2. Image Fusion
Image fusion integrates both spatial and spectral data to hold the superior characteristics of multi-sensor images and improve the knowledge of scene. Therefore, the fused images could improve the accuracy image classification and helps the feature extraction and recognition. The image fusion can be divided into tow classes : spatial domain method and spectral domain method. The last method is used in most application, scum as color space transformation. In this paper, the I HS ( Intensity-Hue-Saturation ) model will be used as a color space and the image fusion is done as follows:

1. The RGB color space of OPS images is transformed to the I HS model [3]:

2. The different gray value of pixel in the black-white of two SAR images (g1 and g2) are added into OPS images intensity:
I' = I+ (g₁-g2)                 (4)
The last term of the above equation is the different of before and during flood. The flood area will be emphasized and non- flood area will be depressed. Adding this term to intensity component in I HS mode means transferring of flood area data to OPS image.

3. The I HS model is inversely transformed to the RGB space and ready to classify in the further

3. Neural network classification
In this paper, the multi-layer perceptron ( MLP) neural network based on back propagation ( BP) algorithm is used as classifier, which consists of set of nodes arranged in multiple layers with connection only between node in adjacent layer by weights. The input information are presented at input layer as the input vector. And the output vector is the processed information, that were retrieved at the output layer. A schematic of a three-layer MLP model is shown in Fig 4 and using in this paper

Fig .4 The 3 -layer ( MLP modle of neural network.
The input and output of the node I in hidden layer of MLP neural network, according to BP algorithm [4], are :


Where, W_ij: the Weight of connection from node I to node j:
B_i : The numerical value called bias.
f ;The activation function

In this work the nonlinear function, sigmoid function given in eq. (6), is used to determine the output state: