Validating the Crop Identification Capability of the Spectral Variance at Key Stages (SVKS) Computed via an Object Self-Reference Combined Algorithm

Crop-distribution information constitutes the premise of precise management for crop cultivation. Euclidean distance and spectral angle mapper algorithms (ED SAM) mostly use similarity difference metric (SSDM) to determine variance associated with spatial location distribution acquisition. These methods are relatively insensitive shape or amplitude variation must reconstruct a reference curve representing entire class, possibly resulting in notable indeterminacy ultimate results. Few studies utilize these compute time define new index identification—namely, at key stages (SVKS)—even though this temporal characteristic could be helpful identification. To integrate advantages sensibility avoid reconstructing curve, an object self-reference combined algorithm comprising ED SAM (CES) was proposed SVKS. objectively validate crop-identification capability SVKS-CES (SVKS computed via CES), SVKS-ED ED), SVKS-SAM SAM), five (SI) types were selected comparison example maize The results indicated that ranges can characterize greater interclass separability attained better identification accuracy compared other indexes. In particular, SVKS-CES2 provided greatest best PA (92.73%), UA (100.00%), OA (98.30%) Compared performance SI, SVKS separability, but more non-maize fields incorrectly identified as usage. Owning accuracy-improvement SVKS-CES, omission commission errors obviously reduced utilization SI. findings suggest application is expected further spread