Tree Counting System in Agricultural Plots
One of the important data used for agricultural managemenet is the data of tree counting in agricultural plots or designated areas. This data can be used to estimate the amount of crops produced per unit. One of the technologies applied for tree counting is the use of remote sensing data such as satellite images which allow for an overview of large area with no need of field survey. Moreover, the cost of survey per unit is not high.
GISTDA has developed a method for tree counting on satellite images by initially using the image of oil palm plots, because oil palm is Thailand’s important economic crop. The main data used for this method is the high resolution multi-spectral satellite image. Another important data used in the tree counting process is agricultural plot boundaries, which can be obtained from shape file in the database or be determined through the use of hands. The operation of the developed system is based on spatial statistics and digital image processing.
In order to check the efficiency of the developed method, GISTDA had used the images of oil palm plots in Phang-nga and Phuket provinces, which are all data from Quickbird and WorldView-2 satellites. From the test, the tree count results are found to be over 90 percent correct, compared with the count by human eyes. The advantage of the developed method is that it can be used with images from other receptors such as aerial images obtained by Unmanned Aerial Vehicle (UAV)
Moreover, to improve the system to be more automatic and can be applied in GISTDA’s production system, the next step is being planned to develop an automatic method for determining the area and boundaries of agricultural plots. This will help to reduce the operators’ work process in determining agricultural plot boundaries on satellite images. The developed procedure can be used for efficient management or monitoring of agricultural products comprehensively or in large-scale provincial areas.
