Abstract: In the image recognition there are scale transform, rotation transform and translation transform, so looking for the features to be invariant to all these transforms becomes very important . In this paper, the method ofaffine transform is used to extract the invariant features of the image contour of several types of airplanes. The imagecontours of several airplanes are acquired by using Cellular Neural Networks. And the inwariant features of airplanecontours are stored in a database. During the recognition of an airplane type the first step is to calculate the invariantfeatures of the real-time acquired image contour of the airplane. The second step is to calculate the correlation coefficient between the real-time acquired image contour invariant features of the target airplane and that of every airplaneimage stored in the database previously. The maximum correlation coefficient indicates the type of the airplane to berecognized.

Key words: invariant features, affine transformation, image recognition