FLOWERING DETECTION OF CANOLA USING DYNAMIC TIME WARPING AND SENTINEL-1 TIME SERIES IMAGES

This paper explored the possibility of using the Sentinel-1 GRD Synthetic Aperture Radar (SAR) data to detect the flowering period of canola. A parameter called pseudo scattering entropy (Hc) was used in this research for canola flowering period detection. The relationship between this parameter and the flowering period of canola was analyzed in this paper. Dynamic Time Warping (DTW) was used to extract the relative flowering periods of canola in two regions. Sentinel-2 data was used for auxiliary validation to compensate for the lack of field data. The results indicate that the Hc has good correlation with canola phenology and is stable in different regions, and DTW has great application value in phenology alignment. Meanwhile, SAR data with higher temporal resolution can improve accuracy.