OilPalm_Dataset.id
Description
All images are from the Palangka Raya oil palm plantation area, Kalimantan Tengah, Indonesia. Data was collected in August 2022. It can be seen in the Table, the data is separated into several small sizes, categories, types, and amounts of data. If you use this data, please cite our journal, available at SSRN: https://ssrn.com/abstract=4596338 or http://dx.doi.org/10.2139/ssrn.4596338
Files
Steps to reproduce
1. Image acquisition using drones (unmanned aircraft) in the Palangka Raya area KM.13, Central Kalimantan, Indonesia. Data from drones with RGB sensors, a total of 30 data. The height of the drone is 100 meters (from ground level), the resulting resolution is 16.9 centimeters per pixel, and the resolution of each photo is 4056x3040 pixels. Length and width of each photo in meters: P = 154 meters and L = 116 meters. 2. Image Preprocessing: combining contrast sharpening and noise reduction, conversion color space. 3. Anotation with the ENVI software, for 4 class features (1: land/road, 2: Other plants, 3: ditch/wallow, 4: oil palm), and 1,2 and 3 represent 3 class features (1: land/road, 2: Other plants, 3: palm oil). 4. Augmentation (Horizontal, Vertical, Combining Horizontal and Vertical) with Python 4. Cropping the image to sizes 256x256 and 512x512 with Python 5. Image mask and image label with Python