Object-based change detection
Characterizations of land-cover dynamics are among the most important applications of Earth observation data, providing insights into management, policy and science. Recent progress in remote sensing and associated digital image processing offers unprecedented opportunities to detect changes in land cover more accurately over increasingly large areas, with diminishing costs and processing time. The advent of high-spatial-resolution remote-sensing imagery further provides opportunities to apply change detection with object-based image analysis (OBIA), that is, object-based change detection (OBCD). When compared with the traditional pixel-based change paradigm, OBCD has the ability to improve the identification of changes for the geographic entities found over a given landscape. In this article, we present an overview of the main issues in change detection, followed by the motivations for using OBCD as compared to pixel-based approaches. We also discuss the challenges caused by the use of objects in change detection and provide a conceptual overview of solutions, which are followed by a detailed review of current OBCD algorithms. In particular, OBCD offers unique approaches and methods for exploiting high-spatial-resolution imagery, to capture meaningful detailed change information in a systematic and repeatable manner, corresponding to a wide range of information needs.
Since the advent of satellite-based Earth observation, land-cover change detection has been a major driver of developments in the analysis of remotely sensed data (Anuta and Bauer 1973, Anderson 1977, Nelson 1983, Singh 1989, Aplin 2004, Coppin et al. 2004, Lu et al. 2004). More recently, high-spatial-resolution imagery has been available from commercial operators, providing unique opportunities for detailed characterization and monitoring of forest ecosystems (Wulder et al. 2004, 2008c, Hay et al. 2005, Falkowski et al. 2009) and urban areas (Herold et al. 2002, Hay et al. 2010) and additional applications developed to address the increasingly detailed information needs (Castilla et al. 2008, Chen et al. 2011). Land-cover change refers to variations in the state or type of physical materials on the Earthrsquo;s surface, such as forests, grass, water,etc., which can be directly observed using remote-sensing techniques (Fisher et al. 2005). As human-induced changes occur at an increasingly rapid pace, and as Earth observation data become ubiquitous, remote-sensing-based monitoring systems are expected to play further crucial roles in environmental policy and decision-making.
Accurate monitoring of land cover is a matter of utmost importance in many different fields. Satelliteor airborne-based monitoring of the Earthrsquo;s surface provides information on the interactions between anthropogenic and environmental phenomena, providing the foundation to use natural resources better (Lu et al. 2004). It enables refined policy development and the capacity to address otherwise inaccessible science questions (Cohen and Goward 2004). Remote-sensing change detection, defined by Singh (1989) as lsquo;the process of identifying differences in the state of an object or phenomenon by observing it at different timesrsquo;, provides a means to study and understand the patterns and processes of ecosystems at a range of geographical and temporal scales. While the knowledge of land-cover conditions at a given point in time is important, the dynamics or trends related to specific change conditions offer unique and often important insights, ranging from natural disaster management to atmospheric pollution dispersion. Indeed, remotely sensed imagery is an important source of data available to characterize change systematically and consistently in terrestrial ecosystems over time (Coops et al. 2006).