Abstract:Fractional vegetation cover (FVC) is an important index of crop growth status, as well as one of the major factors affecting crop photosynthesis, transpiration and water use efficiency. Currently, there are some problems that satellite remote sensing technology widely used is difficult to meet the requirement of fractional vegetation cover extraction in field scale for the low temporal and spatial resolution, the extraction of vegetation coverage based on artificial ground image is time consuming and laborious, the operating cost is high, and the remote sensing image acquired by the unmanned aerial vehicle (UAV) remote sensing system without integrated gimbal is geometrically distorted. To address the issues above, a UAV multi-spectral remote sensing image acquisition system integrated gimbal and position and orientation system（POS）data acquisition modules was developed, which had the ability to acquire the reflection information for red, green and near-infrared bands between 520nm and 920nm. Taking winter wheat as an example, UAV flying experiments were conducted in different growing stages, covering over-wintering period, jointing stage, flag leaf stage and heading date, with 55m flying height and 2.2cm multispectral image resolution. A rapid FVC extraction method was proposed, combining supervised classification with vegetation index histogram, by which the classification thresholds of normalized difference vegetation index (NDVI), soil-adjusted vegetation index (SAVI) and modified soil-adjusted vegetation index (MSAVI) for field wheat were obtained with the value of 0.4756, 0.7056 and 0.6350, respectively. The FVC reference was extracted based on the visible light remote sensing image with a high spatial resolution of 0.8cm captured synchronously with multi-spectral image. The results showed that the fractional vegetation cover of winter wheat could be extracted by multi-spectrum remote sensing technology and vegetation index method with good accuracy. Compared with SAVI and MSAVI, the extraction result based on NDVI classification threshold was the most accurate with the smallest absolute error. The use of UAV carrying a multi-spectral camera and vegetation index threshold method provided a new way to extract fractional vegetation cover, which had certain reference value for the extraction of fractional vegetation cover in field scale.
