The Trimble eCognition Oil Palm Application (OPA) enables a comprehensive platform for analysing aerial data. Simple upload of orthophoto and digital surface model gets users on their way. Upload block boundaries from existing records or use our boundary creation tools to identify an area of analysis, then let the solution automatically detect and index the total number of palms using cutting-edge deep learning technologies.
Individual palm GNSS coordinates, palm height, health indicators and crowne size. This data can be easily exported to an existing GIS database for further analysis. This data can further assist plantation managers, sustainability professionals and finance professionals in understanding fertilizer requirements, yield potential, modeling of greenhouse gas emissions and other aspects of the environmental footprint, as well as asset valuation for accounting and audit purposes.
The eCognition OPA supports a variety of data inputs formats from different UAV sensors to be understood, not simply through automatic palm count, but also through the automated calculation of tree addressing. The Application’s tree addressing tools support multiple planting layouts in flat plantation environments. The user will receive a tree address for each individual tree made up of the block ID, row number and tree number along with ghost tree locations. Ghost trees are identified as points were additional trees could be planted according to neighboring tree distributions - these potential trees are also reflected in the tree address.
Climate change, disease, poor irrigation and fertilization all impact palm health. The OPA provides a tool to automate the detection of stressed palms and other anomalies in a data set, based on NDVI parameters and sensitivities defined by the user. Similarly, location and condition of palms detected as anomalous can be provided to field teams for further inspection and verification, expediting the management of poor palm health and disease spread.
Trimble eCognition OPA provides plantation managers tools to assess land use productivity by automatically detecting land cover not occupied by palms. Following detection of unused land, or “gaps”, users can model the total potential of this unused area by analysing how many palms could occupy the unused land. Users can also assess the land use by looking at the number of ghost trees within each gap polygon - both vector layer deliverables are available for additional GIS-analysis.
Both tools work to identify wasted land that could be better utilized, as well as help the planning and decision making process for new planting.
Plantations are large and complicated places to work. To assist in providing actionable information to stakeholders that do not interact with geospatial data on a daily basis, Trimble eCognition OPA includes a report generation tool that provides a simple template, basic statistics and an overview of analysis parameters. In addition, users are provided with Excel-ready statistics on an individual tree level.