accuracy is noticeable that all methods which

accuracy analysis of different metaheuristic methodsis calculated based on the following equation:Where e is the responsiveness rate, Ac0 is the accuracy of the segmented image withoutnoise, and Aciis the accuracy of the segmented image with i%of noise. Normallythe responsiveness rate is between 0 and 1 where one is the best and zero the worst.Figure 12 shows the responsiveness of the different methods with respect to noiseadding, it is noticeable that all methods which depend on GA have the highest responsivenessrate. In addition, the responsiveness rates is different between differentalgorithms in the low noise and stable for GA and cooperative algorithms Analyzing the graph it is obvious that adding more noise drives SOMs and FCM toward the local optimal solution, and therefore the responsiveness is lower than GA.Analysis of the metaheuristic methods responsiveness to noiseDiscussionThe segmentation process is known to be a very hard NP problem and the result ofthis process is very important for the success of essential steps in image processingsuch as image classification, object detection and object recognition. Many disciplinesdepend on the success of this important step in image processing. Exampleof these are the following: 1-traffic planning and control such as video surveillance58, 2- land use planning such as mapping and natural resources management 58,3- automation such as robotics and object extractions60, 61, 4- Biometrics such asface and finger prints recognition 62, 5- medical image processing for the sake oflocating tumors and virtual surgery simulation 63-65, in assessing the environment66, 67, and it is used in many other disciplines which are not the objective of thischapter. This large variety of use in different disciplines makes this process a criticalone and forces the scientific community to look for more effective methods thatcan enhance the results of image segmentation than the existing one such as supervisedparametric methods. The main concern nowadays is to increase the accuracyof the results and the efficiency with respect to the speed of processing. Combiningboth objectives seems to be a difficult task especially with the appearance of morecomplex type of images such as multicomponent images one of these images is thesatellite images . This chapter has introduced several types of satellites images andillustrated the difference between them based on the spatial, spectral, temporal andradiometric resolution. These difference complicates the tasks of image processing