The Matching Fund PETARUNG is one of the research projects led by a lecturer that develops software for semi-automatic spatial planning map compilation. In our department, one cohort is divided into several groups, each assigned a study area. In these study areas, we are tasked with detecting land use using image classification. Subsequently, we compare the classification results with the actual field observations. Due to the possibility of significant differences between the classified data and the field results, we find it necessary to conduct accuracy testing
I conduct surveys at locations marked as sample points. The field sampling involves the use of the simple random sampling method, marking several land use points visible from the CSRT image. This map serves as a reference during field surveys. To facilitate field surveys, we create routes on Google My Maps
We conducted land use detection using a hybrid classification performed in ENVI. This classification employs supervised classification with maximum likelihood, resulting in a definitive map of land use classification
These are the results of the survey we conducted. We detected only 6 changes from the hybrid classification performed in ENVI.