Semantic Web Techniques for Extracting and Analyzing of Cropland Abandonment in Hilly Areas

Introduction

The growing issue of farmland abandonment in rural regions is escalating, marked by an increase in fallow land and a reluctance among villagers to engage in agriculture. This trend is notably prevalent in impoverished areas. Driven by urbanization and industrial advancement, an exodus of the youthful workforce from rural to urban areas is seeking better employment opportunities. This migration exacerbates the labor shortage and accelerates demographic aging in rural hilly regions (Tian et al., 2010), leading to a pronounced neglect of arable land (Xu et al., 2019). The neglect of farmland is intricately linked to the basic livelihood of farmers and to national food security and has adverse effects on the ecological environment (MacDonald et al., 2000). A thorough examination and detailed analysis of the phenomena, historical context, and forward trends of farmland abandonment are imperative. The realization of these methods requires effective and accurate extraction of abandoned cropland as the premise and basis. Therefore, it is highly important to study the information extraction and identification of abandoned cropland.

Currently, the domestic sphere faces a significant shortfall in statistics pertaining to idle lands (G. Xiao et al., 2019), a gap that substantially impedes advanced understanding and research into the phenomenon of farmland abandonment, thus challenging the development of effective strategies and measures. The methodologies employing traditional interviews and field surveys for analyzing the status quo of farmland neglect are notably inefficient and labor intensive (Zhang et al., 2014; Y. He et al., 2020). These approaches also suffer from limitations in accuracy and comprehensiveness, particularly in pinpointing the exact geographic locations of abandoned agricultural lands and their evolving patterns. In this context, remote-sensing technology has emerged as a pivotal tool, offering distinct advantages for monitoring the situation and identifying trends in the abandonment of arable lands in hilly regions. Remote-sensing technology affords unparalleled efficiency, extensive coverage, and multidimensional monitoring capabilities, enabling the swift identification and observation of abandoned croplands across vast areas. This technology uncovers the temporal and spatial dynamics of land abandonment, offering a robust scientific foundation and support for policy and measure development. Recent studies on monitoring fallow agricultural lands have proliferated, distinguishing themselves primarily through the selection of remote-sensing image data sources and extraction methodologies. Among the key satellite-imagery data employed by researchers for this purpose are Landsat, MODIS(Moderate Resolution Imaging Spectroradiometer), and Sentinel-2, which are instrumental in the detection and analysis of neglected agricultural lands.