There are different methods of Image segmentation. Basically, there are two main types; parallel segmentation and sequential segmentation. In parallel segmentation, the pro-cessing that was done at each point of the picture did not depend on results already obtained at other points. In sequential segmentation, when processing a point, advantage of the results at previously processed points can be used. In parallel approach, the same computation must be performed at every point of the picture. When using the sequential approach, simple inexpensive computations can be fulfilled to detect possible object points. Some sequential segmentation techniques are edge and curve tracking, region growing and partitioning. Partitioning of picture into re-gions can be constructed by starting with an initial parti-tion and allowing both merging and splitting of regions. The initial partition may be trivial or it can be the result of previous segmentation process. The criteria for merging and splitting should depend, if possible, on how well the parti-tion conforms to a model for the given class of pictures. In the absence of such a model one can use general purpose criteria based on such factors as 'region homogenity, distinc-tiveness of adjacent regions, edge strength, size, shape simplicity and soon just as in the case of region growing. The analysis was done for satellite imagery of location(Minea-Egypt). Four classes are considered; each class is specified by its lower and upper limit of its grey level.