Introduce and Compare Several Segmentation Softwar in the Data With High Spatial Resolution Remote Sensing

Common primary method of information extraction from remote sensing data based pixel classification which is designed for images with medium spatial resolution. With the development of remote sensing data and images with high spatial resolution, created has problems in the use of pixel basis. Object - based method reduces spectral changes in elementary and creating homogeneous segment, to consider the possibility of spectral characteristics and to have delivered geometrically. The first stage of this method is that segmentation. The purpose of segmentation creation of separate phenomena and homogeneous regions that are remote sensing images. Segmentation before classification as an important influence on the results of this classification for quality of the parts of this process in any way shall be evaluated. The specific method there are for measuring the quality and quantity of the segmented. Segmentation results are of particular importance in the evaluation of the quality parameters. Important factor in the use of segmentation techniques segmentation quality, so in the first place must have an overview about the software segment. To do so, the 4-band image with high spatial resolution as Ultracam - D is used. The quantity segment of the segmentation by value evaluation index AFI was placed examined. segmentation by visual observation of workflow software (spring5¡ ecognition8.7¡ envi5¡ saga2.1) And quantitative evaluation results showed that software spring5 and ecognition 8.7 Due to the use of different parameters such scale in the segmentation phenomena have been better isolated. For best results tested in different places and on different images with high spatial resolution and with different segmentation index. There are many different methods for segmentation of the performance of each segment depends on the quality and scale is desired.