In this paper, we present a face cartoon generation technique that does not have any limitation of the number of sample images. The technique calculates each value of the face image's characteristics and selects several similar sample part images. Then, the technique synthesizes the part cartoons by applying morphing.