Use of Artificial Neural Networks and Remote Sensing Technologies in Determining the Plant Cover in Urban Areas

The subject of the study is to analyze the urban vegetation using remote sensing. 525 measurement points were determined from the study area. These measurement points were chosen at random, not within rules. Land observations were carried out for the determined points. A database of reflection and plant index values of Sentinel II satellite images belonging to the same points was created. The resulting data was transformed into the “Artificial Neural Network” structure with the Neural Designer software and processed here to create prediction values and significance coefficients for six types of analysis. In order to test the accuracy of these analyzes, 1.774 control points were determined and measurements were made on these points. For the data types of 1.774 points, estimates were created using the "Data Output Function". The highest values given in the estimates suggested by the software and the current structure at the control points were compared and the success rates of the analysis systems were eveluated. As a result of the analysis; It has been determined that the reflection values of vegetation change from season to season. Reflection average values were highest in summer and lowest in winter. It was concluded that the reflection values and plant indexes can be estimated with artificial neural networks using plant properties that are easily obtained by vegetation analysis. Suggestions were given for the data types and methods to be used in the studies to be performed at the end of the study.