, United States)