Remote sensing methods for surveying and extracting abandoned farmlands

Wenjun Li,Shixing Wang,Yi Zhou,Q. Xu,Futao Wang,Yu Han
DOI: https://doi.org/10.1109/CISP.2012.6469955
2012-10-01
Abstract:The increase of agricultural land abandonment is one of the main changes in land use in China over the recent years. In China, the abandonment phenomenon has become one of the most serious problems in the last 20 years. Abandoned farmlands make the land in a vicious circle. Preventing the farmland from abandonment is an urgent issue for China. Thus, abandoned farmland survey is very important; it can provide scientific evidence for food security, land use planning and economic development of China. The conventional investigation approach of abandonment status is field survey. It cannot meet the demand of large area because of the lack of efficiency and timeliness. Remote sensing technology has advantages of large spatial observing area, high real-time, wide spectrum ranges and all weather imaging capability. It has worked successfully in many fields, but was rarely applied in terms of investigation and assessment of arable land abandonment. In this work, Xinyang City in Henan Province was selected as a study area. SPOT-5 images and MODIS NDVI data were used for surveying and extracting the abandoned farmland. This paper introduces the theories, processes and results of three methods for abstracting abandoned farmland on the basic of remote sensing, assesses the efficiency and accuracy of them, and compares the merits and demerits of them, sums up the workable technique for the study area.
What problem does this paper attempt to address?