Fast-growing plantation area extracting based on satellite images: a case of Leizhou City in Southeast China

Fast-growing plantation planting is especially prevalent in the late 10 years in Southeast China, going with the debate on its positive and negative influence. Mastering the status of non-natural fast-growing plantations is essential for both national and local government's decision-makers. This paper presents the methodology used to detect fast-growing plantation in the extent of Leizhou city in southeast China based on CCD/CBERS time-series images. This methodology is on the basis of the assumption that fast-growing plantation's rotation period is 5 years. NDVI differencing as a change detection method is selected to detect the bald land period, which will be considered to be fast-growing plantation. The result indicates that NDVI differencing method is feasible in extracting fast-growing plantation. The area of fast-growing plantation in Leizhou city is 225.54 square kilometers, distributed all over the city. Plantation scale in west coast and northern part connected with Suixi city is larger than southern part connected with Xuwen city. East coast has the least amount of distribution. The paper proposes that texture and shape analysis will help improving extraction precision.