The sign is recognize is useful to the handicapped peoples to speech with others in world. The sign recognize is difficult to process because the hand shape identification is difficult. In recent year there are many research will goes to recognize the sign easily. It is not only help to the handicapped peoples it is use to produce Robert for understand our signs in future and also it helps to give the input of the public computers in sign. In our process we recognize the Arabic words from it is sign. In the Arabic get 28 alphabets and several words so we take and analysis this word easily. In our process we use the Principle component analysis for extract the feature and then use K-Nearest neighbor classifier to classify the features. Here first we preprocess using the Gaussian filter and smooth the input image and then subtract the background of sign image using the morphological method. Then the PCA was extracted to the sign only that type of feature only the sign. It helps to increase the robustness of classification. The PCA extract the feature based on the surface so the sign is easily identify. Then KNN classifier gives the better analysis result to recognize ward for same of the given input sign.