Manual interpretation of medium and high spatial resolution aerial imagery for forestry has evolved in the late 20th century. For this type of application, manual delineation and enumeration of the tree crowns on the aerial images are the main activities. In this context, a new research branch aiming at fully or partly replacing human image interpreter by a “seeing computer” was born. For this new research branch, the application of image processing techniques in delineating the tree crowns on the images has been greatly explored and many delineation techniques has emerged. The launching of IKONOS satellite has opened up another option of the input data. A project utilizing the resolution of the IKONOS satellite image for palm oil tree delineation and enumeration purposes in place of aerial images (which was once a common input data for this application) pursued at Universiti Malaysia Sarawak (UNIMAS) has successfully produced a prototype model for tree crown enumeration. The use of this model, compared with the result of manual delineation, gave an average enumeration accuracy of 92.10%, with the percentage of correctly spotted tree crowns being 84.08%.
- Reduces reliance on manpower needed to work in the estate (site survey).
- Provides accurate and efficient data acquisition through GIS technology.
- Able to determine the actual size of the estates and their layout with the GIS database.
- Enhances efficiency and effectiveness in oil palm estate management (productivity measurement, future planning) with the integration of satellite images and GIS technology
The present product has the advantages of providing ease in data acquisition and accuracy in cost budgeting. Inaccuracy in detecting oil palm density, hectare measurement of planted areas and the measurement of the estate size will cause a higher loss in overall productivity and future maintenance planning will require a larger resources allocation. We aim to market the product to countries such as Malaysia & Indonesia.
Development Process And Product Commercialization
The present product was fully developed by the Centre of Excellence for Image Analysis and Spatial Technology. The product was supported and to be commercialized fully via UNIMAS Holdings Sdn Bhd, a business arm of UNIMAS.